ars.els-cdn.com · Web viewFunnel plot of vitamin D supplementation and CVD and all-cause mortality risk.

Supplementary Material

SUPPLEMENTARY TABLES Supplementary Table 1. Official Global, N America & Europe Recommendation on Supplements for CVD & Cancer (2004-2016) Supplementary Table 2. Search Strategy for vitamin and mineral supplements for cardiovascular diseases & all-cause mortalitySupplementary Table 3. Characteristics of included RCT studies for CVD and all-cause mortalitySupplementary Table 4. GRADE assessment for vitamin DSupplementary Table 5. GRADE assessment for vitamin ASupplementary Table 6. GRADE assessment for beta-caroteneSupplementary Table 7. GRADE assessment for antioxidantsSupplementary Table 8. GRADE assessment for vitamin ESupplementary Table 9. GRADE assessment for vitamin CSupplementary Table 10. GRADE assessment for seleniumSupplementary Table 11. GRADE assessment for vitamin B-complexSupplementary Table 12. GRADE assessment for folic acidSupplementary Table 13. GRADE assessment for niacin (B3)Supplementary Table 14. GRADE assessment for vitamin B6Supplementary Table 15. GRADE assessment for calciumSupplementary Table 16. GRADE assessment for ironSupplementary Table 17. GRADE assessment for multivitaminSupplementary Table 18. GRADE assessment for calcium and vitamin DSupplementary Table 19. Summary of the meta-analayses results for all-cause mortality, CVD mortality, total CVD risk and other significant assocations

SUPPLEMENTARY FIGURES Supplementary Figure 1. Consort statementSupplementary Figure 2. Risk of bias graph for supplements and CVD and total mortality.Supplementary Figure 3. Summary of the pooled effect estimates of RCTs assessing the relationship between vitamin D supplementation and CVD and all cause mortality risk.Supplementary Figure 4. Forest plot of vitamin D supplementation and total CVD risk.Supplementary Figure 5. Forest plot of vitamin D supplementation and total CHD riskSupplementary Figure 6. Forest plot of vitamin D supplementation and MI risk.Supplementary Figure 7. Forest plot of vitamin D supplementation and stroke risk.Supplementary Figure 8. Forest plot of vitamin D supplementation and CVD mortality risk.Supplementary Figure 9. Forest plot of vitamin D supplementation and CHD mortality risk.Supplementary Figure 10. Forest plot of vitamin D supplementation and MI mortality risk. Supplementary Figure 11. Forest plot of vitamin D supplementation and stroke mortality risk.

1

Supplementary Figure 12. Forest plot of vitamin D supplementation and all-cause mortality risk.Supplementary Figure 13. Funnel plot of vitamin D supplementation and CVD and all-cause mortality risk.Supplementary Figure 14. Summary of the pooled effect estimates of RCTs assessing the relationship between vitamin A supplementation and CVD and all cause mortality risk.Supplementary Figure 15. Forest plot of vitamin A supplementation and all-cause mortality risk.Supplementary Figure 16. Summary of the pooled effect estimates of RCTs assessing the relationship between beta-carotene supplementation and CVD and all cause mortality risk.Supplementary Figure 17. Forest plot of beta-carotene supplementation and total CVD risk.Supplementary Figure 18. Forest plot of beta-carotene supplementation and total CHD risk.Supplementary Figure 19. Forest plot of beta-carotene supplementation and MI risk.Supplementary Figure 20. Forest plot of beta-carotene supplementation and stroke risk.Supplementary Figure 21. Forest plot of beta-carotene supplementation and CVD mortality risk.Supplementary Figure 22. Forest plot of beta-carotene supplementation and CHD mortality risk.Supplementary Figure 23. Forest plot of beta-carotene supplementation and MI mortality risk.Supplementary Figure 24. Forest plot of beta-carotene supplementation and stroke mortality risk.Supplementary Figure 25. Forest plot of beta-carotene supplementation and all-cause mortality risk.Supplementary Figure 26. Summary of the pooled effect estimates of RCTs assessing the relationship between antioxidants supplementation and CVD and all cause mortality risk.Supplementary Figure 27. Forest plot of antioxidants supplementation and total CVD risk.Supplementary Figure 28. Forest plot of antioxidants supplementation and total CHD risk.Supplementary Figure 29. Forest plot of antioxidants supplementation and MI risk.Supplementary Figure 30. Forest plot of antioxidants supplementation and stroke risk.Supplementary Figure 31. Forest plot of antioxidants supplementation and CVD mortality risk.Supplementary Figure 32. Forest plot of antioxidants supplementation and CHD mortality risk.Supplementary Figure 33. Forest plot of antioxidants supplementation and MI mortality risk. Supplementary Figure 34. Forest plot of antioxidants supplementation and stroke mortality risk.Supplementary Figure 35. Forest plot of antioxidants supplementation and all-cause mortality risk.Supplementary Figure 36. Funnel plot of antioxidants supplementation and all-cause mortality risk.Supplementary Figure 37. Summary of the pooled effect estimates of RCTs assessing the relationship between vitamin E supplementation and CVD and all cause mortality risk.Supplementary Figure 38. Forest plot of vitamin E supplementation and total CVD risk.

2

Supplementary Figure 39. Forest plot of vitamin E supplementation and total CHD risk.Supplementary Figure 40. Forest plot of vitamin E supplementation and MI risk.Supplementary Figure 41. Forest plot of vitamin E supplementation and stroke risk.Supplementary Figure 42. Forest plot of vitamin E supplementation and CVD mortality risk.Supplementary Figure 43. Forest plot of vitamin E supplementation and CHD mortality risk.Supplementary Figure 44. Forest plot of vitamin E supplementation and MI mortality risk.Supplementary Figure 45. Forest plot of vitamin E supplementation and stroke mortality risk.Supplementary Figure 46. Forest plot of vitamin E supplementation and all-cause mortality risk.Supplementary Figure 47. Funnel plot of vitamin E supplementation and CVD and all-cause mortality risk.Supplementary Figure 48. Summary of the pooled effect estimates of RCTs assessing the relationship between vitamin C supplementation and CVD and all cause mortality risk.Supplementary Figure 49. Forest plot of vitamin C supplementation and total CVD risk.Supplementary Figure 50. Forest plot of vitamin C supplementation and total CHD risk.Supplementary Figure 51. Forest plot of vitamin C supplementation and MI risk.Supplementary Figure 52. Forest plot of vitamin C supplementation and stroke risk.Supplementary Figure 53. Forest plot of vitamin C supplementation and CVD mortality risk.Supplementary Figure 54. Forest plot of vitamin C supplementation and MI mortality risk.Supplementary Figure 55. Forest plot of vitamin C supplementation and stroke mortality risk.Supplementary Figure 56. Forest plot of vitamin C supplementation and all-cause mortality risk.Supplementary Figure 57. Summary of the pooled effect estimates of RCTs assessing the relationship between selenium supplementation and CVD and all cause mortality risk.Supplementary Figure 58. Forest plot of selenium supplementation and total CVD risk.Supplementary Figure 59. Forest plot of selenium supplementation and total CHD risk.Supplementary Figure 60. Forest plot of selenium supplementation and MI risk.Supplementary Figure 61. Forest plot of selenium supplementation and stroke risk.Supplementary Figure 62. Forest plot of selenium supplementation and CVD mortality risk.Supplementary Figure 63. Forest plot of selenium supplementation and MI mortality risk.Supplementary Figure 64. Forest plot of selenium supplementation and stroke mortality risk.Supplementary Figure 65. Forest plot of selenium supplementation and all-cause mortality risk.Supplementary Figure 66. Summary of the pooled effect estimates of RCTs assessing the relationship between vitamin B complex supplementation and CVD and all cause mortality risk.Supplementary Figure 67. Forest plot of vitamin B complex supplementation and total CVD risk.Supplementary Figure 68. Forest plot of vitamin B complex supplementation and total CHD risk.Supplementary Figure 69. Forest plot of vitamin B complex supplementation and MI risk.Supplementary Figure 70. Forest plot of vitamin B complex supplementation and stroke risk.Supplementary Figure 71. Forest plot of vitamin B complex supplementation and CVD mortality risk.

3

Supplementary Figure 72. Forest plot of vitamin B complex supplementation and CHD mortality risk.Supplementary Figure 73. Forest plot of vitamin B complex supplementation and MI mortality risk.Supplementary Figure 74. Forest plot of vitamin B complex supplementation and stroke mortality risk.Supplementary Figure 75. Forest plot of vitamin B complex supplementation and all-cause mortality risk.Supplementary Figure 76. Funnel plot of vitamin B complex supplementation and CVD risk and all-cause mortality risk.Supplementary Figure 77. Summary of the pooled effect estimates of RCTs assessing the relationship between folic acid supplementation and CVD and all cause mortality risk.Supplementary Figure 78. Forest plot of folic acid supplementation and total CVD risk.Supplementary Figure 79. Forest plot of folic acid supplementation and total CHD risk.Supplementary Figure 80. Forest plot of folic acid supplementation and MI risk.Supplementary Figure 81. Forest plot of folic acid supplementation and stroke risk.Supplementary Figure 82. Forest plot of folic acid supplementation and CVD mortality risk.Supplementary Figure 83. Forest plot of folic acid supplementation and MI mortality risk.Supplementary Figure 84. Forest plot of folic acid supplementation and stroke mortality risk.Supplementary Figure 85. Forest plot of folic acid supplementation and all-cause mortality risk.Supplementary Figure 86. Funnel plot of folic acid supplementation and all-cause mortality risk.Supplementary Figure 87. Summary of the pooled effect estimates of RCTs assessing the relationship between niacin (B3) supplementation and CVD and all cause mortality risk.Supplementary Figure 88. Forest plot of niacin (B3) supplementation and total CVD risk.Supplementary Figure 89. Forest plot of niacin (B3) supplementation and total CHD risk.Supplementary Figure 90. Forest plot of niacin (B3) supplementation and MI risk.Supplementary Figure 91. Forest plot of niacin (B3) supplementation and stroke risk.Supplementary Figure 92. Forest plot of niacin (B3) supplementation and CVD mortality risk.Supplementary Figure 93. Forest plot of niacin (B3) supplementation and CHD mortality risk.Supplementary Figure 94. Forest plot of niacin (B3) supplementation and all-cause mortality risk.Supplementary Figure 95. Summary of the pooled effect estimates of RCTs assessing the relationship between vitamin B6 supplementation and CVD and all cause mortality risk.Supplementary Figure 96. Forest plot of vitamin B6 supplementation and total CVD risk.Supplementary Figure 97. Forest plot of vitamin B6 supplementation and MI risk.Supplementary Figure 98. Forest plot of vitamin B6 supplementation and stroke risk.Supplementary Figure 99. Forest plot of vitamin B6 supplementation and MI mortality risk.Supplementary Figure 100. Forest plot of vitamin B6 supplementation and all-cause mortality risk.Supplementary Figure 101. Summary of the pooled effect estimates of RCTs assessing the relationship between calcium supplementation and CVD and all cause mortality risk.Supplementary Figure 102. Forest plot of calcium supplementation and total CVD risk.Supplementary Figure 103. Forest plot of calcium supplementation and total CHD risk.

4

Supplementary Figure 104. Forest plot of calcium supplementation and MI risk.Supplementary Figure 105. Forest plot of calcium supplementation and stroke risk.Supplementary Figure 106. Forest plot of calcium supplementation and CVD mortality risk.Supplementary Figure 107. Forest plot of calcium supplementation and MI mortality risk.Supplementary Figure 108. Forest plot of calcium supplementation and stroke mortality risk.Supplementary Figure 109. Forest plot of calcium supplementation and all-cause mortality risk.Supplementary Figure 110. Summary of the pooled effect estimates of RCTs assessing the relationship between iron supplementation and CVD and all cause mortality risk.Supplementary Figure 111. Forest plot of iron supplementation and total CVD risk.Supplementary Figure 112. Forest plot of iron supplementation and MI risk.Supplementary Figure 113. Forest plot of iron supplementation and CVD mortality risk.Supplementary Figure 114. Forest plot of iron supplementation and all-cause mortality risk.Supplementary Figure 115. Summary of the pooled effect estimates of RCTs studies assessing the relationship between multivitamins supplementation and CVD and all cause mortality risk.Supplementary Figure 116. Forest plot of multivitamins supplementation and total CVD risk.Supplementary Figure 117. Forest plot of multivitamins supplementation and MI risk.Supplementary Figure 118. Forest plot of multivitamins supplementation and stroke risk.Supplementary Figure 119. Forest plot of multivitamins supplementation and CVD mortality risk.Supplementary Figure 120. Forest plot of multivitamins supplementation and MI mortality risk.Supplementary Figure 121. Forest plot of multivitamins supplementation and stroke mortality risk.Supplementary Figure 122. Forest plot of multivitamins supplementation and all-cause mortality risk.Supplementary Figure 123. Funnel plot of multivitamin supplementation and all-cause mortality risk.Supplementary Figure 124. Summary of the pooled effect estimates of RCTs assessing the relationship between calcium and vitamin D supplementation and CVD and all cause mortality risk.Supplementary Figure 125. Forest plot of calcium and vitamin D supplementation and total CVD risk.Supplementary Figure 126. Forest plot of calcium and vitamin D supplementation and total CHD risk.Supplementary Figure 127. Forest plot of calcium and vitamin D supplementation and MI risk.Supplementary Figure 128. Forest plot of calcium and vitamin D supplementation and stroke risk.Supplementary Figure 129. Forest plot of calcium and vitamin D supplementation and CVD mortality risk.Supplementary Figure 130. Forest plot of calcium and vitamin D supplementation and CHD mortality risk.

5

Supplementary Figure 131. Forest plot of calcium and vitamin D supplementation and MI mortality risk.Supplementary Figure 132. Forest plot of calcium and vitamin D supplementation and stroke mortality risk.Supplementary Figure 133. Forest plot of calcium and vitamin D supplementation and all-cause mortality risk.Supplementary Figure 134. Funnel plot of vitamin D & calcium supplementation and all-cause mortality risk.

6

Supplementary Table 1. Official Global, N America & Europe Recommendation on Supplements for CVD & Cancer (2004-2016)

Official Body (Ref) Cardiovascular Disease

Date Supplement Recommendation for Vitamins, Minerals & Multivitamins minerals

Standard Diet recommendations (fruit, vegetables, wholegrain cereals = (+))

World Heart Federation (1) 2016 no supplement recommendation (+) low saturated & trans fats, low sodium, fish, nuts, soy, alcohol in moderation

National Cholesterol Education Program ATP III (2,3)

2001 & Update 2004

no supplement recommendation as such possibly nicotinic acid for low HDL-C and high TG

(DGA 2000) saturated fat <7 %, dietary cholesterol <200 mg/d ± plant sterols 2g/d, viscous fiber 10-25g/d. Maximum dietary therapy can achieve 25-30% LDL-C reduction

American Heart Association’s/American College of Cardiology(4)

2013 no supplement recommendation (+) low fat dairy, poultry, fish, legumes non tropical vegetables oils & nuts. Limit sweets, sugar sweetened beverages & red meats.

Canadian Cardiovascular Society(5)

2012 no supplement recommendation but vitamins, minerals or supplements i.e., CoQ10, not recommended even for statin associated myalgia

(+)PUFA, MUFA, n-3 fats from fish avoid trains, limited saturated and total fat <7%, <30% total energy fiber>30g, cholesterol ≤ 200mg/day plus nuts, soy, viscous fibers, plant sterols

Canadian Cardiovascular Society (6)

2015 no supplement recommendation addition of Niacin to statins with lipids at target not recommended due to AIM-HIGH and HPS-2 THRIVE trial lack of benefit and possible harm. With high TG or low HDL-C or LDL-C above target-possible use of niacin

(+)Mediterranean diet. Do NOT recommend n-3 fatty acid supplements. High doses for marine, algal & yeast sources (2-4g/d) may lower raised TG. Replace sat fat with n-3/n-6

7

mixed sources (canola & soy bean) plus plant source of MUFA (olive, canola, nuts & seeds)

European Society of Cardiology/ European Atherosclerosis Society(7)

2011 no supplement recommendation but Nicotinic acid may be used with statin. But the results of HPS-2 THRIVE and AIM-HIGH trials are awaited

(+) legumes (inc. soy & soy protein), fish poultry without skin, skimmed milk. Moderation in refined cereals, processed fruit (dried, jams, canned etc) red meat, seafood, nuts. But “nutraceuticals” (phytosterols, soy protein, viscous fibers, n-3 fatty acids, polycosanols & red yeast rice) can be used as alternatives or in addition to lipid lowering drugs

European Atherosclerosis Society (Statin associated Muscle Symptoms -SAMS consensus statement)(8)

2015 Supplements Not Recommended. Niacin lowers LDL-C by 15-20% but recent large RCTs showed significant adverse events and no CVD benefit when added to statin.Co-Q10 and Vitamin D failed to reduce SAMS. Therefore not advised.Red Yeast Rice 20-30% LDL-C reduction. Need longer term studies

(+)(as in 2011 assumed since no diet details given) Low saturated fat & Avoid trans-fat plus incorporate a portfolio of plant sterol, soy protein , viscous fibers and nuts, appropriate for SAMS alone or with statin or non-statin drug therapy

British Heart Foundation- Health Eating(9)

N/A No supplement Recommendation (+) plus some dairy products, some meat, fish, eggs, beans and non-dairy product small amount of drinks high in fat or sugar, replace saturated fats mono & poly saturated fats

CVD & CancerUS Preventive Services Task Force (10)

2003 revised 2014

No Supplement Recommendation. Too little evidence on harms versus benefits. Except Recommendation AGAINST β-Carotene & vitamin E supplementation, increased cancer

(+) (Endorse dietary guidelines of 2010) plus Fat free and low fat dairy & seafood.

8

risks in smokers with β-Carotene and possible hemorrhagic stroke with vitamin E

CancerWorld Cancer Research Fund International(11)

N/A Supplements not recommended. Do not rely on supplements for Cancer prevention.Meet nutritional needs through diet alone (special situations, e.g., illness, supplements may be of value)

(+) (fruit and non-starchy (colored) vegetables 600g/d plus unprocessed cereals ≥ 25g fiber/d) pulses (legumes) with every meal

American Cancer Society (12)

20121/11/201220162/5/2016 revised

Supplements not recommended. Evidence not clear therefore not advised e.g.↑ Ca = ↓ colon but ↑ prostate; β-Carotene may ↑ lung cancer in smokers. Folate may increase risk of prostate, and colorectal cancer and possibly breast. Selenium not effective not recommended.Vitamin E may even increase prostate cancer risk

(+) plus less red & processed meat maintain healthy weight, limit alcohol

Canadian Cancer Society (13)

2016 No supplement recommendation except Vitamin D 1000 IU autumn oil winter – many benefit and not increase cancer risk

(+) (as in Canada’s food Guide) plus reduce portion size

European Code against cancer 4th Edition (14)

2015 No Supplement recommendation; Dietary supplements are not recommended for cancer prevention (World Cancer Research Fund 2007)

(+) plus reduce body fat avoid energy dense foods, sugary drinks, eat mostly plant foods, limit red meat, alcohol ≤ 2 drinks/d for men ≤ 1 drink /d for women

Cancer Research UK(15) 2014 (updated 24 Oct)

Supplements not recommended for prevention. Evidence for a healthy diet & cancer prevention.Supplements may interact with treatment therefore advice against antioxidant supplement use. But supplements may have a benefit in life prolongation of patients with advanced cancer

(+) plus a healthy diet provides micronutrients

9

Supplementary Table 2. Search Strategy for vitamin and mineral supplements for cardiovascular diseases & all-cause mortality

Databases Search TermsMEDLINE 1. exp Dietary Supplements/ Or supplement*.mp.

2. exp Vitamin D/ Or exp Cholecalciferol/ Or Vitamin D.mp. Or exp Vitamin A/ Or vitamin A.mp. Or retinol.mp. Or exp beta Carotene/ Or alpha carotene.mp. Or exp Antioxidants/ Or antioxidant.mp. Or Vitamin E.mp. Or exp alpha-Tocopherol/ Or tocopherol.mp. Or Vitamin C.mp. Or ascorbic acid.mp. Or ascorbic acid.mp. Or exp Selenium/ Or selenium.mp. Or exp Vitamin B Complex/ Or B complex.mp. Or exp Folic Acid/ Or folate.mp. Or exp Niacinamide/ Or Vitamin B3.mp. Or exp Niacin/ Or exp Vitamin B 6/ Or pyridoxine.mp. Or Vitamin B6.mp. Or exp Calcium/ Or exp Calcium Carbonate/ Or exp Iron/ Or Iron.mp. Or exp Magnesium/ Or magnesium.mp. Or exp Potassium/ Or potassium.mp. Or Zinc.mp. Or exp Zinc Or Multivitamin.mp. Or multi vitamin.mp. Or (exp Calcium/ Or calcium carbonate.mp. AND exp Vitamin D/ Or exp Cholecalciferol/)3. exp Cardiovascular Diseases/ Or exp Myocardial Infarction/ Or exp Stroke/ Or cardiovascular death.mp. Or exp Mortality/ Or all-cause mortality.mp. Or exp Death/ Or death.mp. Or cancer death.mp. Or cancer mortality.mp.4. 1 AND 2 AND 35. limit 4 to (yr="2012 -Current" and (meta analysis or observational study or randomized controlled trial))

PUB MED 1. ("Dietary Supplements" Or supplements) AND ("Vitamin D" Or Cholecalciferol Or "Vitamin A" Or retinol Or "beta-Carotene" Or "alpha-carotene" Or Antioxidants Or "Vitamin E" Or "alpha-Tocopherol" Or tocopherol Or "Vitamin C" Or "ascorbic acid" Or Selenium Or "Vitamin B-Complex" Or "Folic Acid" Or folate Or Niacinamide Or "Vitamin B3" Or Niacin Or "Vitamin B6" Or pyridoxine Or Calcium Or "Calcium Carbonate" Or Iron Or Magnesium Or Potassium Or Zinc Or Multivitamin Or "multi-vitamin" Or (Calcium Or "calcium carbonate" AND "Vitamin D" Or Cholecalciferol)) AND ("Cardiovascular Diseases" Or "Myocardial Infarction" Or Stroke Or "cardiovascular death" Or Mortality Or "all-cause mortality" Or Death Or "cancer death" Or "cancer mortality")2. limit 1 to (yr="2012/01/01 - 2017/12/31" and (meta analysis or observational study or randomized controlled trial))

10

COCHRANE 1. dietary supplements.mp. [mp=title, abstract, full text, keywords, caption text] Orsupplements.mp. [mp=title, abstract, full text, keywords, caption text] "2. vitamin D.mp. [mp=title, abstract, full text, keywords, caption text] Or cholecalciferol.mp. [mp=title, abstract, full text, keywords, caption text] Or Vitamin A.mp. [mp=title, abstract, full text, keywords, caption text] Or retinol.mp. [mp=title, abstract, full text, keywords, caption text] Or beta-carotene.mp. [mp=title, abstract, full text, keywords, caption text] Or alpha-carotene.mp. [mp=title, abstract, full text, keywords, caption text] Or Antioxidant*.mp. [mp=title, abstract, full text, keywords, caption text] Or Vitamin E.mp. [mp=title, abstract, full text, keywords, caption text] Or alpha-tocopherol.mp. [mp=title, abstract, full text, keywords, caption text] Or tocopherol.mp. [mp=title, abstract, full text, keywords, caption text] Or Vitamin C.mp. [mp=title, abstract, full text, keywords, caption text] Or ascorbic acid.mp. [mp=title, abstract, full text, keywords, caption text] Or Selenium.mp. [mp=title, abstract, full text, keywords, caption text] Or B-complex.mp. [mp=title, abstract, full text, keywords, caption text] Or Folate.mp. [mp=title, abstract, full text, keywords, caption text] Or Folic acid.mp. [mp=title, abstract, full text, keywords, caption text] Or niacinamide.mp. [mp=title, abstract, full text, keywords, caption text] Or vitamin B3.mp. [mp=title, abstract, full text, keywords, caption text] Or niacin.mp. [mp=title, abstract, full text, keywords, caption text] Or vitamin B6.mp. [mp=title, abstract, full text, keywords, caption text] Or pyridoxine.mp. [mp=title, abstract, full text, keywords, caption text] Or calcium.mp. [mp=title, abstract, full text, keywords, caption text] Or calcium carbonate.mp. [mp=title, abstract, full text, keywords, caption text] Or iron.mp. [mp=title, abstract, full text, keywords, caption text] Or magnesium.mp. [mp=title, abstract, full text, keywords, caption text] Or potassium.mp. [mp=title, abstract, full text, keywords, caption text] Or zinc.mp. [mp=title, abstract, full text, keywords, caption text] Or multivitamin.mp. [mp=title, abstract, full text, keywords, caption text] Or multi-vitamin.mp. [mp=title, abstract, full text, keywords, caption text]3. cardiovascular diseases.mp. [mp=title, abstract, full text, keywords, caption text] Or stroke.mp. [mp=title, abstract, full text, keywords, caption text] Or myocardial infarction.mp. [mp=title, abstract, full text, keywords, caption text] Or cardiovascular death.mp. [mp=title, abstract, full text, keywords, caption text] Or all-cause mortality.mp. [mp=title, abstract, full text, keywords, caption text] Or death.mp. [mp=title, abstract, full text, keywords, caption text] Or mortality.mp. [mp=title, abstract, full text, keywords, caption text] Or cancer mortality.mp. [mp=title, abstract, full text, keywords, caption text] Or cancer death.mp. [mp=title, abstract, full text, keywords, caption text]4. 1 AND 2 AND 3

11

5. limit 4 to ((meta analysis or randomized controlled trial) and yr="2012 -Current") [Limit not valid; records were retained]

* Original search date for all databases was Oct 21st 2016; update search date for all databases was October 27th 2017

Supplementary Table 3. Characteristics of included RCT studies for CVD and all-cause mortality

Study (reference) Country Health status

Age (years)

Duration (mean, median

or range)

Supplement intake

assessment

Supplement exposure (median or range, units)

Participants(intervention

/control)

Incident cases

(intervention/control)

Outcome Funding source

Cardiovascular disease (CVD) outcomes

Brohult et al., 1973 (16) Sweden Rheumatoid

arthritis

18-69Mean:

521 y N/A

Vitamin D:100000 IU

calciferol per day25/25 1/0 All-cause

mortality Agency

CDPRG 1975(17)

United States CHD 30-64

Mean:74

months

Follow up visit

consisting of an assessment

to drug adherence

Vitamin B3:3.0 gm/day 1119/2789

914/2333 Total CVD

Agency

114/386 MI95/311 Stroke

238/633 CVD mortality

203/535 CHD mortality

273/709 All-cause mortality

Gillilan et al., 1977 (18) USA

Angina pectoris with obstructive coronary disease

57 (mean)

Vitamin E: 189 ± 15.0 daysPlacebo:

192 ± 13.3 days

capsule count and urine

fluorescence test

Vitamin E: 1600IU daily 26/26 2/2 All-cause

mortality Industry

Inkovaara et al., 1983(19)

Finland Healthy 79.5 1 y Serumcalcium,

phosphate, creatinine,

transaminase and alkaline phosphatase

were measured

Vitamin D:1000 IU /d

45/42 0/1 Total CHD

N/A

12/5 Stroke

0/1 CHD mortality

4/2 Stroke mortality


12


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



Calcium+ Vitamin D:3g calcium

carbonate and 1000 IU vitamin D

46/42

¼ Total CHD

9/5 Stroke

1/3 CHD mortality



Corless et al., 1985 (20)

United Kingdom

Hospital for >4 weeks otherwise

healthy82 40 weeks

Plasma 25(OH)D measured

Vitamin D2 9000 IU /d 41/41 8/8 All-cause

mortality Agency

Aloia et al., 1988(21) USA Postmenopausal

with osteoporosis 50-80 2 y Capsule count

Vitamin D:0.50 µg calcitriol/d

with doseescalation if necessary.

Average: 0.8 µg calcitriol/d

12/15 F 1/1 MI Industry

Korpela et al., 1989(22) Finland Patients with

acute MIMean:

57 6 months

Serum selenium

concentration measured

Selenium(selenium-rich

yeast, 100 µg/day)40/41

1/2 MI

N/A

0/4 CVD mortality

McKeown-Eyssen et al.,

1988 (23)Canada

Patients believed to be free of polyps after

removal of at least

one colorectal polyp

58.0 2 yearsrandom urine

sample 400 mg each of vitamins C and E 96/89 4/3 All-cause

mortality Industry

13


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



Ott et al., 1989(24) USA

Postmenopausal with

>2 compression fractures

50-80 2 y N/A.

Vitamin D:0.25 g calcitriol

twice dailywith dose

escalation ifneeded

43/43 F

0/1 MI mortality

Agency

0/1 F All-cause mortality

Gallagher et al., 1990 (25) USA

postmenopausal women with

vertebral fractures

50-78Mean: 69.7

2 y N/A

Vitamin D:From 0.25µg

calcitriol twice daily until a

maximum dose of 1µg twice daily

25/25 F 1/0 F All-cause mortality Industry

Grady et al., 1991 (26) USA Healthy >69 6 months

Serum calcium was

measured and capsule count at each follow

up visit

0.25 g vitamin D3

orally, twice per day

50/48 1/0 All-cause mortality Industry

Chapuy et al., 1992 (27) France Healthy 84 3 y

Taken with a nurse

supervision

1.2g calcium and 20g (800 IU) of

vitamin D3/d1634/1636 F 258/274 All-cause

mortality Agency

Li et al., 1993 – NIT2(28) China

Under nourished population with a

previous cytologic

diagnosis of esophageal squamous dysplasia

40-69 6 y Pill count Multivitamins 1657/1661

22/35Stroke

mortality

Agency


Bogden et al., 1994 (29) USA Healthy 59-85 1 y n/a Multivitamins 33/32 0/1 MI

Agency &

Industry

14


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases




de la Maza et al., 1995 (30) Chile

Decompensated ambulatoryalcoholic cirrhotics

48.7±8.3

(vitamin E)

50.9±9.9

(placebo)

1year Pill counts500 mg vitamin E

daily 37/37 5/4 All-cause mortality Industry

Hamdy et al., 1995 (31)

Europe (Belgium,

France, Netherlands,

UK)

Mild to moderate chronic renal

failure18-81 2 y

Serum calcium levels

were measured

0.25 g alfacalcidol

(vitamin D) every other daywith dose

escalation ifneeded to 1g to maintain serum

calcium concentrations at

upper limit.

89/87 4/1 All-cause mortality N/A

Ooms et al., 1995 (32) Netherlands Healthy >70 2 y

Questionnaire, by pill

counting, and by

measuring serum 250HD

levels

400 IU of vitamin D3/d

177/171 F 11/21 All-cause mortality Agency

15


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



Pike et al.,1995 (33) Canada

Healthy, non-institutionalized

elderly61-79

1 yearBottle checks

and interviews

Multivitamin 24/23 1/0 All-cause mortality Industry

Steiner et al., 1995 (34) USA

Transient ischemic attacks, minor strokes or

residual neurologic

deficits

71.1 2 y n/a α-tocopherol400 IU daily 52/48 3/6 Stroke Industry

Greenberg et al., 1996 – SCPS (35)

USA

Biopsy-proved basal cell or squamous

cell skin cancer patients

63.2 8.2 y

Patients filled questionnaires

about compliance in

taking the capsules

50 mg β-carotene/d 913/892

68/59 CVD mortality

Agency


Hennekens et al., 1996-PHS

(36)USA Healthy 40-84 12 y

Plasma beta carotene

concentrations in blood were

measuredin three

geographic areas.

50 mg β-carotene on alternate days

11036/11035 M

967/972 Total CVD

Agency &

Industry

468/489 MI

367/382 Stroke



Lips et al., 1996 (37) Netherlands Healthy 80 4 y

Tablet count, questionnaire,

and serum levels.

400 IU vitamin D3/d

1291/1287 223/251 All-cause mortality

Agency &

Industry

Omenn et al., USA Smokers, former 45-74 4 y Weighing of 15mg beta- 9420/8894 544/424 All-cause Agency

16


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



1996 – CARET (38)

smokers and workers exposed

to asbestos

returned bottles/self-

report

carotene/25000IU retinol mortality

Stephens et al., 1996 – CHAOS

(39)

United Kingdom

Coronary atherosclerosis 62 510 d Timing of

refill request400 -800 IU vitamin E/d 1035/967

32/54 MI

Agency

27/23 CVD mortality

18/13 MI mortality



Dawson-Hughes et al.,

1997 (40)USA Healthy ≥65 3 y Pill counts

500 mg of calcium plus 700 IU of

vitamin D3

(cholecalciferol)per day


Agency &

Industry

Girodon et al., 1997 (41) France Elderly,

institutionalized 83.75 2 y Given by nurse

Antioxidants: One of 3

preparations:A: 20mg zinc

sulfate + 100µg selenite

B: 120mg ascorbic acid + 6mg β-

carotene + 15mg α-tocopherolC: A + B


Moon et al., 1997 – SKICAP

AK (42)

United States

History of at least 10 actinic

keratosis & at most 2 squamous cell carcinoma or

basal cell carcinoma skin

cancers

Median: 63 3 y Capsule count 25, 000 IU Retinol 1157/1140 24/24

All-cause mortality

Agency &

Industry

Sano et al., 1997 – ADCS 1 USA Alzheimer’s

disease 73 2 y Serum levels 2000 IU of vitamin E/d 85/84 12/12 All-cause

mortality Agency

17


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



(43)Sato et al., 1997

(44) Japan Hemiplegia after stroke

Mean: 68 6 months N/A Vitamin D

1 μg/d 1α(OH)D 45/39 1/1 All-cause mortality N/A

Baeksgaard et al., 1998 (45) Denmark Healthy

(postmenopausal)

58-67Mean: 65.7

2 yNo formal

assessment of compliance

Calcium-Vitamin D:

1000 mg of calcium carbonate with 14 µg (560

IU) of cholecalciferol

80/80 F 0/1 F All-cause mortality Industry

Shoulson et al., 1998 –

DATATOP(46)

USA and Canada

Early Parkinson’s disease 62 8.2 y N/A

2000IU of vitamin E/d with or without 10mg/d of deprenyl


Agency &

Industry

Virtamo et al., 1998 – ATBC

(47)Finland Smokers 50-69 6.1 y Capsule count 50 mg (~75

IU)Vitamin E 6820/6849 M

519/534 M Total CHD

Agency307/296 MI


Baron et al., 1999 – CPPS

(48)USA History of large

bowel adenomas 61 8 y Tablet count3g (1200mg of

elementary calcium)/d

464/466

50/46 Total CHD

Agency12/11 Stroke


Hoffman et al., 1999 (49) USA Coronary disease

patients

35-80Mean: 67.7

6 months Pill counts

Vitamin E:400 mg/day of D-

α-tocopherol acetate

27/12 M 1/0 M All-cause mortality

Agency &

Industry

Girodon et al., 1999 –

MIN.VIT.AOX (50)

France Institutionalized elderly patients

83.9 2 y Pill count and micronutrient

blood concentration

measured

Antioxidants:1) 20mg Zinc + 100µg Selenium

2) 120mg ascorbic acid + 6 mg β-

carotene+ 15mg α-tocopherol

3) 20mg Zinc + 100µg Selenium +


Industry

18


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



120mg ascorbic acid + 6 mg β-

carotene+ 15mg α-tocopherol4) Placebo

GISSI – Prevenzione Investigators

1999 (51)

ItalyRecent

MI(<3 months)

60 3.5 y Supplement refills

300mg vitamin E (synthetic) or 400

IU vitamin E (natural)

5660/5664

571/584 Total CVD

Agency &

industry

83/95 Stroke



228/251 MI mortality


Green et al., 1999 – NSCPT

(52)Australia

Healthy and patients who

previously had skin cancer

20-69 4.5 y

Tablet counts and

concentration of

betacarotene in the skin

30 mg/d of beta-carotene 801/820

6/12 CVD mortality

Agency and

Industry11/21 All-cause mortality

Komulainen et al., 1999 –

OSTPRE (53)Finland

Postmenopausal with 6–24

months elapsed since last

menstruation

47-56 5 y Participant reporting

300IU/d, and 100IU/d of vitamin D during the fifth

year with or without hormone

replacement therapy

112/115 F

1/0 MI

Industry


93mg Calcium and 300 IU vitamin D (cholecalciferol)/d

116/115 F

1/1 MI

2/1 Stroke

19


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



1/0 MI mortality


Krieg et al., 1999 (54) Switzerland Institutionalized

elderly 84.5 2 yTaken with a

nurse supervision

880 IU of vitamin D3 with 1000 mg of

calcium124/124 21/26 All-cause

mortality N/A

Sato et al.,1999 (55) Japan Parkinson’s

disease

65-88Mean: 70.6

18 months N/A

Vitamin D:1 µg 1α(OH)D3

daily43/43 1/0 All-cause

mortality N/A

Boaz et al., 2000 – SPACE

(56)Israel

Patients with stable

hemodialysis with documented history of CVD

64.6 1.4 y

Serum vitamin E

levels were monitored

800 IU of vitamin E daily (natural) 97/99

15/33 Total CVD

Agency

5/17 MI5/6 Stroke

9/15 CVD mortality

2/8 MI mortality


Correa et al., 2000 (57) Colombia

Subjects with confirmed histologic

diagnoses ofmultifocal

nonmetaplastic atrophy and/or

intestinal metaplasia

29-69 72 months Pill counts

2g Vitamin C 130/117 2/0

All-cause mortality Agency

30 mg Beta-Carotene 117/117 2/0

Antioxidants (Vit C and Beta-

carotene)121/117 2/0

Frazao et al., 2000 (58) USA

Hemodialysis patients with

hyperparathyroidism

52 24 weeksAdministered

with hemodialysis

Vitamin D:10µg 1α-

hydroxyvitamin D2

with dosage adjustments to

maintain plasma iPTH levels

71/67

1/2 CHD mortality Agency


20


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



Lee et al., 1999 – WHS (59) USA Healthy 45 Median

2.1 yFollow-up

questionnaires

50 mg of beta-carotene

given on alternatedays.

19939/19937 74/65 Total CVD

Agency &

Industry

Leppälä et al., 2000 – ATBC

(60)Finland Male smokers 50-69 6 y Tablet counts

Antioxidants:50mg/day Vitamin E and 20mg/day

beta-carotene

7118/7153 M

258/252 Stroke

Agency


20mg β-carotene14246/14273

M

554/503 Stroke


Jacobson et al., 2000 (61) USA Heavy smokers ≥18 6 months Bottle checks

Antioxidant: 500 mg

of vitamin C, 400 IU of a-tocopherol,

and 12 mg of b-carotene.

57/55 0/1 All-cause mortality Agency

Salonen et al., 2000 – ASAP

(62)Finland Hypercholesterol

aemia

45-69Mean: 59.7

3 y Tablet counts

Vitamin E:91 mg of d-α-

tocopherol twice daily


Agency &

IndustryVitamin C 130/130 1/1

AREDS Research Group

2001 (63)USA

Healthy (with additional ocular

eligibility criteria)

68 6.3 y Tablet counts

Antioxidants:One of 3

preparations:A: 500mg Vitamin C + 400IU Vitamin

E + 15mg beta-carotene

B: 80mg zinc oxide + 2mg cupric oxide

C: A + B


Agency &

Industry

21


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



Brown et al., 2001 – HATS

(64)

USA, Canada

CAD with stenosis 53 38

months Tablet counts

Antioxidants: 800IU (Vitamin E + 1000mg Vitamin

C + 25mg β-carotene + 100µg

selenium)

42/38

11/12 Total CVD

Agency &

Industry

1/4 MI

2/2 Stroke

0/1 CVD mortality


de Gaetano et al., 2001 – PPP

(65)Italy At least one risk

factor for CVD 64.4 3.6 y Capsule counts

300 mg syntheticα-tocopherol/d 2231/2264

56/53 Total CVD

Industry

22/25 MI22/18 Stroke

22/26 CVD mortality

3/7 MI mortality



Desnuelle et al., 2001 – ALSRT

(66)France

Amyotrophic Lateral Sclerosis

of less than 5 years duration

>18 1 year n/aa –tocopherol 500 mg capsules b.i.d. 144/144 34/35 All-cause

mortality Industry

de Waart et al., 2001 (67) Netherlands Smokers 60 2 y N/A 400 IU vitamin E 109/109 M 0/1 All-cause

mortality

Agency &

industry

Gallagher et al., 2001 – STOP IT

(68)USA

Healthy with noevidence of osteopenia

65-77 3 y Pill count

Vitamin D:0.25 mg calcitriol

twice a dayHRT alone, or

HRT pluscalcitriol

123/123 F

4/3 Total CVD Agency

& industry

4/3 Stroke


You et al., 2001 – SIT (69)

China Baseline H pylori infection

35-69 39 months

Tablet counts and serum

measurements

Antioxidants:500mg Vitamin C

+ 200IU Vitamin E + 15mg β-carotene + 75µg selenium

1706/1705 9/12 CVD mortality

Agency &

industry

22


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



(β-carotene component was

discontinued after 6 months)

Baker et al., 2002 (70)

United Kingdom CHD N/A 1.7 y N/A 5mg/day folic acid 942/940 23/12 Total

CHD N/A

Chapuy et al., 2002 –

Decalyos II (71)France Healthy 85.2 2 years

Taken in the presence of a

nurse toensure

compliance.

1200mg calcium and 20g (800 IU)

of vitamin D3/d393/190 67/43 All-cause

mortality Industry

Chylack et al., 2002 – REACT

(72)US and UK Early age-related

cataract 67 3 y

Plasma concentrations

measured monthly

Antioxidants:β-carotene

[18mg/d], vitamin C [750mg/d], and

vitamin E [600mg/d]

149/148

2/1 MI mortality Agency

and industry9/3 All-cause

mortality

Graat et al., 2002 (73) Netherlands Healthy

≥60Mean: 73.3

15 months

Returned capsules

counted + plasma samples

Vitamin E:200mg/dL of α-

tocopheryl acetate twice daily


Agency and

IndustryMultivitamin 163/153 0/5

Hodis et al., 2002 – VEAPS

(74)USA

LDL-C ≥3.37 mmol/L and no clinical signs of

CVD

56 3 y

Clinic visits and

participant records

Vitamin E:DL-α-tocopherol

400 IU/d177/176

11/14 Total CVD

Agency &

industry

5/6 MI

0/2 Stroke

1/1 MI mortality


HPS Collaborative Group 2002

(75)

UK Patients with past of occlusive

arterial diseases or diabetes

40-80 5 y Capsule counts

Antioxidants (daily): 600 mg

Vitamin E +250 mg Vitamin C

+20 mg β-carotene

10269/10267 2306/2312 Total CVD

Agency &

industry464/467 MI511/518 Stroke878/840 CVD

mortality

23


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases






Meyer et al., 2002 (76) Norway Residents from

nursing homesMean: 84.7 2 y Assessed by

nurses

Vitamin D:10 µg Vitamin D3

daily569/575 169/163 All-cause

mortality

Agency &

industry

Schnyder et al., 2002 – The Swiss Heart Study (77)

Switzerland CHD (undergone angioplasty) 63 1 y

Homocysteine levels were measured

B-complex: folic acid (1mg/d), vitamin B12,

(cyanocobalamin, 400µg/d) and

vitamin B6

(pyridoxine hydrochloride,

10mg/d)

272/281

7/12 MI

Agency3/6 CHD mortality


Waters et al., 2002 – WAVE

(78)

US/Canada

Postmenopausal women with at

least one 15% to 75% coronary

stenosis at baseline coronary

angiography

Mean: 65 5 y Tablet/capsule

counts

1. Antioxidants: 800 IU/d vitamin E

+ 1000 mg/d Vitamin C

2. HRT: 0.625 mg/d conjugated equine estrogens

(+2.5 mg medroxyprogestero

ne acetate if no hysterectomy)

3. 1+2

105/108 F

3/1 MI

Agency1/3 Stroke

4/2 CVD mortality


Wluka et al., 2002 (79) Australia Knee

osteoarthritis

≥ 40Mean:

642 y

Residual capsule counts

Vitamin E:500 IU daily 67/69 1/0 All-cause

mortality Agency

Collins et al., 2003 (80)

USA Patients with a current diagnosis

ofPAD, a history of

intermittent claudication, and

Mean: 67.1

6 months Patient self-report and measured plasma

vitamin E levels

Vitamin E(400 IU of vitamin

E daily)


Agency &

Industry

24


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



an ABI<0.95 at rest and/or <0.85 after exercise

Cooper et al., 2003 (81) Australia Healthy

postmenopausalMean: 56.3 2 y

Tablet counts and diary

review

Vitamin D(10 000 units of

vitamin D2 weekly)

93/94 F 0/1 F All-cause mortality

Agency &

industry

Liem et al., 2003 (82) Netherlands Coronary artery

lesion and MI >18 24 months N/A Folic acid 0.5 mg/d 300/293

¾ MI

Agency

4/3 Stroke

7/9 CVD mortality


Righetti et al., 2003 (83) Italy End stage renal

disease 64 1 y Pill counts15 mg folic acid or 5 mg folic acid per

day51/30 13/11 Total

CVD N/A

Trivedi et al., 2003 (84)

United Kingdom Healthy 74.75 5 y Complete and

return a form

Vitamin D:100,000IU

cholecalciferol once every 4

months

1345/1341

477/503 Total CVD

Agency

224/233 MI

105/101 Stroke


42/49 MI mortality



Virtamo et al., 2003 – ATBC

(85)Finland Smokers 50-69 8 y Tablet count

20mg/d β-carotene 7282/7287 919/851 All-cause mortality Agency

50 mg/d Vitamin E 7286/7287 868/851

25


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



Dukas et al., 2004 (86) Switzerland Healthy

≥70Mean:

7536 weeks

serum uestionnaire

of 25(OH)D3, 1,25

dihydroxyvitamin D (D-hormone) measured

Vitamin D:1 µg of alfacalcidol 192/186 1/1 All-cause

mortality

Agency and

Industry

Harwood et al., 2004 – NoNOF

(87)

United Kingdom

Within 7 days post-surgery for

hip fracture81.2 1 y N/A

Vitamin D: Single injection of 300,000IU

ergocalciferol

38/37 F 7/5 F

All-cause mortality Industry800IU

cholecalciferol plus 1g calcium

carbonate daily

39/37 F 6/5 F

Coburn et al., 2004 (88) USA Chronic kidney

disease 64.5 8 months Blood and urine samples were collected

Vitamin D:iPTH 85 pg/mL

(ng/L), oral doxercalciferol

27/28

0/2 MI

Agency

0/1 MI mortality

Lange et al., 2004 (89)

Germany and

Netherlands

Patients who had undergone successful

coronary stenting

Folate (mean): 61.4

Placebo

(mean): 61.3

2 yMonthly telephone contact

1.2 mg/d folic acid, 48 mg/d vitamin B6, and 60 μg

mg/d vitamin B12

316/320

53/35 CHD

Agency3/2 MI


Larsen et al., 2004 (90) Denmark Elderly

66-103Mean:

743 y

Plasma levels of 25-

hydroxyvitamin D

[25(OH)D] and 1,25-

dihydroxyvitamin D

[1,25(OH)2D]

Calcium-Vitamin D

(1000 mgof elemental

calcium as calcium carbonate and 400

IU (10 µg) of vitamin D3 daily)


Agency &

Industry

26


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



measured

Liem et al., 2004 (91) Netherlands Acute MI 59 1 y Pill counts 5 mg folic acid per

day 140/143

43/45 Total CVD

Industry

8/10 MI

1/0 Stroke

2/2 MI mortality



Taylor et al., 2004 –

ARBITER 2 (92)

United States CHD Mean:

67 3 y Pill countsVitamin B3

(niacin): 1000mg /day

87/80

4/11 Total CVD

Agency &

Industry

2/2 MI0/1 Stroke

1/2 CVD mortality

Tornwall et al., 2004 – ATBC

(93)Finland Men, smokers 50-69 6 y Capsule

counts

Antioxidants (daily):

50 mg α-tocopherol, 20 mg

β-carotene

6781/6849 M

511/534 Total CHD

Agency

289/296 MI


20 mg/d β-carotene 6821/6849 M

548/534 Total CHD

314/296 MI


Manuel-Y-Keenoy et al.,

2004 – DATOR (94)

BelgiumType 1 Diabetes

Mellitus with high cholesterol

21-69 6 months n/a d-α-tocopherol 750 IU 12/12 1/0 All-cause

mortality Industry

McNeil et al., 2004 – VECAT

Australia Early or no cataract

55-88 4 y Blood samples were

Vitamin E (500 IU)/d


Agency &

27


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



(95)

collected to analyze conc.

Of α-tocopherol and total vitamin E

Industry

Meier et al., 2004 (96)

Southwestern Germany

Healthy men and postmenopausal

women33-78 2 y

Blood and urine samples collected in non-fasting

state

Oral cholecalciferol (500 IU/d) and

calcium (500 mg/d)


Meydani et al., 2004 (97) USA

Respiratory Infections-elderly

nursing home residents

≥65 3 y

Pill count, medication

records, quarterly

measurement of plasma E

levels

Vitamin E (200 IU)/d 311/306 39/44 All-cause

mortalityAgency

& Industry

Aloia et al., 2005 (98) USA Healthy black

postmenopausal 50-75 3 y Pill Count 20 µg/day (800 IU) oral vitamin D3

104/104 F ½ All-cause mortality

Agency &

IndustryAvenell et al.,

2005 – MAVIS (99)

ScotlandElderly with or without chronic

illness≥65 1 y Diaries/tablet

counts Multivitamin 456/454 8/4 All-cause mortality Agency

Brazier et al., 2005 (100) France

Ambulatory with vitamin D deficiency

>65 1 y Blood and urine samples

Calcium carbonate 500 mg and

Vitamin D3 400 IU95/97 F

6/5 Total CVD

Industry3/0 MI1/1 Stroke


Flicker et al., 2005 (101) Australia Elderly Mean:

83.4 2 y

Compliance measured

according to the percentage

of thesupplements

removed from the subject’s medicationcontainer

Vitamin D (ergocalciferol,

initially 10,000 IU given once weekly and then 1,000 IU

daily)


28


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



Graf et al., 2005 (102) Germany

Probable or definite ALS treated with

riluzole

59-57 18 months

Blood samples

Vitamin E (α-tocopherol) 5000

mg per day – distributed over the

active day or placebo.


Grant et al., 2005 –

RECORD (103)UK Low trauma

fracture ≥70 24- 62 months

Blood sample taken for

analysis of 25-OH

vitamin D3

and PTH hormone

Vitamin D3 (800 IU per day) and

calcium (1000 mg per day)

1306/133244/39 MI

Agency &

Industry

60/48 Stroke


Calcium only1113/1128W

39/31 Total CHD

26/13 MI


Vitamin D only 2649/2643

78/84 MI

118/104 stroke


Lee et al., 2005 –WHS (104) USA Healthy 45 Mean:

10.1 yFollow-up

questionnaires

600 IU of natural-source vitamin E

on alternate days or placebo

19937/19939 F

482/517 Total CVD

Agency &

Industry

196/195 MI

241/246 Stroke


12/14 MI mortality



Limburg et al., 2005 (105) China

Patients with mild or moderate

esophagealsquamous dysplasia

26-73 10 months

Direct observation

and pill counts

selenomethionine 200 µg/day 90/90

1/0 MI mortality

Agency1/0 All-cause

mortality

29


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



Lonn et al., 2005 – HOPE and HOPE-TOO (106)

Canada

Patients with vascular disease

or diabetes mellitus

≥55 7 y Pill counts Vitamin E (400IU daily) 4761/4780

1022/985 Total CVD

Agency &

Industry

724/686 MI270/246 Stroke



Mooney et al., 2005 (107) USA

Men and women smokers with ≥10

cigarettesper day and

serum cotinine ≥25 ng/mL

>18years old

(mean: 36.8±0

.6)

15 months

Serum measures and

pill counts

500 mg vitamin C and 400 IU vitamin E 142/142

1/0 MI mortality

Agency1/0 All-cause

mortality

Petersen et al, 2005 – ADCS 2

(108)

USA and Canada

Amnestic mild cognitive

impairment55-90 3 years n/a 2000 IU of vitamin

E daily 257/259 5/5 All-cause mortality

Agency and

Industry

Porthouse et al., 2005 (109) UK

One or morerisk factors for

hip fracture

≥70Mean: 76.9

25 months Self-reported

Calcium+Vitamin D

(1000 mgcalcium with 800 IU cholecaliferol

daily)


Agency &

Industry

Potena et al., 2005 (110) Italy

Heart transplantation

recipients

Mean 54 1 y N/A

Folic acid:15 mg/day of 5-

methyltethrahydrofolate or placebo


Agency and

Industry

Sato et al., 2005 (111) Japan Poststroke

hemiplegiaMean: 74.1 2 y

Supplement administered

in the presence of a study nurse

Vitamin D(1000 IU

ergocalciferol daily)

48/48 F 1/2F All-cause mortality N/A

Bairati et al., 2006 (112) Canada Stage I or II head

and neck cancer 62.5 3 y Pill count 400 IU DL- α –tocopherol/d 194/190 65/47 All-cause

mortality Agency

Bonaa et al., 2006 –

NORVIT (113)

Norway MI 63 3.3 y Questionnaires

Vitamin B6 934/943175/172 Total

CVDAgency

& Industry161/153 MI

22/27 Stroke92/89 All-cause

30


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



mortality

61/59 MI mortality

B-complex 937/943

201/172 Total CVD

182/153 MI21/27 Stroke

68/59 MI mortality


Coyne et al., 2006 (114)

USA, Poland

Chronic kidney disease (stage 3&4) for > 2

months

>186 months (treatmen

t 24 weeks)

Serum levels

2 g paricalcitol thrice weekly; 4 ug thrice weekly, 1 ug daily, or 2 ug daily; avg weekly dose: 9.5+/- 3.81 ug or 1.3 to 1.4 ug per

day


Daly et al.,2006 (115) Australia Healthy

50-87Mean: 61.9

2 y

Participant reporting

(compliance calendar)

Calcium+Vitamin D:

400 ml/day of reduced fatUHT milk

containing 1000 mg of calcium plus 800 IU of vitamin

D3

85/82 M 1/0 M All-cause mortality

Agency &

Industry

Law et al.,2006 (116)

United Kingdom Elderly 85 10 m Serum levels

2.5 mg ergocalciferol

every 3 m1762/1955 347/322 All-cause

mortality Agency

Lonn et al., 2006 – HOPE-

TOO (117)

Canada, USA

Brazil, western EU, and Slovakia

High risk (history of CVD, diabetes

or other risk factors)

69 5 y

Interview and pill count evaluation

every 6 months

B-complex (2.5 mg folic acid/d, 50 mg of vitamin B6/d & 1 mg of vitamin B12/d) or placebo

2758/2764

519/547 Total CVD

Agency &

Industry

341/349 MI



Magliano et al., 2006 – MAVET

Australia Smokers > 55Mean:

4 y Phoned every 3 m, capsule

Vitamin E: 500 IU d- α –tocopherol/d

205/204 2/4 CVD mortality

Agency

31


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



(118) 63.5 count 9/17 All-cause mortality

Schleithoff et al., 2006 (119) Germany CHF 55.5 15 m

Bottle counts and serum

levels

50 g vitamin D3/d with 500 mg Ca/d or placebo with

500mg Ca/d


Stranges et al., 2006 – NPC

(120)USA

Patients with histories of

nonmelanoma skin cancers

within the year prior to

randomization

62.3 7.6 yParticipant

reporting and blood samples

200 µg selenium/d 504/500

103/96 Total CVD

Agency

63/59 Total CHD


40/31 CVD mortality

9/8 MI mortality


Zoungas et al., 2006 –

ASFAST (121)

Australia & New

Zealand

End stage renal disease

Avg: 57

(range 24-79)

3.6 N/A 15 mg folic acid or placebo daily 156/159

46/55 Total CVD

Agency

23/19 Total CHD

8/18 Stroke

21/24 CVD mortality


Bolton-Smith et al., 2007 (122) UK 2 y Pill counts

Calcium/Vitamin D(10µg (400 IU) vitamin D3 plus

1000 mg calcium/day)


Agency &

Industry

CLIPS Group 2007 (123) Europe

stage I-II Peripheral

Arterial Disease

Mean : 65.5 2 y Tablet/capsule

counts

Antioxidants:100 mg/d aspirin600 mg/d vitamin E + 250 mg/d

185/18116/11 Total

CVDIndustry9/4 MI

6/5 Stroke

32


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



vitamin C + 20 mg/d β-carotene both

6/3 CVD mortality

2/2 MI mortality



Cole et al., 2007 – AFPPS (124)

USACANADA

History of colorectal adenomas

Mean 57y 10y

Participant reporting and blood samples

1 mg/d folic acid or placebo 516/505

14/8 MIAgency

& Industry

9/5 Stroke


Cook et al., 2007 – WACS

(125)

USA Patients with history of CVD, or at least three

cardiac risk factors

≥40Mean: 60.6

9.4 y Self-report

β-carotene: 50 mg every other day 4084/4087 F

435/399 Total CVD

Agency &

Industry500/499 Total CHD

135/139 MI161/137 Stroke


10/24 MI mortality



Vitamin C: 500 mg daily

4087/4084 F 419/415 Total CVD

510/489 Total CHD

140/134 MI138/160 Stroke


15/19 MI mortality

15/18 Stroke

33


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



mortality


600 IU Vitamin E every other day 4083/4088 F

399/435 Total CVD

491/508 Total CHD

131/143 MI137/151 Stroke


18/16 MI mortality



Antioxidants: vitamin C (500mg daily), E (600IU every other day), and beta-carotene (50mg every other

day)

1020/1022F 133/124F All-cause mortality

Durga et al., 2007 – FACIT

(126)Netherlands High plasma

homocysteine

50 -70Mean:

603 y

Capsule return count,

diary.800 g folic acid/d 405/413 8/4 All-cause

mortality

Agency &

industry

Jamison et al., 2007 – HOST

(127)USA End stage renal

failure65.8(>20) 3.2 y Serum level

B-complex (100 mg vitamin B6, 2

mg vitamin B12 and 40 mg folic acid)

1032/1024129/150 MI Agency

& industry

37/41 Stroke


Lappe et al., 2007 (128) Nebraska Postmenopausal

women >55 4 y Bottle weight 1500mg Ca/d (Calcium) and 446/288 F 3/2 MI Industry

34


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



1000 IU cholesalciferol/d (Vitamin D) or

placebo

6/4 Stroke

Liu et al., 2007 (129) Canada Long-term care

residents 85 19 months Pill counts Multivitamin 375/373 96/97 All-cause

mortality Agency

Lyons et al., 2007 (130)

United Kingdom

Healthy patients and those with

mobility, cognitive,

visual, hearing or communication

impairments

84 3 y

Dosing was supervised by the research

nurse to record

adherence

Vitamin D:Two 1.25 mg

vitamin D2 tablets (ergocalciferol) or matched placebo three times a year (i.e., 100,000 IU,

four-monthly)


Plummer et al., 2007 (131) Venezuela

Healthy patients from a high-risk for gastric cancer

population

35-69 3 y Capsule counts

Antioxidants: 750 mg vitamin C +

600 mg vitamin E + 18 mg β-carotene

daily

990/990

1/0 TotalCVD Agency

and Industry16/11 All-cause

mortality

Smith et al., 2007 (132)

UK(England) Healthy ≥75 3 y

Intramuscular injections by

nurses

Vitamin D:300 000 IU

intramuscular ergocalciferol injection every autumn over 3

years


Agency &

Industry

Vianna et al., 2007 (133) Brazil End stage renal

disease ≥18 2 y Nurse supervision

10 mg folic acid 3 times / week 93/93

9/9 Total CVD

Agency17/21 CVD mortality


Albert et al., 2008 –

WAFACS (134)USA

Post-menopausal with either a

history of CVD or ≥3 coronary

≥ 42 7.3 y

Calendar packs

containing active agents

B-complex (2.5 mg of folic acid, 50 mg of vitamin B6, and 1 mg of vitamin

2721/2721 F205/211 Total

CVD Agency283/280 Total

CHD

35


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



risk factors or placebos mailed

monthly. Annual

participant reporting

B12)/d

65/74 MI

79/69 Stroke

96/94 CVD mortality


Bjorkman et al., 2008 (135) Finland Aged chronically

immobile patients

65-104Mean: 84.5

6 months Nurses supervision

Vitamin D:400 IU or 1200 IUcholecalciferol per

day


Agency &

industry

Bolland et al., 2008 (136)

New Zealand

Healthy post-menopausal

women74 5 y Tablet counts 1g of elemental

calcium 732/739 F

101/54 Total CVD

Agency and

industry


4/1 CVD mortality


CTNS 2008 (137) Italy Early cataract or

no cataract 55-75 9 y Tablet counts Multivitamin 510/51023/31 CVD

mortality Agency77/81 All-cause

mortality

Ebbing et al., 2008-WENBIT

(138)Norway

Patients with suspected CAD

and/or aortic valve stenosis

61.6 3.2 yCapsule count

and interviews

Vitamin B6 772/780

55/58 MI

Agency &

Industry

20/19 Stroke


B-complex 772/780

59/58 MI

11/19 Stroke


Guyton et al., 2008 (139) USA Type 11a or 11b

Hyperlipidemia 18-7924 weeks (part 1 of 64 week

N/A Vitamin B3 (niacin): 2g/day 670/272 1/0 Total

CHD Industry

36


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



study) 1/1 MI0/1 Stroke

Logan et al., 2008 – UK CAP

(140)

United Kingdom,Denmark

History of colorectal adenomas

Mean 57.8 3 y

Participant reporting and

pill count

0.5 mg of folic acid supplementation

with or without 300 mg enteric-coated

aspirin

470/469

3/0 MI Agency &

Industry1/1 Stroke


Milman et al., 2008 (141) Israel

Patients with type 2 diabetes mellitus

>55

Mean 68.7-69.5

18 months

Participant reporting

Vitamin E (natural source d-alpha

tocopherol) 400 IU/d or placebo

726/708

16/33 Total CVD

Agency

7/17 MI

6/11 Stroke

3/5 CVD mortality


Prince et al., 2008 (142) Australia

History of falling in past 12 months

and plasma 25OHD

concentration of < 24.0 ng/mL

70-90 1 y Pill counts

1000 IU/d of ergocalciferol with

or without 1000 mg/d of calcium as

calcium citrate

151/151 F2/3 MI Agency

& Industry

3/3 Stroke


Reid et al., 2008 (143)

New Zealand Healthy

> 40 (average 57)

2 y

A staff member

dispensed the study

medication into numbered

containers.

600 mg calcium citrate/d, 1200 mg calcium citrate/d,

or placebo

108/107 M

2/0 MI

Agency


Sesso et al., 2008 – PHS II

(144)USA

Healthy and also allowed: MI,

stroke, or cancer 64.4 Mean 8 y Participant reporting

Vitamin E (400 IU synthetic_-

tocopherol) every other day with or without vitamin C

3659/3653 M


& Industry

107/144 MI

133/113 Stroke

37


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



(500 mg synthetic ascorbic acid).


Antioxidants (vitamin E and

vitamin C)

3656/3653 M

310/316 Total CVD

133/144 MI104/113 Stroke


Vitamin C 500 mg 3673/3653 M

309/316 Total CVD

127/144 MI114/113 Stroke


Zhu et al., 2008 – CAIFOS

(145)Australia Elderly living in

a sunny climate 70 -80 5y

Tablet count every 12 months.

Follow-up assessments at 1, 3, and 5 yr at the same time of year

Calcium + Vitamin D:

1200mg/d calcium with placebo with 1000 IU/d vitamin

D2

39/41 F 0/2 All-cause mortality Agency

Anker et al., 2009 – FAIR-

HF (146)USA Chronic heart

failure67.6 years 24 weeks Patients

reporting200 mg of

intravenous iron 304/155

21/22 Total CVD

Industry2/3 MI

4/4 CVD mortality


Hodis et al., 2009 – BVAIT

(147)USA

Men and postmenopausal

women with fasting tHcy≥8.5 μmol/L and no

clinical

40-89 3.1 y Pill count and blood samples

B-complex (folic acid 5 mg +

vitamin B12 0.4 mg + vitamin B6 50 mg) or placebo

254/252


& Industry0/2 All-cause

mortality

38


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



signs/symptoms

Imasa et al., 2009 (148) Phillipines

Unstable angina or NSTEMI in the previous 2

weeks.

Mean 59.1-59.6

6 months

Compliance rate was

measure in both treatment

and control

B-complex (1mg folic acid, 400 µg vitamin B12, and

10 mg vitamin B6) or placebo daily

118/125

15/13 Total CHD

Agency0/1 Stroke


LaCroix et al., 2009 – WHI

(149)USA Healthy

(postmenopausal)

50-79Mean: 62.4

7 yweighing

returned pill bottles during clinic visits

Calcium+Vitamin D

(1000 mg of elemental calcium

as calcium carbonate with 400 IU of vitamin D3

daily)

18176/18106 F

226/244 F CVD mortality

Agency &

Industry130/128 F CHD mortality

54/60 F Stroke mortality

Lippman et al., 2009 –

SELECT (150)

USA,Canada,

Puerto Rico

No prior prostate-cancer diagnosis,

serum prostate specific antigen 4

ng/ml or less

50-75+ Median: 5.46 y (range,

4.17 – 7. 33 y).

Pill counts and serum

levels

Selenium 200 μg/d 8856/8910 M

1080/1050 Total CVD

Agency &

Industry

82/100 Stroke




Vitamin E400 IU/d

8863/8910 M 1034/1050 Total CVD

70/92 Stroke


9/8 Stroke

39


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



mortality


Selenium and Vitamin E

8904/8910M

1041/1050 Total CVD

111/100 Stroke




Saposnik et al., 2009 – HOPE 2

(151)

Canada, USA,Brazil,

western EU, and Slovakia

History of CHD, CVD, or

peripheral arterial disease, or

diabetes mellitus and at least 1

additional CVD risk factor

Mean 68.9 5 y

Participant reporting and

pill count

B-complex (2.5 mg of folic acid, 50 mg of vitamin B6, and

1 mg of vitamin B12) or placebo

2758/2764

111/147 Stroke Agency &

Industry


Sang et al., 2009 (152) China

Coronary artery disease Mean:

7112

months N/A

10 mg/d atorvastatin +

extended-releaseniacin initially 500 mg/d for 30 days, then 1000 mg/d


Wu et al., 2009 – NHS/HPFS

(153)USA

History of colorectal adenoma

50-78Mean 64.1

months

Capsule count and blood samples to

assess adherence

1mg/d of folic acid or placebo

338/334

6/1 MI

Agency &

Industry

4/3 Stroke

0/3 CVD mortality


Armitage et al., 2010 –

SEARCH (154)UK History of MI 64 6.7 y

Adherencewas assessed by reviewing the calendar-

packedtablets

remaining

B-complex (2 mg folic acid + 1 mg vitamin B12/d) or matching placebo

6033/6031

1537/1493 Total CVD

Agency &

Industry

1229/1185 Total CHD

431/429 MI

269/265 Stroke

40


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases





185/170 MI mortality



Chailurkit et al., 2010 (155) Thailand Postmenopausal

w/o osteoporosis 60-97 2 y Tablet count at each visit

500 mg elemental Ca/d 201/196 F


& Industry1/1 W All-cause

mortality

de Zeeuw et al., 2010 – VITAL

(156)

USA, Europe

Type 2 Diabetes Mellitus,

oncentrati & consuming ACE inhibitors for ≥3

months

> 20 24 weeks

Collect blood sample 5x for

plasma paricalcitrol

concentration.

1 μg paricalcitol & 2 μg paricalcitol/d 95/93

1/0 Total CHD

Industry2/0 MI1/0 Stroke


Galan et al., 2010 –

SU.FOL.OM3(157)

France

History of MI, unstable angina,

or ischaemic stroke

45-80 4.7 y

Compliance with treatment

was self-reported in the

biannual questionnaires

(or by telephone interview).

B-complex (560ug 5-

methyltetrahydrofolate, 3mg B6, 20ug

B12) daily

1242/1259

75/82 Total CVD

Agency &

Industry

49/55 Total CHD

28/32 MI

35/48 Stroke

26/14 CVD mortality


Heinz et al., 2010 (158)

Germany End stage renal disease

20-80 6 y Dosing was supervised by the research

B-complex :1 tablet contained 2.5 mg folic acid,

327/323 83/98 Total CVD

Agency &

Industry20/19 MI

41


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



nurse to record

adherence

25ug cobalamin,& 10 mg vitamin B6 (2 tablets were taken 3 times per

11/15 Stroke

37/32 CHD mortality


Hercberg et al., 2010 –

SU.VI.MAX (159)

France Healthy 35-60 Mean: 7.5 y

Monthly questionnaire

Antioxidants:120 mg/d ascorbic

acid + 30 mg/d vitamin E + 6 mg/d β-carotene + 100

µg/d selenium + 20 mg/d zinc

6377/6364


& Industry


House et al., 2010 – DIVINe

(160)Canada

History of Type 1or 2 diabetes & a clinical diagnosis

ofdiabetic

nephropathy

> 18 6 y

Pill counts aevery clinic

visit &annual

measurements of plasma

total homocysteine, serum folate, & serum B12.

B-complex (folic acid 2.5mg/d; B6

25mg/d; B12 1mg/d)

119/119

8/4 MI

Agency &

Industry6/1 Stroke


Janssen et al., 2010 (161) Netherlands Geriatric patients >65 6 months

Compliance calculated as a percentage of

trial medication

taken

Vitamin Dcholecalciferol 400

IU/day36/34 F 1/0 F All-cause

mortality Agency

Salovaara et al., 2010 – OSTRE-

FPS (162)Finland Postmenopausal 65-71 3 y

Tablet count at end of

study

800 IU cholecalciferol &

1000 mg elemental Ca/d


Agency &

Industry

Sanders et al., 2010 – Vital D

(163)Australia Elderly with high

risk of fractured ≥70 5 y

Study staff confirmed by telephone the ingestion of

study medication

within 2

Vitamin D:500 000 IU

cholecalciferol/year

1131/1127 F17/13 Total

CVDAgency


42


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



weeks.

Sanyal et al., 2010 – PIVENS

(164)USA Nonalcoholic

steatohepatitisMean: 46.3

120 weeks

(96 weeks of treatment

+ 24 weeks of follow-

up)

N/A Vitamin E800 IU /d 84/83 1/0 All-cause

mortality

Agency &

Industry

VITATOPS Trial Study Group 2010

(165)

AustraliaNorth-

AmericaAsia

South-AmericaEurope

History of stroke/TIA 63 3.4 years N/A

B-complex (2 mg folic acid/d , 25mg

B6, 0.5mg B12)4089/4075

616/678 Total CVD

Agency &

Industry

118/114 MI360/388 Stroke



Boden et al., 2011 – AIM HIGH (166)

USA and Canada CVD ≥45 3 y Pill counts

Vitamin B3 (niacin): 1500-

2000mg per day1718/1696

171/158 Total CVD

Agency &

Industry

104/93 MI

30/18 Stroke

45/38 CVD mortality

38/34 CHD mortality


Cherniack et al., 2011 (167) USA elderly veterans ≥ 70 6 months serum Vitamin D: 2000

IU/d 23/23

1/1 MIAgency

& industry

1/0 MI mortality


Grimnes et al., 2011 (168) Norway Healthy Mean:

53 6 months

Phone call by study nurses at month 1 and 3 + pill

counts

Vitamin D:20,000 IU vitamin D3 twice weekly


43


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



Lewis et al., 2011 – CAIFOS

(169)Australia Healthy

postmenopausal 75.1 5 y

Medication return and

tablet counting

1200 mg calcium/d 730/730 F

104/103 Total CVD

Agency


18/24 CVD mortality

13/9 MI mortality



Gallagher et al., 2012 – VIDOS

(170)USA

≥7 years postmenopausal with Vitamin D

insufficiency

67 1 y

Food diary, serum

analysis, urine sample every

3months

Vitamin D3 400-4800 IU once daily 20/21 F 1/0 Stroke Agency

Marshall et al., 2011 (171) USA

HGPIN with no evidence of

cancer≥40 3 y Pill counts 200 µg/d selenium 212/211 4/6 All-cause

mortality Agency

Alvarez et al., 2012 (172) USA Early chronic

kidney disease 62 1 y Food records, serum levels.

50,000 IU Vitamin D3/wk for 12 wk

followed by 50,000 IU every other wk

for 40 wk.


Glendenning et al., 2012 (173) Australia Healthy

(postmenopausal)

≥70Mean: 76.7

9 monthsUnder

supervision of study staff

Vitamin D:150,000 IU oral cholecalciferol every 3 months


Agency &

Industry

44


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



Lehouck et al., 2012 (174) Belgium Hospitalized

COPD patients 68 1 y Oral syringe dispensers

100,000 IU vitamin D/4 weeks 91/91 9/6 All-cause

mortality Agency

Ma et al., 2012 – SIT (175) China

Seropositive for Helicobacte

pylori35-64 7.3 y

Tablet counts and serum

measurements

Antioxidants:500mg Vitamin C

+ 200IU Vitamin E + 15mg β-carotene + 75µg selenium

(β-carotene component was

discontinued after 6 months)

1706/1705


Agency


Punthakee et al., 2012 – TIDE

(176)33 countries

Type 2 Diabetes Mellitus and

other CVD risk factors

≥50 5.5 y N/A 1,000 IU vitamin D/day 607/614

2/3 Total CVD

Industry

1/1 MI1/1 Stroke

0/1 CVD mortality


Sesso et al., 2012 – PHS II

(177)USA Healthy 64.3 11.2 y

Annual mailed

questionnaireMultivitamin 7317/7324 M

876/856 Total CVD

Agency &

industry

317/335 MI

332/311 Stroke


27/43 MI mortality


1345/1412 All-cause

45


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



mortality

Delanaye et al., 2013 (178) Belgium Hemodialysed

Patients ≥18 1 y

Supplements were taken

with a nurse ensuring

compliance

Vitamin D:Cholecalciferol

(25000 IU) therapy or placebo every 2

weeks


Hewitt et al., 2013 (179) Australia

Patients on a thrice-weekly

hemodialysis for ≥ 3 months & with screening

levels of 25(OH)D ≤24ng/ml

(60nmol/L)

≥ 18 ≥6 months

Supplements provided by dialysis staff

to ensure compliance

10 ml of an oral solution of medium chain triglyceride containing 50000

IU of cholecalciferol or placebo. Taken

weekly for the first 8 weeks, followed by monthly doses

for 4 months.


Lamas et al., 2013 – TACT

(180)

USA & Canada Post MI 65 3 y

Unused pills returned to

site

Multivitamins (28-component mixture)

853/855

94/115 Total CVD

Agency

58/61 MI8/15 Stroke

45/56 CVD mortality


Manning et al., 2013 (181)

New-Zealand

Metabolic syndrome 27-80 1 y Pill counts Vitamin E

(100 IU/day) 36/40 0/1 All-cause mortality

Agency &

industry

Prentice et al., 2013 –WHI CaD (182) USA Postmenopausal 50-79 Mean 7.2

y Semiannual pill counts

1000 mg of calcium + 400 IU Vitamin D3 (daily, given in two equal doses) or placebo.

Concurrent supplementation was permitted.

7718/7584 F(no personal supplement

users)

848/813 Total CVD

Agency

229/211 Total CHD

193/167 MI

184/162 Stroke


46


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



Witham et al., 2013 – VitDISH

(183)

United Kingdom

Isolated systolic hypertension &

baseline 25OHD level < 30 ng/ml

≥ 70 mean:

772 y

Patients were observed ingesting

supplement

Vitamin D:A total of 100 000

U of oral cholecalciferol or placebo every 3

months for 1 year.

80/79

2/2 Total CHD

Agency3/1 Stroke


Witham et al., 2013 (184) UK

History of myocardial infarction

66 (mean) 6 months n/a

Vitamin D3300,000

units39/36



HPS2-THRIVE Collaborative Group 2014

(185)

United Kingdom,

China, Scandinavia

History of MI, CVD, peripheral arterial disease,

DM, & symptomatic

coronary disease

50-80 Median:3.9 y

Self-reported consumption & blood tests

2 g vitamin B3 (niacin)/d 12838/12835

1696/1758 Total CVD

Agency &

industry

668/694 Total CHD

402/431 MI498/499 Stroke



Wang et al., 2014 – OPERA

(186)Hong Kong

Stage 3-5 CKD with LV

hypertrophy18-75 52 weeks

Recorded during visits at weeks 6,

12, 24, 36, 48, and 52.

Vitamin D: Oral paricalcitrol (1ug)

once daily30/30

0/6 Total CVD

Agency0/2 MI

0/2 Stroke

Van Wijngaarden et al., 2014 – B-PROOF (187)

Netherlands

Living independently with elevated homocysteine (12-50µmol)

≥65 2 yTablet counts,

periodic phone calls

B-complex (500µg vitamin B12 plus 400µg folic acid)

or placebo


Wang et al., 2014 – PHS II

(188)USA Healthy

≥50(mean: 64.3)

Mean: 10.3 y

Self-reported (questionnaire

)

Vitamin E400 IU synthetic α-

tocopherol on alternate days

3659/3653M 397/406 M All-cause mortality

Agency &

industryVitamin C 3673/3653 M 414/406 MAntioxidants

(Vitamin E and C) 3656/3653 M 440/406 M

47


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



Huo et al., 2015 – CSPPT (189) China Hypertension 60 4.5 y Participant

reporting0.8 mg /d of folic

acid 10348/10354

324/405 Total CVD

Agency &

industry

25/24 MI282/355 Stroke

43/43 CVD mortality

5/4 MI mortality



Ponikowski et al., 2015 –

CONFIRM-HF (190)

Austria, Italy,

Poland, Portugal, Russia, Spain,

Sweden, UK, Ukraine

Heart failure 68 52 weeksIntervention delivered by

injection

i.v.iron, as ferric

carboxymaltose150/151

11/12CVD

mortality

Industry

12/14All-cause mortality

Wang et al., 2015 (191) China Healthy 60–74 1 y

Number of capsules

consumedwas recorded

to monitor compliance

400 µg folic acid, 2 mg vitamin B6, 10

µg vitamin B12195/195 4/6

All-cause mortality Agency

Van Dijk et al., 2015 –

B-PROOF (192)Netherlands

Living independently with elevated homocysteine (12-50µmol)

≥65 2 yTablet counts,

periodic phone calls

B-complex (500µg vitamin B12 plus 400µg folic acid)

or placebo

1516/1511181/170

Total CVD

Agency46/60 Stroke

45/43 MI

Gupta et al., 2016 (193) India

Ischaemic stroke survivors with

low serum vitamin D

(<75nmol/L)

60±11.3 6 months

Administered by healthcare

providers

Calcium + Vitamin D 25/28 4/11

All-cause mortality

None

Lappe et al., 2017 (194) USA Healthy

(postmenopausal) ≥556 y

(supplementation

Pill counts and returned

bottles

Calcium + Vitamin D (2000 IU/d

of vitamin D3 and 1156/1147 W 7/9 W All-cause

mortality

Agency &

industry

48


Age (years)


or range)

Supplement intake

assessment



/control)

Incident cases



during 4 y) weighing 1500mg/d of

calcium)CHD, Coronary heart disease; CVD, cardiovascular disease; N/A, not available; T2DM, type two diabetes mellitus; CDPRG, The Coronary Drug Project Research Group; STOCKHOLM, Stockholm lschaemic Heart Disease Secondary Prevention Study; FATS, Familial Atherosclerosis Treatment Study;UCSF-SCOR, University of California, San Francisco, Arteriosclerosis Specialized Center of Research; CHAOS, Cambridge Heart Antioxidant Study; ADCS, Alzheimer’s Disease Cooperative Study; DATATOP, Deprenyl And Tocopherol AntioxidativeTherapy Of Parkinsonism; ATBC , The Alpha- Tocopherol, Beta-Carotene Cancer Prevention Study; CPPS, Calcium Polyp Prevention Study;OSTPRE, Kuopio Osteoporosis Study; SPACE, Secondary Prevention with Antioxidants of Cardiovascular disease in Endstage renal disease; HOPE, Heart Outcomes Prevention Evaluation; AREDS, Age-Related Eye Disease Study; HATS, The HDL-Atherosclerosis Treatment Study; PPP, Primary Prevention Project; VEAPS, The Vitamin E Atherosclerosis Prevention Study; The Swiss Heart Study, The Swiss Heart Study: A Randomized Control Trial; REACT, The Roche European American Cataract Trial; WAVE, Women’s Angiographic Vitamin and Estrogen; FITNESS, The Frailty Interventions Trial iN Elderly SubjectS; NoNOF, The Notthinghan Neck Of Femur Study; VECAT, Vitamin E, Cataract and Age-related Maculopathy Trial;ARBITER2, Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol; RECORD, Randomized Evaluation of Calcium Or vitamin D; VIP, Visual ImPairment Trial; WHI, Women’s Health Study; NORVIT, The NORwegian Viamin Trial; HOPE, Heart Outcomes Prevention Evaluation; ASFAST, Atherosclerosis and Folic Acid Supplementation Trial; MAVET, The Melbourne Atherosclerosis Vitamin E Trial; FACIT, Folic Acid and Carotid Intima-media Thickness; HOST, Homocysteinemia in Kidney and End Stage Renal Disease; ASCENT, AIPC (Androgenindependent Prostate Cancer) Study of Calcitriol Enhancing Taxotere; WAFACS, Women’s Antioxidant and Folic Acid Cardiovascular Study; CAIFOS, Calcium Intake Fracture Outcome Study UK CAP, The United Kingdom Colorectal Adenoma Prevention Trial; FERRIC- HF, Ferric Iron Sucrose in Heart Failure; PHS, Physicians Health Study; FAIR-HF, Ferinject Assessment in Patients with Iron Deficiency and Chronic Heart Failure Ttrial; BVAIT, The B-Vitamin Atherosclerosis Intervention Trial; WACS, The Women’s Antioxidant Cardiovascular Study; NHS/HPFS, Nurse’s Health Study/ Health Professionals Follow-up Study; SELECT, The Selenium and Vitamin E Cancer Prevention Trial; ARBITER, Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol; SEARCH, Study of the Effectiveness of Additional Reductions in Cholesterol and Homocysteine; SU.VI.MAX, The Supplémentation en Vitamines et Minéraux AntioXydants study; VITAL, VITamins And Lifestyle cohort CAIFOS, Calcium Intake Fracture Outcome Study; WAFACS, Women’s Antioxidant and Folic Acid Cardiovascular Study; UK CAP, The United Kingdom Colorectal Adenoma Prevention Trial; SELECT, The Selenium and Vitamin E Cancer Prevention

.

49

Supplementary Table 4. GRADE assessment for vitamin D

Certainty assessment Effect

Certainty№ of

studiesStudy design Risk of bias Inconsistency Indirectness Imprecision Other considerations Relative

(95% CI)Absolute(95% CI)

Vitamin D and total CVD risk (random effects model) - RCTs

6 randomised trials

not serious not serious not serious a serious b none c RR 0.95(0.86 to 1.05)

8 fewer per 1,000

(from 8 more to 23

fewer)

⨁⨁⨁◯MODERATE

Vitamin D and total CHD risk (random effects model) - RCTs

3 randomised trials

not serious not serious not serious d very serious e none c RR 0.97(0.22 to 4.22)

0 fewer per 1,000

(from 11 fewer to 45

more)

⨁⨁◯◯LOW

Vitamin D and MI risk (random effects model) - RCTs

12 randomised trials

not serious not serious not serious f serious g none RR 0.95(0.83 to 1.10)

3 fewer per 1,000

(from 6 more to 10

fewer)


Vitamin D and stroke risk (random effects model) - RCTs


not serious not serious not serious h serious i none RR 1.12(0.94 to 1.34)

5 more per 1,000

(from 2 fewer to 14

more)


Vitamin D and CVD mortality risk - (random effects model) - RCTs

2 randomised trials

not serious not serious not serious j serious k none c RR 0.86(0.66 to 1.10)

8 fewer per 1,000

(from 6 more to 21

fewer)


Vitamin D and CHD mortality risk - (random effects model) - RCTs

2 randomised trials

not serious l not serious serious m very serious n none c RR 0.41(0.06 to 2.72)

16 fewer per 1,000(from 26

fewer to 47 more)

⨁◯◯◯VERY LOW

Vitamin D and stroke mortality risk - (random effects model) - RCTs

2 randomised trials

not serious o not serious not serious p serious q none c RR 1.13(0.68 to 1.87)

3 more per 1,000

(from 6 fewer to 18

more)


Vitamin D and all-cause mortality risk (random effects model) - RCTs


not serious not serious not serious r not serious none RR 0.99(0.95 to 1.03)

2 fewer per 1,000

(from 5 more to 8

fewer)

⨁⨁⨁⨁HIGH

50

CI: Confidence interval; RR: Risk ratio

Explanationsa. No serious indirectness for vitamin D supplementation and total CVD risk (RCTs). Five of six studies were primary prevention and one study was secondary prevention. b. Serious imprecision for vitamin D supplementation and total CVD risk (RCTs), as the 95% CI (RR, 0.86-1.05) overlaps with the minimally important difference for clinical benefit (RR<0.95). c. Publication bias was not assessed since there were <10 studies. d. No serious indirectness for vitamin D supplementation and total CHD risk (RCTs). All 3 studies were primary prevention. e. Very serious imprecision for vitamin D supplementation and total CHD risk (RCTs) as the 95% CI (RR, 0.22-4.22) include both clinically important benefit (RR<0.95) and harm (RR>1.05). f. No serious indirectness for vitamin D supplementation and MI risk (RCTs). All 12 studies were primary prevention. g. Serious imprecision for vitamin D supplementation and MI risk (RCTs) as the 95% CI (RR, 0.83-1.10) include both clinically important benefit (RR<0.95) and harm (RR>1.05). h. No serious indirectness for vitamin D supplementation and stroke risk (RCTs). All 11 included studies were primary prevention. i. Serious imprecision for vitamin D supplementation and stroke risk (RCTs) as the 95% CI (RR, 0.94-1.34) include both clinically important benefit (RR<0.95) and harm (RR>1.05). j. No serious indirectness for vitamin D supplementation and CVD mortality risk (RCTs). Both studies were primary prevention. k. Serious imprecision for vitamin D supplementation and CVD mortality risk (RCTs) as the 95% CI (RR, 0.66-1.10) include both clinically important benefit (RR<0.95) and harm (RR>1.05). l. No serious risk of bias for vitamin D supplementation and total CHD mortality risk (RCTs) despite one study being high risk and the other being unclear for random sequence allocation. m. Serious indirectness for vitamin D supplementation and CHD mortality risk (RCTs) as 61% of the study population were hemodialysis patients. n. Very serious imprecision for vitamin D supplementation and CHD mortality risk (RCTs) as the 95% CI (RR, 0.06-2.72) include both clinically important benefit (RR<0.95) and harm (RR>1.05). o. No serious risk of bias for vitamin D supplementation and stroke mortality risk (RCTs) despite one study being high risk and the other being unclear for random sequence allocation. p. No serious indirectness for vitamin D supplementation and stroke mortality risk (RCTs). Both studies were primary prevention. q. Serious imprecision for vitamin D supplementation and stroke mortality risk (RCTs) as the 95% CI (RR, 0.68-1.87) include both clinically important benefit (RR<0.95) and harm (RR>1.05). r. No serious indirectness for vitamin D supplementation and all-cause mortality risk (RCTs). All 44 studies were primary prevention.

Supplementary Table 5. GRADE assessment for vitamin A


Certainty№ of



Vitamin A and all-cause mortality (random effects model) - RCTs

1 randomised trials

not serious not serious a serious c serious d none b RR 0.99(0.56 to 1.72)

0 fewer per 1,000

(from 9 fewer to 15

more)

⨁⨁◯◯LOW


Explanationsa. Unable to assess inconsistency as only one study was included. b. Publication bias was not assessed as there were < 10 studies. c. Serious indirectness for vitamin A supplementation and all-cause mortality risk (RCTs), as the only included study was conducted among a population with actinic keratoses and squamous cell carcinoma or basal cell carcinoma skin cancers. d. Serious imprecision for vitamin A and all-cause mortality risk (RCTs), as the 95% CI (RR, 0.56-1.72) include both clinically important benefit (RR<0.95) and harm (RR>1.05).

Supplementary Table 6. GRADE assessment for beta-carotene

Certainty assessment Effect Certainty

№ of studies

Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations Relative


Beta-carotene and total CVD risk (random effects model) - RCTs

3 randomised trials


1 more per 1,000

(from 2 fewer to 4

more)


Beta-carotene and total CHD risk (random effects model) - RCTs

51


№ of studies



2 randomised trials

not serious not serious serious e serious f none c RR 1.02(0.94 to 1.10)

2 more per 1,000

(from 6 fewer to 9

more)

⨁⨁◯◯LOW

Beta-carotene and MI risk (random effects model) - RCTs

3 randomised trials

not serious not serious not serious g serious h none c RR 0.99(0.91 to 1.09)

0 fewer per 1,000

(from 4 fewer to 4

more)


Beta-carotene and stroke risk (random effects model) - RCTs

3 randomised trials

not serious not serious not serious i serious j none c RR 1.06(0.95 to 1.19)

2 more per 1,000

(from 2 fewer to 7

more)


Beta-carotene and CVD mortality risk (random effects model) - RCTs

4 randomised trials

not serious not serious not serious k serious l none c RR 1.10(0.98 to 1.23)

3 more per 1,000

(from 1 fewer to 8

more)


Beta-carotene and CHD mortality risk (random effects model) - RCTs

1 randomised trials

not serious not serious d serious m serious n none c RR 0.99(0.83 to 1.18)

0 fewer per 1,000

(from 6 fewer to 6

more)

⨁⨁◯◯LOW

Beta-carotene and MI mortality risk (random effects model) - RCTs

1 randomised trials

not serious not serious d serious o not serious none c RR 0.42(0.20 to 0.87)

3 fewer per 1,000

(from 1 fewer to 5

fewer)


Beta-carotene and stroke mortality risk (random effects model) - RCTs

2 randomised trials

not serious not serious serious p serious q none c RR 1.33(0.73 to 2.44)

2 more per 1,000

(from 1 fewer to 7

more)

⨁⨁◯◯LOW

Beta-carotene and all-cause mortality risk (random effects model) - RCTs

6 randomised trials

not serious not serious not serious r serious s none c RR 1.03(0.96 to 1.11)

3 more per 1,000

(from 4 fewer to 11

more)



Explanationsa. No serious indirectness for beta-carotene supplementation and total CVD risk (RCTs). Majority of studies were primary prevention with the exception of one study (Cook et al. 2007) which included individuals with a history of CVD or at least three cardiac risk factors. b. Serious imprecision for beta-carotene supplementation and total CVD risk (RCTs), as the 95% CI (RR, 0.96-1.10) overlaps with the minimally important difference for clinical harm (RR>1.05). c. Publication bias was not assessed since there were <10 studies.

52

d. Not able to assess consistency as only one study was included. e. Serious indirectness for beta-carotene supplementation and CHD risk (RCTs), as the 2 included studies were conducted either in middle-aged, Finnish male smokers or middle-aged females with a history of CVD, or at least three cardiac risk factors. f. Serious imprecision for beta-carotene supplementation and total CHD risk (RCTs), as the 95% CI (RR, 0.94, 1.10) include both clinically important benefit (RR<0.95) and harm (RR>1.05). g. No serious indirectness for beta-carotene supplementation and MI risk (RCTs). Majority of studies were primary prevention with the exception of one study (Cook et al. 2007) which included individuals with history of CVD or at least three cardiac risk factors. h. Serious imprecision for beta-carotene supplementation and MI risk (RCTs), as the 95% CI (RR, 0.91, 1.09) include both clinically important benefit (RR<0.95) and harm (RR>1.05). i. No serious indirectness for beta-carotene supplementation and stroke risk (RCTs). Majority of studies were primary prevention with the exception of one study (Cook et al., 2007) which included individuals with history of CVD or at least three cardiac risk factors. j. Serious imprecision for beta-carotene supplementation and stroke risk (RCTs), as the 95% CI (RR, 0.95-1.19) overlaps with the minimally important difference for clinical harm (RR>1.05). k. No serious indirectness for beta-carotene and CVD mortality risk (RCTs). Majority of the studies were primary prevention, with the exception of one study (Cook et al., 2007) which included individuals with history of CVD, or at least three cardiac risk factors. l. Serious imprecision for beta-carotene and CVD mortality risk (RCTs), as the 95% CI (RR, 0.98-1.23) overlaps with the minimally important difference for clinical harm (RR>1.05). m. Serious indirectness for beta-carotene supplementation and CHD mortality risk (RCTs), as the 1 included study was conducted in middle-aged, Finnish male smokers. n. Serious imprecision for beta-carotene supplementation and CHD mortality risk (RCTs), as the 95% CI (RR, 0.83, 1.18) include both clinically important benefit (RR<0.95) and harm (RR>1.05). o. Serious indirectness for beta-carotene supplementation and MI mortality risk (RCTs), as the one included study was conducted in a female population at high risk for CVD. This was a secondary prevention study. p. Serious indirectness for beta-carotene supplementation and stroke mortality risk (RCTs), as the 2 included studies were conducted in middle-aged, Finnish male smokers or middle-aged females with a history of CVD or at least 3 cardiac risk factors. q. Serious imprecision for beta-carotene supplementation and stroke mortality risk (RCTs), as the 95% CI (RR 0.73, 2.44), include both clinically important benefit (RR<0.95) and harm (RR>1.05). r. No serious indirectness for beta-carotene supplementation and all-cause mortality risk (RCTs). Majority of the studies were primary prevention, with the exception of one study (Cook et al., 2007) which included individuals with of CVD, or at least three cardiac risk factors. s. Serious imprecision for beta-carotene supplementation and all-cause mortality risk (RCTs), as the 95% CI (RR, 0.96-1.11) overlaps with the minimally important difference for clinical harm (RR>1.05).

Supplementary Table 7. GRADE assessment for antioxidants


№ of studies



Antioxiants and total CVD (random effects model) - RCTs

7 randomised trials

not serious not serious not serious a not serious none b RR 0.99(0.95 to 1.04)

1 fewer per 1,000

(from 5 more to 6

fewer)

⨁⨁⨁⨁HIGH

Antioxidants and total CHD (random effects model) - RCTs

1 randomised trials

not serious not serious c serious d serious e none b RR 0.97(0.86 to 1.09)

2 fewer per 1,000

(from 7 more to 11

fewer)

⨁⨁◯◯LOW

Antioxidants and MI (random effects model) - RCTs

6 randomised trials

not serious not serious not serious f serious g none b RR 0.98(0.90 to 1.08)

1 fewer per 1,000

(from 3 more to 4

fewer)


Antioxidants and stroke (random effects model) - RCTs

7 randomised trials

not serious not serious not serious h serious i none b RR 1.00(0.92 to 1.09)

0 fewer per 1,000

(from 3 fewer to 3

more)


Antioxidants and CVD mortality (random effects model) - RCTs

7 randomised trials

not serious not serious not serious j serious k none b RR 1.02(0.94 to 1.10)

1 more per 1,000

(from 3 fewer to 5

more)


Antioxidants and CHD mortality (random effects model) - RCTs

53


№ of studies



2 randomised trials

not serious not serious serious l serious m none b RR 1.02(0.93 to 1.13)

1 more per 1,000

(from 4 fewer to 7

more)

⨁⨁◯◯LOW

Antioxidants and MI mortality (random effects model) - RCTs

3 randomised trials

not serious not serious serious n very serious o none b RR 1.51(0.39 to 5.93)

3 more per 1,000

(from 4 fewer to 31

more)


Antioxidants and stroke mortality (random effects model) - RCTs

5 randomised trials

not serious not serious not serious p serious q none b RR 1.10(0.87 to 1.39)

1 more per 1,000

(from 1 fewer to 2

more)


Antioxidants and all-cause mortality (random effects model) - RCTs


not serious not serious not serious r serious s none RR 1.06(1.00 to 1.12)

5 more per 1,000

(from 0 fewer to 9

more)



Explanationsa. No serious indirectness for antioxidant supplementation and total CVD risk (RCTs). Two studies were secondary prevention, four studies were primary prevention and one study was a combination of both. b. Publication bias was not assessed since there were < 10 studies. c. Unable to assess inconsistency as only one study was included. d. Serious indirectness for antioxidants supplementation and total CHD risk (RCTs) as there was only one study included (Tornwall et al., 2004) and it was conducted among Finnish male smokers. e. Serious imprecision for antioxidant supplementation and total CHD risk (RCTs), as the 95% CI (RR, 0.86-1.09) include both clinically important benefit (RR<0.95) and harm (RR>1.05). f. No serious indirectness for antioxidant supplementation and MI risk (RCTs). Three studies were for secondary prevention, two studies were primary prevention and one study was a combination of both. g. Serious imprecision for antioxidant supplementation and MI risk (RCTs), as the 95% CI (RR, 0.90-1.08) include both clinically important benefit (RR<0.95) and harm (RR>1.05). h. No serious indirectness for antioxidant supplementation and stroke risk (RCTs). Three studies were secondary prevention, three studies were primary prevention and one study was a combination of both. i. Serious imprecision for antioxidant supplementation and stroke risk (RCTs), as the 95% CI (RR, 0.92-1.09) include both clinically important benefit (RR<0.95) and harm (RR>1.05). j. No serious indirectness for antioxidant supplementation and CVD mortality risk (RCTs). Three studies were for secondary prevention, three studies were primary prevention and one study was a combination of both. k. Serious imprecision for antioxidants supplementation and CVD mortality risk (RCTs), as the 95% CI (RR, 0.94-1.10) include both clinically important benefit( (RR<0.95) harm (RR>1.05). l. Serious indirectness for antioxidants supplementation and CHD mortality risk (RCTs) as there was only two studies included and one of them (Tornwall et al., 2004) was conducted among Finnish male smokers only, which accounted for ~30% of the population. m. Serious imprecision for antioxidant supplementation and CHD mortality risk (RCTs), as the 95% CI (RR, 0.93-1.13) include both clinically important benefit (RR<0.95) and harm (RR>1.05). n. Serious indirectness for antioxidant supplementation MI mortality risk (RCTs), as the included studies were in specific populations: early age-related cataract (Chylack et al., 2002), smokers >10 cigarettes/day (Mooney et al., 2005) and peripheral artery disease (CLIPS group 2007). o. Very serious imprecision for antioxidant supplementation and MI mortality risk (RCTs), as the 95% CI (RR, 0.39-5.93) include both appreciable clinically important benefit (RR<0.95) and harm (RR>1.05). p. No serious indirectness for antioxidant supplementation and stroke mortality risk (RCTs). One study was secondary prevention, 3 studies were primary prevention and 1 study was a combination of both. q. Serious imprecision for antioxidant supplementation and stroke mortality risk (RCTs), as the 95% CI (RR, 0.87-1.39) include both clinically important benefit (RR<0.95) and harm (RR>1.05). r. No serious indirectness for antioxidant supplementation and all-cause mortality (RCTs). Sixteen studies were primary prevention, three were secondary prevention and two were a combination of both. s. Serious imprecision for antioxidants supplementation and all-cause mortality risk (RCTs)-random effects model, as the 95% CI (RR, 1.00-1.12) overlaps with the minimally important difference for clinical harm (RR>1.05).

Supplementary Table 8. GRADE assessment for vitamin E

54


Certainty№ of



Vitamin E and total CVD risk (random effects model) - RCTs


not serious not serious a not serious b serious c publication bias strongly suspected d

RR 0.96(0.89 to 1.03)

3 fewer per 1,000

(from 2 more to 9

fewer)

⨁⨁◯◯LOW

Vitamin E and total CHD risk (random effects model) - RCTs

2 randomised trials

not serious not serious serious e serious f none g RR 0.97(0.90 to 1.06)

3 fewer per 1,000

(from 6 more to 10

fewer)

⨁⨁◯◯LOW

Vitamin E and MI risk (random effects model) - RCTs


not serious serious h not serious i serious j publication bias strongly suspected d

RR 0.87(0.75 to 1.01)

5 fewer per 1,000

(from 0 fewer to 9

fewer)


Vitamin E and stroke risk (random effects model) - RCTs


not serious not serious not serious k serious l none g RR 0.98(0.88 to 1.08)

0 fewer per 1,000

(from 2 fewer to 2

more)

⨁⨁◯◯MODERATE

Vitamin E and CVD mortality risk (random effects model) - RCTs


not serious not serious not serious m serious n none g RR 0.95(0.89 to 1.02)

1 fewer per 1,000

(from 1 more to 3

fewer)


Vitamin E and CHD Mortality risk (random effects model) - RCTs

2 randomised trials

not serious not serious serious o serious p none g RR 0.90(0.80 to 1.02)

4 fewer per 1,000

(from 1 more to 8

fewer)

⨁⨁◯◯LOW

Vitamin E and MI mortality risk (random effects model) - RCTs

7 randomised trials

not serious not serious not serious q serious r none g RR 0.91(0.78 to 1.07)

1 fewer per 1,000

(from 1 more to 2

fewer)


Vitamin E and stroke mortality risk (random effects model) - RCTs

5 randomised trials

not serious not serious not serious s serious t none g RR 0.97(0.66 to 1.43)

0 fewer per 1,000

(from 0 fewer to 1

more)


Vitamin E and all-cause mortality risk (random effects model) - RCTs


not serious not serious not serious u not serious none RR 1.00(0.97 to 1.04)

0 fewer per 1,000

(from 2 fewer to 3

more)

⨁⨁⨁⨁HIGH

55


Explanations

a. Although there was evidence of serious inconsistency for vitamin E and total CVD risk (RCTs) (I2=50%, p=0.04), independent removal of one trial (Boaz et al.) explained half of the heterogeneity (I2=24%, p=0.24). b. No serious indirectness for vitamin E and total CVD risk (RCTs). Two studies were secondary prevention, five studies were primary prevention and three were a combination of both. c. Serious imprecision for vitamin E supplementation and total CVD risk (RCTs), as the 95% CI (RR 0.89, 1.03) overlaps with the minimally important difference for clinical benefit (RR<0.95). d. Publication bias strongly detected due to funnel plot asymmetry and Egger’s and Begg’s tests (p<0.05) e. Serious indirectness for vitamin E supplementation and total CHD risk (RCTs), as the two included studies were conducted either in Finnish, middle-aged male smokers or middle-aged women with history of CVD, or at least three cardiac risk factors. f. Serious imprecision for vitamin E supplementation and total CHD risk (RCTs), as the 95% CI (RR, 0.90, 1.06) include both clinically important benefit (RR<0.95) and harm (RR>1.05). g. Publication bias was not assessed since there were <10 studies. h. Serious inconsistency for vitamin E supplementation and MI risk (RCTs), as I2=59%, p=0.01. i. No serious indirectness for vitamin E and total MI risk (RCTs). Two studies were secondary prevention, five studies were primary prevention and three studies were a combination of both. j. Serious imprecision for vitamin E supplementation and MI risk (RCTs), as the 95% CI (RR, 0.75-1.01) overlaps with the minimally important difference for clinical benefit (RR<0.95) k. No serious indirectness for vitamin E and stroke risk (RCTs). Three studies were secondary prevention, five studies were primary prevention and three studies were a combination of both. l. Serious imprecision for vitamin E supplementation and stroke risk (RCTs), as the 95% CI (RR 0.88, 1.08) include both clinically important benefit (RR<0.95) and harm (RR>1.05). m. No serious indirectness for vitamin E supplementation and CVD mortality risk (RCTs). Three studies were secondary prevention, five studies were primary prevention and three studies were a combination of both. n. Serious imprecision for vitamin E supplementation and CVD mortality risk (RCTs), as the 95% CI (RR, 0.89-1.02) overlaps with the minimally important difference for clinical benefit (RR<0.95) o. Serious indirectness for vitamin E supplementation and CHD mortality risk (RCTs), as the two included studies were conducted in Italian, middle-aged individuals with a recent MI or Finnish, middle-aged male smokers. p. Serious imprecision for vitamin E supplementation and CHD mortality risk (RCTs), as the 95% CI (RR, 0.80-1.02) overlaps with the minimally important difference for clinical benefit (RR<0.95) q. No serious indirectness for vitamin E and MI mortality risk (RCTs). Three studies were secondary prevention, three studies were primary prevention and two studies were a combination of both. r. Serious imprecision for vitamin E supplementation and MI mortality risk (RCTs), as the 95% CI (RR, 0.78-1.07) include both clinically important benefit (RR <0.95) and harm (RR>1.05). s. No serious indirectness for vitamin E supplementation and stroke mortality risk (RCTs). One study was secondary prevention, three studies were primary prevention and two studies were a combination of both. t. Serious imprecision for vitamin E supplementation and stroke mortality risk (RCTs), as the 95% CI (RR, 0.66-1.43) include both clinically important benefit (RR <0.95) and harm (RR>1.05). u. No serious indirectness for vitamin E supplementation and all-cause mortality risk (RCTs). Nine studies were secondary prevention and 23 studies were primary prevention.

Supplementary Table 9. GRADE assessment for vitamin C


Certainty№ of



Vitamin C and total CVD risk (random effects model) - RCTs

2 randomised trials


1 fewer per 1,000

(from 9 fewer to 9

more)


Vitamin C and total CHD risk (random effects model) - RCTs

1 randomised trials

not serious not serious d serious e serious f none c RR 1.04(0.93 to 1.17)

5 more per 1,000

(from 8 fewer to 20

more)

⨁⨁◯◯LOW

Vitamin C and MI risk (random effects model) - RCTs

2 randomised trials

not serious not serious not serious g serious h none c RR 0.96(0.81 to 1.14)

1 fewer per 1,000

(from 5 more to 7

fewer)


Vitamin C and stroke risk (random effects model) - RCTs

56


Certainty№ of



2 randomised trials


3 fewer per 1,000

(from 3 more to 8

fewer)


Vitamin C and CVD mortality risk (random effects model) - RCTs

2 randomised trials

not serious not serious not serious k serious l none c RR 1.07(0.92 to 1.25)

3 more per 1,000

(from 3 fewer to 10

more)


Vitamin C and MI mortality risk (random effects model) - RCTs

1 randomised trials

not serious not serious not serious m serious n none c RR 0.79(0.40 to 1.55)

1 fewer per 1,000

(from 3 fewer to 3

more)


VItamin C and stroke mortality risk (random effects model) - RCTs

1 randomised trials

not serious not serious not serious o serious p none c RR 0.83(0.42 to 1.65)

1 fewer per 1,000

(from 3 fewer to 3

more)


Vitamin C and all-cause mortality risk (random effects model) - RCTs

4 randomised trials

not serious not serious not serious q serious r none c RR 1.02(0.94 to 1.11)

2 more per 1,000

(from 7 fewer to 12

more)



Explanationsa. No serious indirectness for vitamin C supplementation and total CVD risk (RCTs). Both studies were a combination of primary and secondary prevention and had large populations, one of men and the other of women. b. Serious imprecision for vitamin C supplementation and total CVD risk (RCTs), as the 95% CI (RR, 0.90-1.10) include both clinically important benefit (RR<0.95) and harm (RR>1.05). c. Publication bias was not assessed as there were < 10 studies. d. Unable to assess inconsistency as only one study was included. e. Serious indirectness for vitamin C supplementation and total CHD risk (RCTs), as the one included study was conducted in women only. f. Serious imprecision for vitamin C supplementation and total CHD risk (RCTs), as the 95% CI (RR, 0.93-1.17) include both clinically important benefit (RR<0.95) and harm (RR>1.05). g. No serious indirectness for vitamin C supplementation and MI risk (RCTs). Both studies were a combination of primary and secondary prevention and had large populations, one of men and the other of women. h. Serious imprecision for vitamin C supplementation and MI risk (RCTs), as the 95% CI (RR, 0.81-1.14) include both clinically important benefit (RR<0.95) and harm (RR>1.05). i. No serious indirectness for vitamin C supplementation and total CVD risk (RCTs). Both studies were a combination of primary and secondary prevention and had large populations, one of men and the other of women. j. Serious imprecision for vitamin C supplementation and stroke risk (RCTs), as the 95% CI (RR, 0.78-1.09) include both clinically important benefit (RR<0.95) and harm (RR>1.05). k. No serious indirectness for vitamin C supplementation and CVD mortality risk (RCTs). Both studies were a combination of primary and secondary prevention and had large populations, one of men and the other of women. l. Serious imprecision for vitamin C supplementation and CVD mortality risk (RCTs), as the 95% CI (RR, 0.92-1.25) include both clinically important benefit (RR<0.95) and harm (RR>1.05). m. No serious indirectness for vitamin C supplementation and MI mortality risk (RCTs). Both studies were a combination of primary and secondary prevention and had large populations, one of men and the other of women. n. Serious imprecision for vitamin C supplementation and MI mortality risk (RCTs), as the 95% CI (RR, 0.40-1.55) includes both clinically important benefit (RR<0.95) and harm (RR>1.05). o. No serious indirectness for vitamin C supplementation and stroke mortality risk (RCTs). Both studies were a combination of primary and secondary prevention and had large populations, one of men and the other of women. p. Serious imprecision for vitamin C supplementation and stroke mortality risk (RCTs), as the 95% CI (RR, 0.42-1.65) include both clinically important benefit (RR<0.95) and harm (RR>1.05). q. No serious indirectness for vitamin C supplementation and all-cause mortality risk (RCTs). Three studies were primary prevention and one was a combination of both primary and secondary prevention. r. Serious imprecision for vitamin C supplementation and all-cause mortality risk (RCTs), as the 95% CI (RR, 0.94-1.11) include both clinically important benefit (RR<0.95) and harm (RR>1.05).

57

Supplementary Table 10. GRADE assessment for selenium


Certainty№ of



Selenium and total CVD risk (random effects model) - RCTs

2 randomised trials

not serious not serious serious a serious b none c RR 1.04(0.96 to 1.12)

5 more per 1,000

(from 5 fewer to 15

more)

⨁⨁◯◯LOW

Selenium and total CHD risk (random effects model) - RCTs

1 randomised trials

not serious not serious d serious e serious f none c RR 1.06(0.76 to 1.48)

7 more per 1,000


more)

⨁⨁◯◯LOW

Selenium and MI risk (random effects model) - RCTs

2 randomised trials

not serious not serious serious g serious h none c RR 0.93(0.62 to 1.39)

6 fewer per 1,000


more)

⨁⨁◯◯LOW

Selenium and stroke risk (random effects model) - RCTs

2 randomised trials

not serious not serious serious i serious j none c RR 0.89(0.70 to 1.14)

2 fewer per 1,000

(from 2 more to 4

fewer)

⨁⨁◯◯LOW

Selenium and CVD mortality risk (random effects model) - RCTs

3 randomised trials

not serious not serious serious k serious l none c RR 1.00(0.66 to 1.51)

0 fewer per 1,000

(from 6 fewer to 10

more)

⨁⨁◯◯LOW

Selenium and MI mortality risk (random effects model) - RCTs

2 randomised trials

not serious not serious serious m very serious n none c RR 1.21(0.49 to 2.99)

3 more per 1,000

(from 7 fewer to 27

more)


Selenium and stroke mortality risk (random effects model) - RCTs

1 randomised trials

not serious not serious d serious o very serious p none c RR 1.13(0.44 to 2.93)

0 fewer per 1,000

(from 1 fewer to 2

more)


Selenium and all-cause mortality risk (random effects model) - RCTs

4 randomised trials

not serious not serious serious q serious r none c RR 0.99(0.88 to 1.12)

1 fewer per 1,000

(from 6 fewer to 6

more)

⨁⨁◯◯LOW


58

Explanationsa. Serious indirectness for selenium supplementation and total CVD risk (RCTs) since both included studies were conducted in specific populations, that is, Lippman et al., 2009 included males only (at higher CVD risk) and Stranges et al., 2006 was conducted in patients with history of non-melanoma skin cancers. b. Serious imprecision for selenium supplementation and total CVD risk (RCTs), as the 95% CI (RR, 0.96-1.12) overlaps with the minimally important difference for clinical harm (RR>1.05). c. Publication bias was not assessed as there were < 10 studies. d. Unable to assess inconsistency as only one study was conducted. e. Serious indirectness for selenium supplementation and total CHD risk (RCTs) as there was only one study which included individuals with a history of non-melanoma skin cancer. f. Serious imprecision for selenium supplementation and total CHD risk (RCTs) as the 95% CI (RR, 0.76-1.48) include both clinically important benefit (RR<0.95) and harm (RR>1.05). g. Serious indirectness for selenium supplementation and MI risk (RCTs) as one of the included studies, Stranges et al., 2006 - NPC, which represented ~ 96% of the population, was conducted in patients with history of non-melanoma skin cancers. h. Serious imprecision for selenium supplementation and MI risk (RCTs) as the 95% CI (RR, 0.62-1.39) include both clinically important benefit (RR<0.95) and harm (RR>1.05). i. Serious indirectness for selenium supplementation and stroke risk (RCTs), since both included studies were conducted in specific populations, that is, Lippman et al., 2009 included males only (at higher CVD risk) and Stranges et al., 2006 was conducted in patients with history of non-melanoma skin cancers. j. Serious imprecision for selenium supplementation and stroke risk (RCTs) as the 95% CI (RR, 0.70-1.14) include both clinically important benefit (RR<0.95) and harm (RR>1.05). k. Serious indirectness for selenium supplementation and total CVD mortality risk (RCTs) since one of the included studies, Lippman et al., 2009-SELECT, which represented ~94% of the population was conducted in males. l. Serious imprecision for selenium supplementation and CVD mortality risk (RCTs) as the 95% CI (RR, 0.66-1.51) include both clinically important benefit (RR<0.95) and harm (RR>1.05). m. Serious indirectness for selenium supplementation and MI mortality risk (RCTs) as one of the included studies, Lippman et al., 2009-SELECT, which represented ~95% of the population was conducted in males. n. Very serious imprecision for selenium supplementation and MI mortality risk (RCTs) as the 95% CI (RR, 0.49-2.99) include both appreciable clinically important benefit (RR<0.95) and harm (RR>1.05). o. Serious indirectness for selenium supplementation and stroke mortality risk (RCTs) as the one included study was conducted in males with higher CVD risks. p. Very serious imprecision for selenium supplementation and stroke mortality risk (RCTs) as the 95% CI (RR, 0.44-2.93) include both appreciable clinically important benefit (RR<0.95) and harm (RR>1.05). q. Serious indirectness for selenium supplementation and all-cause mortality risk (RCTs), as one of the included studies was conducted in males and accounted for 94% of the population. r. Serious imprecision for selenium supplementation and all-cause mortality risk (RCTs) as the 95% CI (RR, 0.88-1.12) include both clinically important benefit (RR<0.95) and harm (RR>1.05).

Supplementary Table 11. GRADE assessment for vitamin B-complex


№ of studies



B-Complex and total CVD risk (random effects model) - RCTs

9 randomised trials


3 fewer per 1,000

(from 7 more to 12

fewer)


B-complex and total CHD risk (random effects model) - RCTs

5 randomised trials

not serious not serious not serious d serious e none c RR 1.04(0.96 to 1.14)

6 more per 1,000

(from 6 fewer to 21

more)


B-Complex and MI risk (random effects model) - RCTs


not serious not serious not serious f serious g none RR 1.00(0.93 to 1.07)

0 fewer per 1,000

(from 5 fewer to 5

more)


B-complex and stroke risk (random effects model) - RCTs


not serious not serious not serious h serious i none RR 0.90(0.81 to 1.00)

5 fewer per 1,000

(from 0 fewer to 9

fewer)


B-Complex and CVD mortality risk (random effects model) - RCTs

5 randomised trials

not serious not serious not serious j serious k none c RR 0.98(0.87 to 1.11)

2 fewer per 1,000

(from 9 more to 10

fewer)


59


№ of studies



B-complex and CHD mortality risk (random effects model) - RCTs

3 randomised trials

not serious not serious not serious l serious m none c RR 1.09(0.97 to 1.23)

6 more per 1,000

(from 2 fewer to 16

more)


B-complex and MI mortality risk (random effects model) - RCTs

2 randomised trials

not serious not serious not serious n serious o none c RR 1.11(0.93 to 1.32)

4 more per 1,000

(from 2 fewer to 11

more)


B-complex and stroke mortality risk (random effects model) - RCTs

2 randomised trials

not serious not serious not serious p serious q none c RR 0.91(0.68 to 1.21)

1 fewer per 1,000

(from 2 more to 3

fewer)


B-Complex and all-cause mortality risk (random effects model) - RCTs


not serious not serious not serious r serious s none RR 1.02(0.97 to 1.06)

3 more per 1,000

(from 4 fewer to 8

more)



Explanationsa. No serious indirectness for vitamin B-complex supplementation and total CVD risk (RCTs). Three studies were primary prevention and six studies were secondary prevention. b. Serious imprecision for vitamin B-complex supplementation and total CVD risk (RCTs) as the 95% CI (RR, 0.93-1.04) overlaps with the minimally important difference for clinical benefit (RR<0.95). c. Publication bias was not assessed since there were < 10 studies. d. No serious indirectness for vitamin B-complex supplementation and total CHD risk (RCTs). All studies were secondary prevention. e. Serious imprecision for vitamin B-complex supplementation and total CHD risk (RCTs) as the 95% CI (RR, 0.96-1.14) overlaps with the minimally important difference for clinical harm (RR>1.05). f. No serious indirectness for vitamin B-complex supplementation and MI risk (RCTs). Four studies were primary prevention, seven studies were secondary prevention and two studies were a combination of both. g. Serious imprecision for vitamin B-complex supplementation and MI risk (RCTs) as the 95% CI (RR, 0.93-1.07) include both clinically important benefit (RR <0.95) and harm (RR>1.05). h. No serious indirectness in vitamin B-complex supplementation and stroke risk (RCTs). Four studies were primary prevention and eight studies were secondary prevention. i. Serious imprecision for vitamin B-complex supplementation and stroke risk (RCTs) as the 95% CI (RR, 0.81-1.00) overlaps with the minimally important difference for clinical benefit (RR<0.95). j. No serious indirectness for vitamin B-complex supplementation and total CVD mortality risk (RCTs). Three studies were secondary prevention and two studies were a combination of primary and secondary prevention. k. Serious imprecision for vitamin B-complex supplementation and CVD mortality risk (RCTs) as the 95% CI (RR, 0.87, 1.11) include both clinically important benefit (RR<0.95) and harm (RR>1.05). l. No serious indirectness for vitamin B-complex supplementation and CHD mortality risk (RCTs). One study was primary prevention and two studies were secondary prevention. m. Serious imprecision for vitamin B-complex supplementation and CHD mortality risk (RCTs) as the 95% CI (RR, 0.97-1.23) overlaps with the minimally important difference for clinical harm (RR>1.05). n. No serious indirectness for vitamin B-complex supplementation and MI mortality risk (RCTs). Both studies were secondary prevention. o. Serious imprecision for vitamin B-complex supplementation and MI mortality risk (RCTs) as the 95% CI (RR, 0.93-1.32) include both clinically important benefit (RR<0.95) and harm (RR>1.05). p. No serious indirectness for vitamin B-complex supplementation and stroke mortality risk (RCTs), although both studies were secondary prevention. q. Serious imprecision for vitamin B-complex supplementation and stroke mortality risk (RCTs) as the 95% CI (RR, 0.68-1.21) include both clinically important benefit (RR<0.95) and harm (RR>1.05). r. No serious indirectness for vitamin B-complex supplementation and all-cause mortality risk (RCTs). Seven studies were primary prevention, seven studies were secondary prevention and two studies were a combination of both. s. Serious imprecision for vitamin B-complex supplementation and all-cause mortality risk (RCTs) as the 95% CI (RR, 0.97-1.06) overlaps with the minimally important difference for clinical harm (RR>1.05).

Supplementary Table 12. GRADE assessment for folic acid

60


Certainty№ of



Folic acid and total CVD risk (random effects model) - RCTs

5 randomised trials

not serious not serious serious a not serious none b RR 0.83(0.73 to 0.93)

8 fewer per 1,000

(from 3 fewer to 13

fewer)


Folic acid and total CHD risk (random effects model) - RCTs

2 randomised trials

not serious not serious not serious c serious d none b RR 1.47(0.95 to 2.28)

13 more per 1,000

(from 1 fewer to 36

more)


Folic acid and MI risk (random effects model) - RCTs

6 randomised trials

not serious not serious serious e serious f none b RR 1.21(0.78 to 1.88)

1 more per 1,000

(from 1 fewer to 3

more)

⨁⨁◯◯LOW

Folic acid and stroke risk (random effects model) - RCTs

7 randomised trials

not serious not serious serious g not serious none b RR 0.80(0.69 to 0.93)

6 fewer per 1,000

(from 2 fewer to 10

fewer)


Folic acid and CVD mortality risk (random effects model) - RCTs

5 randomised trials

not serious not serious serious h serious i none b RR 0.89(0.68 to 1.17)

1 fewer per 1,000

(from 2 more to 3

fewer)

⨁⨁◯◯LOW

Folic acid and MI mortality risk (random effects model) - RCTs

2 randomised trials

not serious not serious serious j very serious k none b RR 1.17(0.39 to 3.49)

0 fewer per 1,000

(from 0 fewer to 1

more)


Folic acid and stroke mortality risk (random effects model) - RCTs

2 randomised trials

not serious not serious serious l serious m none b RR 1.85(0.88 to 3.93)

1 more per 1,000

(from 0 fewer to 3

more)

⨁⨁◯◯LOW

Folic acid and all-cause mortality risk (random effects model) - RCTs


not serious not serious serious n serious o none RR 0.87(0.72 to 1.05)

5 fewer per 1,000

(from 2 more to 10

fewer)

⨁⨁◯◯LOW


Explanationsa. Serious indirectness for folic acid supplementation and CVD risk (RCTs), as the benefit appears to be driven by one study (Huo et al. 2015-CSPPT) conducted in a Chinese population-removal of this one study changes the pooled effect estimate from a beneficial lowering to no lowering in CVD risk (RR=0.89; 95%CI: 0.72, 1.10, p=0.28). One study was secondary prevention and four studies were primary prevention.

61

b. Publication bias was not assessed since there were <10 studies. c. No serious indirectness for folic acid supplementation and total CHD risk (RCTs). One study was primary prevention, while the other one was secondary prevention. d. Serious imprecision for folic acid supplementation and total CHD risk (RCTs), as the 95% CI (RR, 0.95-2.28) overlaps with the minimally important difference for clinical harm (RR>1.05). e. Serious indirectness for folic acid supplementation and MI risk (RCTs). The majority of the population came from one study (Huo et al., 2015-CSPPT) that was conducted in a chinese population only. Two studies were secondary prevention and four studies were primary prevention. f. Serious imprecision for folic acid supplementation and MI risk (RCTs) as the 95% CI (RR, 0.78-1.88) include both clinically important benefit (RR<0.95) and harm (RR>1.05). g. Serious indirectness for folic acid supplementation and stroke risk (RCTs), as the benefit appears to be driven by one study (Huo et al. 2015-CSPPT) conducted in a Chinese population -removal of this one study changes the pooled effect estimate from a beneficial lowering to no lowering in stroke risk (RR=0.92; 95%CI: 0.52, 1.63, p=0.78). Five studies were primary prevention and two were secondary prevention. h. Serious indirectness for folic acid supplementation and CVD mortality risk (RCTs). The majority of the population came from one study (Huo et al., 2015-CSPPT) that was conducted in a chinese population only. One study was secondary prevention and four studies were primary prevention. i. Serious imprecision for folic acid supplementation and CVD mortality risk (RCTs) as the 95% CI (RR, 0.68-1.17) include both clinically important benefit (RR<0.95) and harm (RR>1.05). j. Serious indirectness for folic acid supplementation and MI mortality risk (RCTs). The majority of the population came from one study (Huo et al., 2015-CSPPT) that was conducted in a chinese population only. One study was secondary prevention and one study was primary prevention. k. Very serious imprecision for folic acid supplementation and MI mortality risk (RCTs) as the 95% CI (RR, 0.39-3.49) include both appreciable clinically important benefit (RR<0.95) and harm (RR>1.05). l. Serious indirectness for folic acid supplementation and stroke mortality risk (RCTs). The majority of the population came from one study (Huo et al., 2015-CSPPT) that was conducted in a chinese population only. One study was secondary prevention and one study was primary prevention. m. Serious imprecision for folic acid supplementation and stroke mortality risk (RCTs) as the 95% CI (RR, 0.88-3.93) include both clinically important benefit (RR<0.95) and harm (RR>1.05). n. Serious indirectness for folic acid supplementation and all-cause mortality risk (RCTs). The majority of the population came from one study (Huo et al., 2015-CSPPT) that was conducted in a chinese population only. Two studies were secondary prevention, seven studies were primary prevention and one study was conducted in heart transplantation recipients. o. Serious imprecision for folic acid supplementation and all-cause mortality risk (RCTs), as the 95% CI (RR, 0.72-1.05) overlaps with the minimally important difference for clinical benefit (RR<0.95).

Supplementary Table 13. GRADE assessment for niacin (B3)


Certainty№ of



Vitamin B3 (Niacin) and total CVD risk (random effects model) - RCTs

4 randomised trials


7 fewer per 1,000

(from 7 more to 17

fewer)


Vitamin B3 (Niacin) and total CVD risk - No statins (random effects model) - RCTs

1 randomised trials

not serious not serious d serious e not serious none c RR 0.98(0.95 to 1.01)

17 fewer per 1,000

(from 8 more to 42

fewer)


Vitamin B3 (Niacin) and total CVD risk - Background statin treatment (random effects model) - RCTs

3 randomised trials

not serious not serious not serious f serious g none c RR 0.97(0.81 to 1.17)

4 fewer per 1,000

(from 22 more to 25

fewer)


Vitamin B3 (Niacin) and total CHD risk (random effects model) - RCTs

2 randomised trials

not serious not serious not serious serious h none c RR 0.96(0.87 to 1.07)

2 fewer per 1,000

(from 4 more to 7

fewer)


Vitamin B3 (Niacin) and MI risk (random effects model) - RCTs

5 randomised trials


5 fewer per 1,000

(from 3 more to 13

fewer)


62


Certainty№ of



Vitamin B3 (Niacin) and MI risk - No statins (random effects model) - RCTs

1 randomised trials

not serious not serious d serious k not serious none c RR 0.74(0.60 to 0.90)


fewer to 55 fewer)


Vitamin B3 (Niacin) and MI risk- Background statin treatment (random effects model) - RCTs

4 randomised trials

not serious not serious not serious l serious m none c RR 0.96(0.85 to 1.08)

1 fewer per 1,000

(from 3 more to 5

fewer)


Vitamin B3 (Niacin) and stroke risk (random effects model) - RCTs

5 randomised trials

not serious not serious n not serious o serious p none c RR 0.95(0.72 to 1.26)

2 fewer per 1,000

(from 12 more to 13

fewer)


Vitamin B3 (Niacin) and stroke risk - No statins (random effects model) - RCTs

1 randomised trials

not serious not serious d not serious q not serious none c RR 0.76(0.61 to 0.95)

27 fewer per 1,000

(from 6 fewer to 43

fewer)

⨁⨁⨁⨁HIGH

Vitamin B3 (Niacin) and stroke risk - Background statin treatment (random effects model) - RCTs

4 randomised trials

not serious not serious not serious r serious s none c RR 1.10(0.70 to 1.72)

3 more per 1,000


more)


Vitamin B3 (Niacin) and CVD mortality risk (random effects model) - RCTs

3 randomised trials

not serious not serious not serious t serious u none c RR 0.95(0.84 to 1.08)

7 fewer per 1,000

(from 12 more to 24

fewer)


Vitamin B3 (Niacin) and CVD mortality risk - No statins (random effects model) - RCTs

1 randomised trials

not serious not serious d serious v serious w none c RR 0.94(0.82 to 1.07)


more to 41 fewer)

⨁⨁◯◯LOW

Vitamin B3 (Niacin) and CVD mortality risk - Background statin treatment (random effects model) - RCTs

2 randomised trials

not serious not serious not serious x serious y none c RR 1.14(0.75 to 1.73)

3 more per 1,000

(from 6 fewer to 16

more)


Vitamin B3 (Niacin) and CHD mortality risk (random effects model) - RCTs

63


Certainty№ of



3 randomised trials

not serious not serious not serious z serious aa none c RR 0.99(0.89 to 1.10)

0 fewer per 1,000

(from 5 fewer to 5

more)


Vitamin B3 (Niacin) and CHD mortality risk - No statins (random effects model) - RCTs

1 randomised trials

not serious not serious d not serious ab serious ac none c RR 0.95(0.82 to 1.09)


more to 35 fewer)


Vitamin B3 (Niacin) and CHD mortality risk - Background statin treatment (random effects model) - RCTs

2 randomised trials

not serious not serious not serious ad serious ae none c RR 1.04(0.90 to 1.21)

1 more per 1,000

(from 2 fewer to 5

more)


Vitamin B3 (Niacin) and all-cause mortality risk (random effects model) - RCTs

4 randomised trials

not serious not serious not serious af serious ag none c RR 1.04(0.95 to 1.14)

4 more per 1,000

(from 4 fewer to 12

more)


Vitamin B3 (Niacin) and all-cause mortality - No statins (random effects model) - RCTs

1 randomised trials

not serious not serious d serious ah serious ai none c RR 0.96(0.85 to 1.08)


more to 38 fewer)

⨁⨁◯◯LOW

Vitamin B3 (Niacin) and all-cause mortality risk - Background statin treatment (random effects model) - RCTs

3 randomised trials

not serious not serious not serious aj serious ak none c RR 1.1(1.0 to 1.2)

6 more per 1,000

(from 0 fewer to 11

more)



Explanationsa. No serious indirectness for vitamin B3 (niacin) supplementation and total CVD risk (RCTs), despite all studies being in populations with a history of CVD or CHD; the total population was also high (29,254). b. Serious imprecision for vitamin B3 (niacin) supplementation and total CVD risk (RCTs), as the 95% CI (RR, 0.93-1.03) overlaps with the minimally important difference for clinical benefit (RR<0.95). c. Publication bias was not assessed since there were <10 studies. d. Not able to assess inconsistency as only one study was included. e. Serious indirectness for vitamin B3 (niacin) supplementation and total CVD risk (RCTs-no statins) as the one included study was conducted in a population with CHD. f. No serious indirectness for vitamin B3 (niacin) supplementation and total CVD risk (RCTs-background statin). All three included studies were secondary prevention. g. Serious imprecision for vitamin B3 (niacin) supplementation and total CVD risk (RCTs-background statin), as the 95% CI (RR, 0.81-1.17) include both clinically important benefit (RR<0.95) and harm (RR>1.05). h. Serious imprecision for vitamin B3 (niacin) supplementation and total CHD risk (RCTs), as the 95% CI (RR, 0.87-1.07) include both clinically important benefit (RR<0.95) and harm (RR>1.05). i. No serious indirectness for vitamin B3 (niacin) supplementation and MI risk (RCTs). Four studies were secondary prevention and one study was primary prevention. j. Serious imprecision for vitamin B3 (niacin) supplementation and MI risk (RCTs-random effects model), as the 95% CI (RR, 0.75-1.06) include both clinically important benefit (RR<0.95) and harm (RR>1.05).

64

k. Serious indirectness for vitamin B3 (niacin) supplementation and MI risk (RCTs-no statins) as the one included study was conducted in a population with CHD. l. No serious indirectness for vitamin B3 (niacin) supplementation and MI risk (RCTs-background statin). All four included studies were secondary prevention. m. Serious imprecision for vitamin B3 (niacin) supplementation and MI risk (RCTs-background statin), as the 95% CI (RR 0.85-1.08) include both clinically important benefit (RR<0.95) and harm (RR>1.05). n. Although there was evidence of serious inconsistency for vitamin B3 (niacin) supplementation and stroke risk (RCTs) (I²=60%, p=0.04), independent removal of one trial (CDPRG 1975) explained almost half of the heterogeneity (I²=37%, P=0.19). o. No serious indirectness for vitamin B3 (niacin) supplementation and stroke risk (RCTs). Four studies were secondary prevention and one study was primary prevention. p. Serious imprecision for vitamin B3 (niacin) supplementation and stroke risk (RCTs-random effects model), as the 95% CI (RR, 0.72-1.26) include both clinically important benefit (RR<0.95) and harm (RR>1.05). q. No serious indirectness for vitamin B3 (niacin) supplementation and stroke risk (RCTs-no statins). The included study was secondary prevention. r. No serious indirectness for vitamin B3 (niacin) supplementation and stroke risk (RCTs-backgroud statin). Three studies were secondary prevention and one study was primary prevention. s. Serious imprecision for vitamin B3 (niacin) supplementation and stroke risk (RCTs-background statin) (random effects model) as the 95% CI (RR, 0.70-1.72) include both clinically important benefit (RR<0.95) and harm (RR>1.05). t. No serious indirectness for vitamin B3 (niacin) supplementation and CVD mortality risk (RCTs), despite all studies being in populations with a history of CVD or CHD; the total population was also high (7,489). u. Serious imprecision for vitamin B3 (niacin) supplementation and CVD mortality risk (RCTs), as the 95% CI (RR, 0.84-1.08) include both clinically important benefit (RR<0.95) and harm (RR> 1.05). v. Serious indirectness for niacin (B3) supplementation and CVD mortality risk (RCTs-no statins) as the one included study was conducted in a population with CHD. w. Serious imprecision for vitamin B3 (niacin) supplementation and CVD mortality risk (RCTs-no statins), as the 95% CI (RR, 0.82-1.07) include both clinically important benefit (RR<0.95) and harm (RR>1.05). x. No serious indirectness for vitamin B3 (niacin) supplementation and CVD mortality risk (RCTs-background statin). Both included studies were secondary prevention. y. Serious imprecision for vitamin B3 (niacin) supplementation and CVD mortality risk (RCTs-background statin), as the 95% CI (RR 0.75, 1.73) include both clinically important benefit (RR<0.95) and harm (RR>1.05). z. No serious indirectness for vitamin B3 (niacin) supplementation and CHD mortality risk (RCTs). All three studies were secondary prevention. aa. Serious imprecision for vitamin B3 (niacin) supplementation and CHD mortality risk (RCTs), as the 95% CI (RR, 0.89-1.10) include both clinically important benefit (RR<0.95) and harm (RR>1.05). ab. No serious indirectness for vitamin B3 (niacin) supplementation and CHD mortality risk (RCTs-no statin). The one included study was secondary prevention. ac. Serious imprecision for vitamin B3 (niacin) supplementation and CHD mortality risk (RCTs-no statins), as the 95% CI (RR, 0.82-1.09) includes both clinically important benefit (RR<0.95) and harm (RR>1.05). ad. No serious indirectness for vitamin B3 (niacin) supplementation and CHD mortality risk (RCTs-background statin). Both studies were secondary prevention. ae. Serious imprecision for vitamin B3 (niacin) supplementation and CHD mortality risk (RCTs-background statin), as the 95% CI (RR, 0.90-1.21) include both clinically important benefit (RR<0.95) and harm (RR>1.05). af. No serious indirectness for vitamin B3 (niacin) supplementation and all-cause mortality risk (RCTs), despite all studies being in populations with a history of CVD or CHD; the total population was also high (33,103). ag. Serious imprecision for vitamin B3 (niacin) supplementation and all-cause mortality risk (RCTs), as the 95% CI (RR, 0.95-1.14) overlaps with the minimally important difference for clinical harm (RR>1.05). ah. Serious indirectness for vitamin B3 (niacin) supplementation and all-cause mortality risk (RCTs-no statin) as the one included study was conducted in a population with CHD. ai. Serious imprecision for vitamin B3 (niacin) supplementation and all-cause mortality risk (RCTs-no statin), as the 95% CI (RR, 0.85-1.08) include both clinically benefit (RR<0.95) and harm (RR>1.05) aj. No serious indirectness for vitamin B3 (niacin) supplementation and all-cause mortality risk (RCTs-background statin). All 3 studies were secondary prevention. ak. Serious imprecision for vitamin B3 (niacin) supplementation and all-cause mortality risk (RCTs-background statin) as the 95% CI (RR, 1.0-1.2) overlaps with the minimally important difference for clinical harm (RR>1.05)

Supplementary Table 14. GRADE assessment for vitamin B6


Certainty№ of



Vitamin B6 (Pyridoxine) and total CVD risk (random effects model) - RCTs

1 randomised trials

not serious not serious a not serious b serious c none d RR 1.03(0.85 to 1.24)

5 more per 1,000


more)


Vitamin B6 (Pyridoxine) and MI risk (random effects model) - RCTs

2 randomised trials

not serious not serious not serious e serious f none d RR 1.04(0.87 to 1.23)

5 more per 1,000


more)


Vitamin B6 (Pyridoxine) and stroke risk (random effects model) - RCTs

2 randomised trials

not serious not serious not serious g serious h none d RR 0.92(0.61 to 1.40)

2 fewer per 1,000


more)


Vitamin B6 (Pyridoxine) and MI mortality risk (random effects model) - RCTs

65


Certainty№ of



1 randomised trials

not serious not serious a not serious i serious j none d RR 1.04(0.74 to 1.48)

3 more per 1,000


more)


Vitamin B6 (Pyridoxine) and all-cause mortality risk (random effects model) - RCTs

2 randomised trials

not serious not serious not serious k serious l none d RR 1.02(0.80 to 1.30)

1 more per 1,000


more)



Explanationsa. Unable to assess inconsistency as only one study was included. b. No serious indirectness for vitamin B6 (pyridoxine) supplementation and total CVD risk (RCTs), although the single study was secondary prevention. c. Serious imprecision for vitamin B6 (pyridoxine) supplementation and total CVD risk (RCTs), as the 95% CI (RR, 0.85-1.24) include both clinically important benefit (RR<0.95) and harm (RR>1.05). d. Publication bias was not assessed since there were <10 studies. e. No serious indirectness for vitamin B6 (pyridoxine) supplementation and MI risk (RCTs), although both included studies were secondary prevention. f. Serious imprecision for vitamin B6 (pyridoxine) supplementation and MI risk (RCTs), as the 95% CI (RR, 0.87- 1.23) include both clinically important benefit (RR<0.95) and harm (RR>1.05). g. No serious indirectness for vitamin B6 (pyridoxine) supplementation and stroke risk (RCTs), although both included studies were secondary prevention. h. Serious imprecision for vitamin B6 (pyridoxine) supplementation and stroke risk (RCTs), as the 95% CI (RR, 0.61-1.40) include both clinically important benefit (RR<0.95) and harm (RR>1.05). i. No serious indirectness for vitamin B6 (pyridoxine) supplementation and MI mortality risk (RCTs), although the one included study was conducted in a population with MI. It was a secondary prevention study. j. Serious imprecision for vitamin B6 (pyridoxine) supplementation and MI mortality risk (RCTs), as the 95% CI (RR, 0.74-1.48) include both clinically important benefit (RR<0.95) and harm (RR>1.05). k. No serious indirectness for vitamin B6 (pyridoxine) supplementation and all-cause mortality risk (RCTs), although both included studies were conducted in a population with MI and suspected CAD. Both were secondary prevention studies. l. Serious imprecision for vitamin B6 (pyridoxine) supplementation and all-cause mortality risk (RCTs), as the 95% CI (RR 0.80, 1.30) include both clinically important benefit (RR<0.95) and harm (RR>1.05).

Supplementary Table 15. GRADE assessment for calcium


№ of studies



Calcium and total CVD risk (random effects model) - RCTs

3 randomised trials

not serious serious a serious b serious c none d RR 1.43(0.79 to 2.59)

41 more per 1,000


more)


Calcium and total CHD risk (random effects model) - RCTs

2 randomised trials

not serious not serious serious e serious f none d RR 1.16(0.87 to 1.56)

8 more per 1,000

(from 6 fewer to 27

more)

⨁⨁◯◯LOW

Calcium and MI risk (random effects model) - RCTs

4 randomised trials

not serious serious g not serious h serious i none d RR 1.69(0.94 to 3.04)

24 more per 1,000

(from 2 fewer to 72

more)

⨁⨁◯◯LOW

66


№ of studies



Calcium and stroke risk (random effects model) - RCTs

3 randomised trials

not serious not serious serious j serious k none d RR 1.29(0.96 to 1.72)

12 more per 1,000

(from 2 fewer to 29

more)

⨁⨁◯◯LOW

Calcium and CVD mortality risk (random effects model) - RCTs

2 randomised trials

not serious not serious l serious m very serious n none d RR 1.24(0.27 to 5.65)

4 more per 1,000


more)


Calcium and MI mortality risk (random effects model) - RCTs

1 randomised trials

not serious not serious o serious p serious q none d RR 1.44(0.62 to 3.36)

5 more per 1,000

(from 5 fewer to 29

more)

⨁⨁◯◯LOW

Calcium and stroke mortality risk (random effects model) - RCTs

1 randomised trials

not serious not serious o serious r very serious s none d RR 0.75(0.26 to 2.15)

3 fewer per 1,000

(from 8 fewer to 13

more)


Calcium and all-cause mortality risk (random effects model) - RCTs

6 randomised trials

not serious not serious not serious t serious u none d RR 1.08(0.97 to 1.21)

9 more per 1,000

(from 3 fewer to 22

more)



Explanationsa. Serious inconsistency for calcium supplementation and total CVD risk (RCTs), as I²=80% and p=0.007. b. Serious indirectness for calcium supplementation and total CVD risk (RCTs), as all of the included studies were conducted in post-menopausal women. c. Serious imprecision for calcium supplementation and total CVD risk (RCTs), as the 95% CI (RR, 0.79-2.59) include both clinically important benefit (RR<0.95) and harm (RR>1.05). d. Publication bias was not assessed as there were <10 studies. e. Serious indirectness for calcium supplementation and total CHD risk (RCTs), as 29% of the population have a history of large bowel adenomas and the remaining 71% have low trauma fracture. f. Serious imprecision for calcium supplementation and total CHD risk (RCTs), as the 95% CI (RR, 0.87-1.56) include both clinically important benefit (RR<0.95) and harm (RR>1.05). g. Serious inconsistency for calcium supplementation and MI risk (RCTs), as I²=69% and p=0.02. h. No serious indirectness for calcium supplementation and MI risk (RCTs). All studies were primary prevention. i. Serious imprecision for calcium supplementation and MI risk (RCTs), as the 95% CI (RR, 0.94-3.04) include both clinically important benefit (RR<0.95) and harm (RR>1.05). j. Serious indirectness for calcium supplementation and stroke risk (RCTs), as 24% of total participants had a history of large bowel adenomas and 76% of total participants are post-menopausal women. k. Serious imprecision for calcium supplementation and stroke risk (RCTs), as the 95% CI (RR, 0.96-1.72) overlaps with the minimally important difference for clinical harm (RR>1.05). l. No serious inconsistency for calcium supplementation and CVD mortality risk (RCTs), although I²=53% and p=0.14. m. Serious indirectness for calcium supplementation and total CVD mortality risk (RCTs), as both included studies were conducted in post-menopausal women. n. Very serious imprecision for calcium supplementation and CVD mortality risk (RCTs), as the 95% CI (RR, 0.27-5.65) include both appreciable clinically important benefit (RR<0.95) and harm (RR>1.05). o. Unable to assess inconsistency as only one study was included. p. Serious indirectness for calcium supplementation and MI mortality risk (RCTs), as the only study was conducted in post-menopausal women. q. Serious imprecision for calcium supplementation and MI mortality risk (RCTs), as the 95% CI (RR, 0.62-3.36) include both clinically important benefit (RR<0.95) and harm (RR>1.05). r. Serious indirectness for calcium supplementation and stroke mortality risk (RCTs), as the only included study was conducted in post-menopausal women. s. Very serious imprecision for calcium supplementation and stroke mortality risk (RCTs), as the 95% CI (RR, 0.26-2.15) include both appreciable clinically important benefit (RR<0.95) and harm (RR>1.05). t. No serious indirectness for calcium supplementation and all-cause mortality risk (RCTs). All studies were primary prevention. u. Serious imprecision for calcium supplementation and all-cause mortality risk (RCTs), as the 95% CI (RR, 0.97-1.21) overlaps with the minimally important difference for clinical harm (RR>1.05).

67

Supplementary Table 16. GRADE assessment for iron


№ of studies



Iron and total CVD risk (IV iron for iron deficiency in heart failure) (random effects model) - RCTs

1 randomised trials

not serious not serious a serious b not serious none c RR 0.49(0.28 to 0.86)


fewer to 102 fewer)


Iron and MI risk (IV iron for iron deficiency in heart failure) (random effects model) - RCTs

1 randomised trials

not serious not serious a serious d very serious e none c RR 0.34(0.06 to 2.01)


fewer to 20 more)


Iron and CVD mortality risk (IV iron for iron deficiency in heart failure) (random effects model) - RCTs

2 randomised trials

not serious not serious not serious f serious g none c RR 0.80(0.40 to 1.58)


more to 31 fewer)


Iron and all-cause mortality risk (Iron for iron deficiency in heart failure) (random effects model) - RCTs

2 randomised trials

not serious not serious serious h serious i none c RR 0.79(0.42 to 1.51)


more to 34 fewer)

⨁⨁◯◯LOW


Explanationsa. Unable to assess inconsistency as only one study was included. b. Serious indirectness for iron supplementation and total CVD risk (RCTs), as the one included study was conducted in patients with chronic heart failure. c. Publication bias was not assessed as there were <10 studies. d. Serious indirectness for iron supplementation and MI risk (RCTs), as the one included study was conducted in patients with chronic heart failure. e. Very serious imprecision for iron supplementation and MI risk (RCTs), as the 95% CI (RR, 0.06-2.01) include both appreciable clinically important benefit (RR<0.95) and harm (RR>1.05). f. No serious indirectness for iron supplementation and total CVD mortality risk (RCTs), even though there were only two studies, one study was a multi-center study and included several countries and both studies were conducted in high risk populations for CVD mortality. g. Serious imprecision for iron supplementation and CVD mortality risk (RCTs), as the 95% CI (RR, 0.40-1.58) include both clinically important benefit (RR<0.95) and harm (RR>1.05). h. Serious indirectness for iron supplementation and all-cause mortality risk (RCTs), as both studies were conducted in patients with heart failure. i. Serious imprecision for iron supplementation and all-cause mortality risk (RCTs), as the 95% CI (RR, 0.42-1.51) include both clinically important benefit (RR<0.95) and harm (RR>1.05).

Supplementary Table 17. GRADE assessment for multivitamin


Certainty№ of



Multivitamin and total CVD risk (random effects model) - RCTs

2 randomised trials


6 fewer per 1,000

(from 20 more to 27

fewer)

⨁⨁◯◯LOW

Multivitamin and MI risk (random effects model) - RCTs

68


Certainty№ of



3 randomised trials

not serious not serious not serious e serious f none c RR 0.95(0.82 to 1.09)

2 fewer per 1,000

(from 4 more to 9

fewer)


Multivitamin and stroke (random effects model) - RCTs

2 randomised trials

not serious serious g serious h serious i none c RR 0.86(0.46 to 1.62)

6 fewer per 1,000


more)


Multivitamin and CVD mortality risk (random effects model) - RCTs

3 randomised trials

not serious not serious serious j serious k none c RR 0.94(0.83 to 1.06)

4 fewer per 1,000

(from 4 more to 10

fewer)

⨁⨁◯◯LOW

Multivitamin and MI mortality risk (random effects model) - RCTs

1 randomised trials

not serious not serious d serious l serious m none c RR 0.63(0.39 to 1.02)

2 fewer per 1,000

(from 0 fewer to 4

fewer)

⨁⨁◯◯LOW

Multivitamins and stroke mortality risk (random effects model) - RCTs

2 randomised trials

not serious not serious not serious n serious o none c RR 0.88(0.51 to 1.51)

1 fewer per 1,000

(from 6 fewer to 6

more)


Multivitamin and all-cause mortality risk (random effects model) - RCTs


not serious not serious not serious p serious q none RR 0.95(0.90 to 1.01)

8 fewer per 1,000

(from 2 more to 16

fewer)



Explanationsa. Serious indirectness for multivitamin supplementation and total CVD risk (RCTs) as one of the included studies (Sesso et al., 2012- PHS II), which represented ~90% of the population, was conducted in males only. One study was secondary prevention and the other one was primary prevention. b. Serious imprecision for multivitamin supplementation and total CVD risk (RCTs), as the 95% CI (RR, 0.77-1.17) include both clinically important benefit (RR<0.95) and harm (RR>1.05). c. Publication bias was not assessed since there were <10 studies. d. Inconsistency was not assessed as only one study was included. e. No serious indirectness for multivitamin supplementation and MI risk (RCTs). Majority of the studies were primary prevention, with the exception of one (Lamas et al., 2013) which included individuals post MI. f. Serious imprecision for multivitamin supplementation and MI risk (RCTs), as the 95% CI (RR, 0.82-1.09) include both clinically important benefit (RR<0.95) and harm (RR>1.05). g. Serious inconsistency for multivitamin supplementation and stroke risk (RCTs), as I²=59% and p=0.12 (effect estimates for the two included studies go in opposite directions). h. Serious indirectness for multivitamin supplementation and stroke risk (RCTs), as one of the included studies (Sesso et al., 2012- PHS II), which represented ~83% of the population, was conducted in males only. One included study was primary prevention and the other one was for secondary prevention. i. Serious imprecision for multivitamin supplementation and stroke risk (RCTs), as the 95% CI (RR, 0.46-1.62) include both clinically important benefit (RR<0.95) and harm (RR>1.05). j. Serious indirectness for multivitamin supplementation and CVD mortality risk (RCTs), as one of the included studies (Sesso et al., 2012- PHS II), which represented ~83% of the population, was conducted in males only. k. Serious imprecision for multivitamin supplementation and CVD mortality risk (RCTs), as the 95% CI (RR, 0.83-1.06) include both clinically important benefit (RR<0.95) and harm (RR>1.05). l. Serious indirectness for multivitamin supplementation and MI mortality risk (RCTs), as the one included study (Sesso et al., 2012- PHS II), was conducted in males only. m. Serious imprecision for multivitamin supplementation and MI mortality risk (RCTs), as the 95% CI (RR, 0.39-1.02) overlaps with the minimally important difference for clinical benefit (RR<0.95). n. No serious indirectness for multivitamin supplementation and stroke mortality risk (RCTs). One study was for primary prevention and the second one, Sesso et al., 2012 was a combination of both.

69

o. Serious imprecision for multivitamin supplementation and stroke mortality risk (RCTs), as the 95% CI (RR, 0.49-1.63) include both clinically important benefit (RR<0.95) and harm (RR>1.05). p. No serious indirectness for multivitamin supplementation and all-cause mortality (RCTs). Majority of the studies were for primary prevention with the exception for one study (Lamas et al., 2013) which included individuals post MI. q. Serious imprecision for multivitamin supplementation and all-cause mortality risk (RCTs), as the 95% CI (RR, 0.90-1.01) overlaps with the minimally important difference for clinical benefit (RR<0.95).

Supplementary Table 18. GRADE assessment for calcium and vitamin D


Certainty№ of



VitD+Ca and total CVD risk (random effects model) - RCTs

2 randomised trials


3 more per 1,000

(from 6 fewer to 13

more)

⨁⨁◯◯LOW

VitD+Ca and total CHD risk (random effects model) - RCTs

2 randomised trials

not serious not serious serious e very serious f none c RR 0.73(0.20 to 2.68)

8 fewer per 1,000


more)


VitD+Ca and MI risk (random effects model) - RCTs

5 randomised trials

not serious not serious serious g serious h none c RR 1.14(0.95 to 1.37)

3 more per 1,000

(from 1 fewer to 8

more)

⨁⨁◯◯LOW

VitD+Ca and stroke risk (random effects model) - RCTs

7 randomised trials


4 more per 1,000

(from 0 fewer to 9

more)


Vit D+Ca and CVD mortality risk (random effects model) - RCTs

1 randomised trials

not serious not serious d serious k serious l none c RR 0.92(0.77 to 1.10)

1 fewer per 1,000

(from 1 more to 3

fewer)

⨁⨁◯◯LOW

VitD+Ca and CHD mortality risk (random effects model) - RCTs

2 randomised trials

not serious not serious serious m serious n none c RR 0.94(0.54 to 1.64)

0 fewer per 1,000

(from 3 fewer to 5

more)

⨁⨁◯◯LOW

Vit D+Ca and MI mortality risk (random effects model) - RCTs

1 randomised trials

not serious not serious d serious o very serious p none c RR 2.97(0.12 to 72.26)

0 fewer per 1,000

(from 0 fewer to 0

fewer)


VitD+Ca and stroke mortality risk (random effects model) - RCTs

70


Certainty№ of



2 randomised trials

not serious not serious serious q serious r none c RR 0.90(0.63 to 1.29)

0 fewer per 1,000

(from 1 fewer to 1

more)

⨁⨁◯◯LOW

VitD+Ca and all-cause mortality (random effects model) - RCTs


not serious not serious not serious s serious t none RR 0.95(0.89 to 1.01)

4 fewer per 1,000

(from 1 more to 10

fewer)



Explanationsa. Serious indirectness for calcium and vitamin D supplementation and total CVD risk (RCTs), as both studies were in women only. b. Serious imprecision for calcium and vitamin D supplementation and total CVD risk (RCTs), as the 95% CI (RR, 0.94-1.12) include both clinically important benefit (RR<0.95) and harm (RR>1.05). c. Publication bias was not assessed as there were < 10 studies. d. Unable to assess inconsistency as there was only 1 study included. e. Serious indirectness for calcium and vitamin D supplementation and total CHD risk (RCTs), as nearly 100% of the population were women. f. Very serious imprecision for calcium and vitamin D supplementation and total CHD risk (RCTs), as the 95% CI (RR, 0.20-2.68) include both appreciable clinically important benefit (RR<0.95) and harm (RR>1.05). g. Serious indirectness for calcium and vitamin D supplementation and MI risk (RCTs), as 95% of the population were women. All studies were primary prevention. h. Serious imprecision for calcium and vitamin D supplementation and MI risk (RCTs), as the 95% CI (RR, 0.95-1.37) overlaps with minimally important difference for clinical harm (RR>1.05). i. Serious indirectness for calcium and vitamin D supplementation and stroke risk (RCTs). All studies were primary prevention. j. Serious imprecision for calcium and vitamin D supplementation and stroke (RCTs), as the 95% CI (RR, 0.98-1.39) overlaps with minimally important difference for clinical harm (RR>1.05). k. Serious indirectness for calcium and vitamin D supplementation and CVD mortality risk (RCTs), as the only included study was conducted in women only. l. Serious imprecision for calcium and vitamin D supplementation and CVD mortality (RCTs), as the 95% CI (RR, 0.77-1.10) include both clinically important benefit (RR<0.95) and harm (RR>1.05). m. Serious indirectness for calcium and vitamin D supplementation and CHD mortality risk (RCTs), as >99% of the population were women. n. Serious imprecision for calcium and vitamin D supplementation and CHD mortality risk (RCTs), as the 95% CI (RR, 0.54-1.64) include both clinically important benefit (RR<0.95) and harm (RR>1.05). o. Serious indirectness for calcium and vitamin D supplementation and MI mortality risk (RCTs), as the only study was conducted in early menopausal women only. p. Very serious imprecision for calcium and vitamin D supplementation and MI mortality risk (RCTs), as the 95% CI (RR, 0.12-72.26) include both appreciable clinically important benefit (RR<0.95) and harm (RR>1.05). q. Serious indirectness for calcium and vitamin D supplementation and stroke mortality risk (RCTs), as >99% of the population were women. r. Serious imprecision for calcium and vitamin D supplementation and stroke mortality risk (RCTs), as the 95% CI (RR, 0.63-1.29) include both clinically important benefit (RR<0.95) and harm (RR>1.05). s. No serious indirectness for calcium and vitamin D supplementation and all-cause mortality risk (RCTs). All but one study were primary prevention, the sole secondary prevention study was conducted in ischemic stroke survivors (Gupta et al., 2016). t. Serious imprecision for calcium and vitamin D supplementation and all-cause mortality risk (RCTs), as the 95% CI (RR, 0.89-1.01) overlaps with minimally important difference for clinical benefit (RR<0.95).

71

Supplementary Table 19. Summary of the meta-analayses results for all-cause mortality, CVD mortality, total CVD risk and other significant assocations

Test Control

All-cause mortality 0 43 RCTs 6m-5.5y 0.99 0.95, 1.03 0.58 2,908/18,719 2,968/18,831 0 Random HighCVD mortality 0 2 RCTs 5-5.5y 0.86 0.66, 1.10 0.23 101/1,952 118/1,955 0 Random LowTotal CVD risk 0 6 RCTs 6m-5.5y 0.95 0.86, 1.05 0.31 505/3,275 532/3,271 0 Random Moderate

All-cause mortality 0 1 RCT 3y 0.99 0.56, 1.72 0.96 24/1,157 24/1,140 N/A Random Low

All-cause mortality 0 6 RCTs 4.5-12y 1.03 0.96, 1.11 0.37 2,181/21,188 2,103/21,154 17 Random ModerateCVD mortality 0 4 RCTs 4.5-12y 1.10 0.98, 1.23 0.11 623/16,834 568/16,834 0 Random ModerateTotal CVD risk 0 3 RCTs 2.1-12y 1.03 0.96, 1.10 0.44 1,476/35,059 1,436/35,059 0 Random ModerateMI mortality - 1 RCT 9.4y 0.42 0.20, 0.87 0.02 10/4,084 24/4,087 N/A Random Moderate

All-cause mortality + 21 RCTs 6m-9.4y 1.06 1.00, 1.12 0.05 4,372/53,153 4,100/52,627 17 Random ModerateCVD mortality 0 7 RCTs 3.3-8y 1.02 0.94, 1.10 0.65 1,141/2,4867 1,122/2,4862 0 Random ModerateTotal CVD risk 0 7 RCTs 2-8y 0.99 0.95, 1.04 0.77 3,822/30,423 3,844/30,403 0 Random High

All-cause mortality 0 32 RCTs 6m-10.1y 1.00 0.97, 1.04 0.87 4,497/61,506 4,487/61,495 0 Random HighCVD mortality 0 11 RCTs 1.4y-10.1y 0.95 0.89, 1.02 0.16 1,404/51,257 1,483/51,276 0 Random ModerateTotal CVD risk 0 10 RCTs 1.4y-10.1y 0.96 0.89, 1.03 0.23 3,916/50,194 4,020/50,281 50 Random Low

All-cause mortality 0 4 RCTs 3-10.3y 1.02 0.94, 1.11 0.63 921/8,020 898/7,984 0 Random ModerateCVD mortality 0 2 RCTs 8-9.4y 1.07 0.92, 1.25 0.35 335/7,760 311/7,737 0 Random ModerateTotal CVD risk 0 2 RCTs 1.4-10.1y 0.99 0.90, 1.10 0.89 728/7,760 731/7,737 0 Random Moderate

All-cause mortality 0 4 RCTs 10m-7.6y 0.99 0.88, 1.12 0.87 493/9,662 499/9,711 0 Random LowCVD mortality 0 3 RCTs 6m-7.6y 1.00 0.66, 1.51 1.00 169/9,400 177/9,451 48 Random LowTotal CVD risk 0 2 RCTs 5.5-7.6y 1.04 0.96, 1.12 0.34 1,183/9,360 1,146/9,410 0 Random Low

All-cause mortality 0 16 RCTs 6m-7.3y 1.02 0.97, 1.06 0.41 3,153/22,701 3,092/22,723 0 Random ModerateCVD mortality 0 5 RCTs 3.4-7.3y 0.98 0.87, 1.11 0.74 1,304/16,843 1,338/16,850 52 Random ModerateTotal CVD risk 0 9 RCTs 2-7.3y 0.98 0.93, 1.04 0.58 3,426/19,877 3,462/19,879 29 Random Moderate

Stroke - 12 RCTs 6m-7.3y 0.90 0.81, 1.00 0.04 986/21,664 1,081/21,675 16 Random Moderate

All-cause mortality 0 10 RCTs 1-10y 0.87 0.72, 1.05 0.14 415/12,792 462/12,788 17 Random ModerateCVD mortality 0 5 RCTs 2-5.3y 0.89 0.68, 1.17 0.41 88/11,235 100/11,233 0 Random ModerateTotal CVD risk - 5 RCTs 1-4.5y 0.83 0.73, 0.93 0.002 435/10,788 525/10,779 0 Random High

Stroke risk - 7 RCTs 1-10y 0.80 0.69, 0.93 0.003 309/12,268 385/12,257 0 Random Moderate

All-cause mortality 0 4 RCTs 1-6.2y 1.04 0.95, 1.14 0.36 1,167/15,727 1,524/17,376 16 Random ModerateCVD mortality 0 3 RCTs 3-6.2y 0.95 0.84, 1.08 0.46 284/2,924 673/4,565 0 Random ModerateTotal CVD risk 0 4 RCTs 3-6.2y 0.97 0.93, 1.03 0.33 2,785/15,762 4,260/17,400 34 Random Moderate

MI risk 0 5 RCTs 6m-6.2y 0.90 0.75, 1.06 0.20 623/16,432 913/17,672 41 Random ModerateStroke risk 0 5 RCTs 6m-6.2y 0.95 0.72, 1.26 0.73 623/16,432 830/17,672 60 Random Moderate

All-cause mortality 0 1 RCT 6.2y 0.96 0.85, 1.08 0.51 273/1,119 709/2,789 N/A Random LowCVD mortality 0 1 RCT 6.2y 0.94 0.82, 1.07 0.33 238/1,119 633/2,789 N/A Random LowTotal CVD risk 0 1 RCT 6.2y 0.98 0.95, 1.01 0.15 914/1,119 2,333/2,789 N/A Random Moderate

MI risk - 1 RCT 6.2y 0.74 0.60, 0.90 0.002 114/1,119 386/2,789 N/A Random ModerateStroke risk - 1 RCT 6.2y 0.76 0.61, 0.95 0.01 95/1,119 311/2,789 N/A Random High

All-cause mortality + 3 RCTs 1-3.9y 1.10 1.00,1.20 0.05 894/14,608 815/14,587 0 Random ModerateCVD mortality 0 2 RCTs 3y 1.14 0.75, 1.73 0.55 46/1,805 40/1,776 0 Random ModerateTotal CVD risk 0 3 RCTs 3-3.9y 0.97 0.81, 1.17 0.77 1,871/14,643 1,927/14,611 55 Random Moderate

MI risk 0 4 RCTs 6m-3.9y 0.96 0.85, 1.08 0.53 509/15,313 527/14,883 0 Random ModerateStroke risk 0 4 RCTs 6m-3.9y 1.10 0.70, 1.72 0.68 528/15,313 519/14,883 37 Random Moderate

Study lengths I2 (%)

Model (R/F)

Grade95% CI P-valueSupplements EffectNumber of

RCTs/CohortsRisk

Ratio Participant number

Beta carotene

Vitamin A

Vitamin D

Vitamin C

Vitamin B3 (Niacin) - overall effect

Folic acid

Selenium

B-Complex

Vitamin E

Antioxidants

Vitamin B3 (Niacin) - no statin

Vitamin B3 (Niacin)- with background statin

72

Supplementary Table 20 (continued). Summary of the meta-analayses results for all-cause mortality, CVD mortality, total CVD risk and other significant assocations

All-cause mortality 0 2 RCTs 3.2-3.3y 1.02 0.80, 1.30 0.88 120/1,706 119/1,723 0 Random ModerateTotal CVD risk 0 1 RCT 3.3y 1.03 0.85, 1.24 0.78 175/934 172/943 N/A Random Moderate

All-cause mortality 0 6 RCTs 2-8y 1.08 0.97, 1.21 0.16 558/4,852 526/4,913 0 Random ModerateCVD mortality 0 2 RCTs 5y 1.24 0.27, 5.65 0.78 22/1,462 25/1,469 53 Random Very lowTotal CVD risk 0 3 RCTs 2-5y 1.43 0.79, 2.59 0.23 207/1,663 157/1,665 80 Random Very low

MI risk 0 4 RCTs 2-5y 1.69 0.94, 3.04 0.08 136/2,683 95/2,704 69 Random Low

All-cause mortality 0 2 RCTs 24-52wks 0.79 0.42, 1.51 0.48 17/456 18/306 0 Random LowCVD mortality 0 2 RCT 24-52wks 0.80 0.40, 1.58 0.51 15/454 16/306 0 Random ModerateTotal CVD risk - 1 RCT 24wks 0.49 0.28, 0.86 0.01 21/304 22/155 N/A Random Moderate

All-cause mortality 0 10 RCTs 1-11.2y 0.95 0.90, 1.01 0.12 1,771/11,436 1,862/11,433 0 Random ModerateCVD mortality 0 3 RCTs 3-11.2y 0.94 0.83, 1.06 0.30 476/8,671 508/8,680 0 Random ModerateTotal CVD risk 0 2 RCTs 3-11.2y 0.95 0.77, 1.17 0.61 970/8,170 971/8,179 62 Random Low

All-cause mortality 0 20 RCTs 6m-7.2y 0.95 0.89, 1.01 0.09 1,833/20,999 1,857/21,073 0 Random ModerateCVD mortality 0 1 RCTs 7y 0.92 0.77, 1.11 0.38 226/18,176 244/18,106 N/A Random LowTotal CVD risk 0 2 RCTs 1-7.2y 1.03 0.94, 1.12 0.58 854/7813 818/7,681 0 Random Low

Calcium and Vitamin D

Iron

Calcium

Multivitamins

Vitamin B6

73

Vitamin and mineral supplements and CVD & all-cause mortality

TOTAL: 1496 PAPERS 284 COCHRANE 716 MEDLINE 402 PUBMED 94 Manual Searches

1383 Excluded 479 Duplicates 2 Foreign Language 96 Less than 6 months 158 No outcome of interest 100 No supplement of interest 8 Non-human 59 Observational studies 9 SRMA of cohorts 181 Non-supplemental 74 Protocol 8 Review 209 Wrong population

113 PAPERS 55 SRMA (952 RCTs, 2 RCTs/cohorts)

54 single RCTs 4 single RCT/cohorts

833 Excluded (Duplicates, not the nutrient of interest, no control arm, foreign language, no supplement use, less than 6 months)

Studies identified through systematic search.

All individual RCT and cohort studies.

179 INCLUDED RCTs Outcome Vit D Vit A β-carotene Antioxidants Vit E Vit C Selenium B-complex Folic acid Total CVD 6 0 3 7 10 2 2 9 5 Total CHD 3 0 2 1 2 1 1 5 2 MI 12 0 3 6 10 2 2 13 6 Stroke 11 0 3 7 11 2 2 12 7 CVD mortality 2 0 4 7 11 2 3 5 5 CHD mortality 2 0 1 2 2 0 0 3 0 MI mortality 4 0 1 3 8 2 2 2 2 Stroke mortality 2 0 2 5 6 2 1 2 2 All-cause mortality

43 1 6 21 32 4 4 16 10

Outcome Vit B3 Vit B6 Calcium Iron Magnesium Potassium Zinc Multivitamins Ca & Vit D Total CVD 4 1 3 1 0 0 0 2 2 Total CHD 2 0 2 0 0 0 0 0 2 MI 5 2 4 1 0 0 0 3 5 Stroke 5 2 3 0 0 0 0 2 7 CVD mortality 3 0 2 2 0 0 0 3 1 CHD mortality 3 0 0 0 0 0 0 0 2 MI mortality 0 1 1 0 0 0 0 1 1 Stroke mortality 0 0 1 0 0 0 0 2 2 All-cause mortality

4 2 6 2 0 0 0 10 20

Supplementary Figure 1. Consort statement

74

Ran

dom

seq

uenc

e ge

nera

tion

(sel

ectio

n bi

as)

Albert et al., 2008 - WAFACS ?

Aloia et al., 1988 ?

Aloia et al., 2005 +

Alvarez et al., 2012 ?

Anker et al., 2009 - FAIR-HF +

AREDS Research Group 2001 +

Armitage et al., 2010 - SEARCH +

Avenell et al., 2005 - MAVIS +

Baeksgaard et al., 1998 ?

Bairati et al., 2006 +

Baker et al., 2002 ?

Baron et al., 1999 - CPPS +

Bjorkman et al., 2008 ?

Boaz et al., 2000 - SPACE +

Boden et al., 2011 - AIM HIGH ?

Bogden et al., 1994 +

Bolland et al., 2008 +

Bolton-Smith et al., 2007 +

Bonaa et al., 2006 - NORVIT ?

Brazier et al., 2005 ?

Brohult et al., 1973 ?

Brown et al., 2001 - HATS ?

CDPRG 1975 ?

Chailurkit et al., 2010 ?

Chapuy et al., 1992 ?

Chapuy et al., 2002 - Decalyos II ?

Cherniack et al., 2011 ?

Chylack et al.., 2002 - REACT +

CLIPS Group 2007 ?

Coburn et al., 2004 ?

Cole et al., 2007 - AFPPS +

Collins et al., 2003 +

Cook et al., 2007 - WACS +

Cooper et al., 2003 ?

Corless et al., 1985 +

Correa et al., 2000 +

Coyne et al., 2006 +

CTNS 2008 ?

Daly et al., 2006 +

Dawson-Hughes et al., 1997 ?

de Gaetano et al., 2001 - PPP +

de la Maza et al., 1995 +

Delanaye et al., 2013 ?

Desnuelle et al., 2001 - ALSRT ?

de Waart et al., 2001 ?

de Zeeuw et al., 2010 - VITAL +

Dukas et al., 2004 ?

Durga et al., 2007 - FACIT ?

Ebbing et al., 2008 - WENBIT +

Flicker et al., 2005 +

Frazao et al., 2000 ?

Galan et al., 2010 - SU.FOL.OM3 +

Gallagher et al., 1990 +

Gallagher et al., 2001 - STOP IT –

Gallagher et al., 2012 - VIDOS +

Gillilan et al., 1997 ?

Girodon et al., 1997 ?

Girodon et al., 1999 - MIN.VIT.AOX ?

GISSI-Prevenzione Investigators 1999 +

Glendenning et al., 2012 +

Graat et al., 2002 +

Grady et al., 1991 ?

Graf et al., 2005 ?

Grant et al., 2005 - RECORD +

Greenberg et al., 1996 - SCPS +

Green et al., 1999 - NSCPT +

Grimnes et al. 2011 ?

Gupta et al., 2016 +

Guyton et al., 2008 ?

Hamdy et al., 1995 ?

Harwood et al., 2004 - NoNOF +

Heinz et al., 2010 ?

Hennekens et al., 1996 - PHS +

Hercberg et al., 2010 - SU.VI.MAX ?

Hewitt et al., 2013 +

Hodis et al., 2002 - VEAPS +

Hodis et al., 2009 - BVAIT +

Hoffman et al., 1999 ?

House et al., 2010- DIVINe +

HPS2-THRIVE Collaborative Group 2014 +

HPS Collaborative Group 2002 +

Huo et al., 2015 - CSPPT +

Imasa et al., 2009 ?

Inkovaara et al., 1983 –

Jacobson et al., 2000 ?

Jamison et al., 2007 - HOST +

Janssen et al.,2010 ?

Komulainen et al., 1999 - OSTPRE ?

Korpela et al., 1989 ?

Krieg et al., 1999 ?

Lacroix et al., 2009 - WHI +

Lamas et al., 2013 - TACT +

Lange et al., 2004 ?

Lappe et al., 2007 +


Larsen et al., 2004 +

Law et al., 2006 +

Lee et al., 1999 - WHS +


Lehouck et al., 2012 +

Leppälä et al., 2000 - ATBC ?

Lewis et al., 2011 - CAIFOS +

Liem et al., 2003 +

Liem et al., 2004 ?

Li et al., 1993 - NIT 2 ?

Limburg et al., 2005 +

Lippman et al., 2009 - SELECT ?

Lips et al., 1996 +

Liu et al., 2007 +

Logan et al., 2008 - UKCAP +

Lonn et al., 2005 - HOPE & HOPE TOO ?

Lonn et al., 2006 - HOPE TOO ?

Lyons et al., 2007 +

Ma et al., 2012 - SIT ?

Magliano et al., 2006 - MAVET ?

Manning et al., 2013 +

Manuel-Y-Kennoy et al., 2004 - DATOR ?

Marshall et al., 2011 ?

McKeown-Eyssen et al., 1988 ?

McNeil et al., 2004 - VECAT ?

Meier et al., 2004 ?

Meydani et al., 2004 +

Meyer et al., 2002 –

Milman et al., 2008 - ICARE +

Moon et al., 1997 - SKICAP AK ?

Mooney et al., 2005 +

Omenn et al., 1996 - CARET ?

Ooms et al., 1995 ?

Ott et al., 1989 ?

Petersen et al., 2005 - ADCS 2 +

Pike et al., 1995 +

Plummer et al., 2007 +

Ponikowski et al. 2015 - CONFIRM-HF +

Porthouse et al., 2005 +

Potena et al., 2005 +

Prentice et al., 2013 - WHI CaD +

Prince et al., 2008 +

Punthakee et al., 2012 - TIDE +

Reid et al., 2008 +

Righetti et al., 2003 ?

Salonen et al., 2000 - ASAP ?

Salovaara et al., 2010 - OSTRE-FPS +

Sanders et al., 2010 - Vital D +

Sang et al., 2009 ?

Sano et al., 1997 - ADCS 1 ?

Sanyal et al., 2010 - PIVENS ?

Saposnik et al., 2009 - HOPE 2 +

Sato et al., 1997 ?

Sato et al., 1999 +

Sato et al., 2005 +

Schleitoff et al., 2006 +

Schnyder et al., 2002 - The Swiss Heart Study ?

Sesso et al., 2008 - PHS II +


Shoulson et al., 1998 - DATATOP ?

Smith et al., 2007 ?

Steiner et al., 1995 ?

Stephens et al., 1996 - CHAOS +

Stranges et al., 2006 - NPC ?

Taylor et al., 2004 - ARBITER 2 +

Tornwall et al., 2004 - ATBC ?

Trivedi et al., 2003 ?

Van Dijk et al., 2015 - B-PROOF +

Van Wijngaarden et al., 2014 - B-PROOF +

Vianna et al., 2007 ?

Virtamo et al., 1998 - ATBC ?


VITATOPS Trial Study Group 2010 +

Wang et al., 2014 - OPERA +

Wang et al., 2014 - PHS II +

Wang et al., 2015 +

Waters et al., 2002 - WAVE +

Witham et al., 2013 +

Witham et al., 2013 - VitDISH +

Wluka et al., 2002 +

Wu et al., 2009 - NHS/HPFS +

You et al., 2001 - SIT ?

Zhu et al., 2008 - CAIFOS +

Zoungas et al., 2006 - ASFAST ?

Allo

catio

n co

ncea

lmen

t (se

lect

ion

bias

)

+

?

?

+

+

+

+

?

?

?

?

?

+

?

+

?

+

+

+

?

?

?

?

+

?

?

?

+

+

+

+

+

+

?

?

?

+

?

+

?

+

?

?

?

?

+

+

+

+

+

+

+

+

+

+

?

?

?

+

+

+

?

?

+

?

+

+

–

+

?

+

+

?

?

?

?

?

?

+

+

+

+

+

+

?

+

+

?

?

?

?

+

?

–

–

–

?

?

?

+

?

+

?

?

?

+

+

+

+

+

+

+

+

?

+

+

?

?

?

?

?

+

?

+

?

?

+

+

?

+

?

+

+

+

+

?

+

+

+

+

+

?

+

+

?

+

+

?

?

+

?

?

?

?

?

+

?

?

+

+

?

?

+

+

?

?

?

+

?

?

+

+

?

+

?

+

+

+

?

Blin

ding

of p

artic

ipan

ts a

nd p

erso

nnel

(per

form

ance

bia

s)+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

–

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Inco

mpl

ete

outc

ome

data

(attr

ition

bia

s)

+

+

+

+

+

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

?

+

+

?

+

?

+

+

+

+

+

?

+

+

+

–

+

+

+

+

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

+

?

+

+

+

+

+

?

+

?

?

+

+

+

+

?

?

+

+

+

+

+

?

+

?

+

?

?

?

–

+

?

+

?

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

+

?

+

+

?

?

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

?

+

?

+

?

+

?

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

+

+

+

Sel

ectiv

e re

porti

ng (r

epor

ting

bias

)

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Supplementary Figure 2. Risk of bias summary for supplements and CVD and total mortality. Review authors' judgments about each risk of bias item for each included study.

75

Ran

dom

seq

uenc

e ge

nera

tion

(sel

ectio

n bi

as)























CDPRG 1975 ?


Chandra et al., 1992 +





CLIPS Group 2007 ?









CTNS Study Group 2008 ?

Daly et al., 2008 +
























Graf et al., 2005 ?


































Law et al., 2006 +






Liem et al., 2003 +

Liem et al., 2004 ?

Li et al., 1993 - NIT 2 ?



Lips et al., 1996 +

Liu et al., 2007 +





Ma et al., 2012 - SIT ?














Ooms et al., 1995 ?

Ott et al., 1989 ?


Pike et al., 1995 +








Reid et al., 2008 +





Sang et al., 2009 ?




Sato et al., 1997 ?

Sato et al., 1999 +

Sato et al., 2005 +





















Wang et al., 2015 +









Allo

catio

n co

ncea

lmen

t (se

lect

ion

bias

)

+

?

?

+

+

+

+

?

?

?

?

?

+

?

+

?

+

+

+

?

?

?

?

+

?

?

?

?

+

+

+

+

+

+

?

?

?

+

?

+

?

+

?

?

?

?

+

+

+

+

+

+

+

+

+

+

?

?

?

+

+

+

?

?

+

?

+

+

–

+

?

+

+

?

?

?

?

?

?

+

+

+

+

+

+

?

+

+

?

?

?

?

+

?

–

–

–

?

?

?

+

?

+

?

?

?

+

+

+

+

+

+

+

+

?

+

+

?

?

?

?

?

+

?

+

?

?

+

+

?

+

?

+

+

+

+

?

+

+

+

+

+

?

+

+

?

+

+

?

?

+

?

?

?

?

?

+

?

?

+

+

?

?

+

+

?

?

?

+

?

?

+

+

?

+

?

+

+

+

?

Blin

ding

of p

artic

ipan

ts a

nd p

erso

nnel

(per

form

ance

bia

s)

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

–

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Inco

mpl

ete

outc

ome

data

(attr

ition

bia

s)

+

+

+

+

+

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

?

+

+

?

+

+

?

+

+

+

+

+

?

+

+

+

–

+

+

+

+

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

+

?

+

+

+

+

+

?

+

?

?

+

+

+

+

?

?

+

+

+

+

+

?

+

?

+

?

?

?

–

+

?

+

?

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

+

?

+

+

?

?

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

?

+

?

+

?

+

?

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

+

+

+

Sel

ectiv

e re

porti

ng (r

epor

ting

bias

)

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Ran

dom

seq

uenc

e ge

nera

tion

(sel

ectio

n bi

as)























CDPRG 1975 ?







CLIPS Group 2007 ?









CTNS 2008 ?

Daly et al., 2006 +
























Graf et al., 2005 ?


































Law et al., 2006 +






Liem et al., 2003 +

Liem et al., 2004 ?

Li et al., 1993 - NIT 2 ?



Lips et al., 1996 +

Liu et al., 2007 +





Ma et al., 2012 - SIT ?














Ooms et al., 1995 ?

Ott et al., 1989 ?


Pike et al., 1995 +








Reid et al., 2008 +





Sang et al., 2009 ?




Sato et al., 1997 ?

Sato et al., 1999 +

Sato et al., 2005 +





















Wang et al., 2015 +









Allo

catio

n co

ncea

lmen

t (se

lect

ion

bias

)

+

?

?

+

+

+

+

?

?

?

?

?

+

?

+

?

+

+

+

?

?

?

?

+

?

?

?

?

+

+

+

+

+

+

?

?

?

+

?

+

?

+

?

?

?

?

+

+

+

+

+

+

+

+

+

+

?

?

?

+

+

+

?

?

+

?

+

+

–

+

?

+

+

?

?

?

?

?

?

+

+

+

+

+

+

?

+

+

?

?

?

?

+

?

–

–

–

?

?

?

+

?

+

?

?

?

+

+

+

+

+

+

+

+

?

+

+

?

?

?

?

?

+

?

+

?

?

+

+

?

+

?

+

+

+

+

?

+

+

+

+

+

?

+

+

?

+

+

?

?

+

?

?

?

?

?

+

?

?

+

+

?

?

+

+

?

?

?

+

?

?

+

+

?

+

?

+

+

+

?

Blin

ding

of p

artic

ipan

ts a

nd p

erso

nnel

(per

form

ance

bia

s)

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

–

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Inco

mpl

ete

outc

ome

data

(attr

ition

bia

s)

+

+

+

+

+

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

?

+

+

?

+

+

?

+

+

+

+

+

?

+

+

+

–

+

+

+

+

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

+

?

+

+

+

+

+

?

+

?

?

+

+

+

+

?

?

+

+

+

+

+

?

+

?

+

?

?

?

–

+

?

+

?

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

+

?

+

+

?

?

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

?

+

?

+

?

+

?

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

+

+

+

Sel

ectiv

e re

porti

ng (r

epor

ting

bias

)

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Supplementary Figure 2 (Continued). Risk of bias summary for supplements and CVD and total mortality. Review authors' judgments about each risk of bias item for each included study.

76

Ran

dom

seq

uenc

e ge

nera

tion

(sel

ectio

n bi

as)























CDPRG 1975 ?







CLIPS Group 2007 ?










Daly et al., 2008 +
























Graf et al., 2005 ?


































Law et al., 2006 +






Liem et al., 2003 +

Liem et al., 2004 ?

Li et al., 1993 - NIT 2 ?



Lips et al., 1996 +

Liu et al., 2007 +





Ma et al., 2012 - SIT ?














Ooms et al., 1995 ?

Ott et al., 1989 ?


Pike et al., 1995 +








Reid et al., 2008 +





Sang et al., 2009 ?




Sato et al., 1997 ?

Sato et al., 1999 +

Sato et al., 2005 +





















Wang et al., 2015 +









Allo

catio

n co

ncea

lmen

t (se

lect

ion

bias

)

+

?

?

+

+

+

+

?

?

?

?

?

+

?

+

?

+

+

+

?

?

?

?

+

?

?

?

?

+

+

+

+

+

+

?

?

?

+

?

+

?

+

?

?

?

?

+

+

+

+

+

+

+

+

+

+

?

?

?

+

+

+

?

?

+

?

+

+

–

+

?

+

+

?

?

?

?

?

?

+

+

+

+

+

+

?

+

+

?

?

?

?

+

?

–

–

–

?

?

?

+

?

+

?

?

?

+

+

+

+

+

+

+

+

?

+

+

?

?

?

?

?

+

?

+

?

?

+

+

?

+

?

+

+

+

+

?

+

+

+

+

+

?

+

+

?

+

+

?

?

+

?

?

?

?

?

+

?

?

+

+

?

?

+

+

?

?

?

+

?

?

+

+

?

+

?

+

+

+

?

Blin

ding

of p

artic

ipan

ts a

nd p

erso

nnel

(per

form

ance

bia

s)

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

–

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Inco

mpl

ete

outc

ome

data

(attr

ition

bia

s)

+

+

+

+

+

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

?

+

+

?

+

+

?

+

+

+

+

+

?

+

+

+

–

+

+

+

+

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

+

?

+

+

+

+

+

?

+

?

?

+

+

+

+

?

?

+

+

+

+

+

?

+

?

+

?

?

?

–

+

?

+

?

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

+

?

+

+

?

?

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

?

+

?

+

?

+

?

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

+

+

+

Sel

ectiv

e re

porti

ng (r

epor

ting

bias

)

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Ran

dom

seq

uenc

e ge

nera

tion

(sel

ectio

n bi

as)























CDPRG 1975 ?







CLIPS Group 2007 ?










Daly et al., 2008 +
























Graf et al., 2005 ?


































Law et al., 2006 +






Liem et al., 2003 +

Liem et al., 2004 ?

Li et al., 1993 - NIT 2 ?



Lips et al., 1996 +

Liu et al., 2007 +





Ma et al., 2012 - SIT ?














Ooms et al., 1995 ?

Ott et al., 1989 ?


Pike et al., 1995 +








Reid et al., 2008 +





Sang et al., 2009 ?




Sato et al., 1997 ?

Sato et al., 1999 +

Sato et al., 2005 +





















Wang et al., 2015 +









Allo

catio

n co

ncea

lmen

t (se

lect

ion

bias

)

+

?

?

+

+

+

+

?

?

?

?

?

+

?

+

?

+

+

+

?

?

?

?

+

?

?

?

?

+

+

+

+

+

+

?

?

?

+

?

+

?

+

?

?

?

?

+

+

+

+

+

+

+

+

+

+

?

?

?

+

+

+

?

?

+

?

+

+

–

+

?

+

+

?

?

?

?

?

?

+

+

+

+

+

+

?

+

+

?

?

?

?

+

?

–

–

–

?

?

?

+

?

+

?

?

?

+

+

+

+

+

+

+

+

?

+

+

?

?

?

?

?

+

?

+

?

?

+

+

?

+

?

+

+

+

+

?

+

+

+

+

+

?

+

+

?

+

+

?

?

+

?

?

?

?

?

+

?

?

+

+

?

?

+

+

?

?

?

+

?

?

+

+

?

+

?

+

+

+

?

Blin

ding

of p

artic

ipan

ts a

nd p

erso

nnel

(per

form

ance

bia

s)

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

–

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Inco

mpl

ete

outc

ome

data

(attr

ition

bia

s)

+

+

+

+

+

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

?

+

+

?

+

+

?

+

+

+

+

+

?

+

+

+

–

+

+

+

+

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

+

?

+

+

+

+

+

?

+

?

?

+

+

+

+

?

?

+

+

+

+

+

?

+

?

+

?

?

?

–

+

?

+

?

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

+

?

+

+

?

?

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

?

+

?

+

?

+

?

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

+

+

+

Sel

ectiv

e re

porti

ng (r

epor

ting

bias

)

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Ran

dom

seq

uenc

e ge

nera

tion

(sel

ectio

n bi

as)























CDPRG 1975 ?







CLIPS Group 2007 ?










Daly et al., 2008 +
























Graf et al., 2005 ?


































Law et al., 2006 +






Liem et al., 2003 +

Liem et al., 2004 ?

Li et al., 1993 - NIT 2 ?



Lips et al., 1996 +

Liu et al., 2007 +





Ma et al., 2012 - SIT ?














Ooms et al., 1995 ?

Ott et al., 1989 ?


Pike et al., 1995 +








Reid et al., 2008 +





Sang et al., 2009 ?




Sato et al., 1997 ?

Sato et al., 1999 +

Sato et al., 2005 +





















Wang et al., 2015 +









Allo

catio

n co

ncea

lmen

t (se

lect

ion

bias

)

+

?

?

+

+

+

+

?

?

?

?

?

+

?

+

?

+

+

+

?

?

?

?

+

?

?

?

?

+

+

+

+

+

+

?

?

?

+

?

+

?

+

?

?

?

?

+

+

+

+

+

+

+

+

+

+

?

?

?

+

+

+

?

?

+

?

+

+

–

+

?

+

+

?

?

?

?

?

?

+

+

+

+

+

+

?

+

+

?

?

?

?

+

?

–

–

–

?

?

?

+

?

+

?

?

?

+

+

+

+

+

+

+

+

?

+

+

?

?

?

?

?

+

?

+

?

?

+

+

?

+

?

+

+

+

+

?

+

+

+

+

+

?

+

+

?

+

+

?

?

+

?

?

?

?

?

+

?

?

+

+

?

?

+

+

?

?

?

+

?

?

+

+

?

+

?

+

+

+

?

Blin

ding

of p

artic

ipan

ts a

nd p

erso

nnel

(per

form

ance

bia

s)

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

–

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Inco

mpl

ete

outc

ome

data

(attr

ition

bia

s)

+

+

+

+

+

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

?

+

+

?

+

+

?

+

+

+

+

+

?

+

+

+

–

+

+

+

+

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

+

?

+

+

+

+

+

?

+

?

?

+

+

+

+

?

?

+

+

+

+

+

?

+

?

+

?

?

?

–

+

?

+

?

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

+

?

+

+

?

?

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

?

+

?

+

?

+

?

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

+

+

+

Sel

ectiv

e re

porti

ng (r

epor

ting

bias

)

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Ran

dom

seq

uenc

e ge

nera

tion

(sel

ectio

n bi

as)























CDPRG 1975 ?







CLIPS Group 2007 ?










Daly et al., 2008 +
























Graf et al., 2005 ?


































Law et al., 2006 +






Liem et al., 2003 +

Liem et al., 2004 ?

Li et al., 1993 - NIT 2 ?



Lips et al., 1996 +

Liu et al., 2007 +





Ma et al., 2012 - SIT ?














Ooms et al., 1995 ?

Ott et al., 1989 ?


Pike et al., 1995 +








Reid et al., 2008 +





Sang et al., 2009 ?




Sato et al., 1997 ?

Sato et al., 1999 +

Sato et al., 2005 +





















Wang et al., 2015 +









Allo

catio

n co

ncea

lmen

t (se

lect

ion

bias

)

+

?

?

+

+

+

+

?

?

?

?

?

+

?

+

?

+

+

+

?

?

?

?

+

?

?

?

?

+

+

+

+

+

+

?

?

?

+

?

+

?

+

?

?

?

?

+

+

+

+

+

+

+

+

+

+

?

?

?

+

+

+

?

?

+

?

+

+

–

+

?

+

+

?

?

?

?

?

?

+

+

+

+

+

+

?

+

+

?

?

?

?

+

?

–

–

–

?

?

?

+

?

+

?

?

?

+

+

+

+

+

+

+

+

?

+

+

?

?

?

?

?

+

?

+

?

?

+

+

?

+

?

+

+

+

+

?

+

+

+

+

+

?

+

+

?

+

+

?

?

+

?

?

?

?

?

+

?

?

+

+

?

?

+

+

?

?

?

+

?

?

+

+

?

+

?

+

+

+

?

Blin

ding

of p

artic

ipan

ts a

nd p

erso

nnel

(per

form

ance

bia

s)

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

–

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Inco

mpl

ete

outc

ome

data

(attr

ition

bia

s)

+

+

+

+

+

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

?

+

+

?

+

+

?

+

+

+

+

+

?

+

+

+

–

+

+

+

+

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

+

?

+

+

+

+

+

?

+

?

?

+

+

+

+

?

?

+

+

+

+

+

?

+

?

+

?

?

?

–

+

?

+

?

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

+

?

+

+

?

?

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

?

+

?

+

?

+

?

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

+

+

+

Sel

ectiv

e re

porti

ng (r

epor

ting

bias

)

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Ran

dom

seq

uenc

e ge

nera

tion

(sel

ectio

n bi

as)























CDPRG 1975 ?







CLIPS Group 2007 ?










Daly et al., 2008 +
























Graf et al., 2005 ?


































Law et al., 2006 +






Liem et al., 2003 +

Liem et al., 2004 ?

Li et al., 1993 - NIT 2 ?



Lips et al., 1996 +

Liu et al., 2007 +





Ma et al., 2012 - SIT ?














Ooms et al., 1995 ?

Ott et al., 1989 ?


Pike et al., 1995 +








Reid et al., 2008 +





Sang et al., 2009 ?




Sato et al., 1997 ?

Sato et al., 1999 +

Sato et al., 2005 +





















Wang et al., 2015 +









Allo

catio

n co

ncea

lmen

t (se

lect

ion

bias

)

+

?

?

+

+

+

+

?

?

?

?

?

+

?

+

?

+

+

+

?

?

?

?

+

?

?

?

?

+

+

+

+

+

+

?

?

?

+

?

+

?

+

?

?

?

?

+

+

+

+

+

+

+

+

+

+

?

?

?

+

+

+

?

?

+

?

+

+

–

+

?

+

+

?

?

?

?

?

?

+

+

+

+

+

+

?

+

+

?

?

?

?

+

?

–

–

–

?

?

?

+

?

+

?

?

?

+

+

+

+

+

+

+

+

?

+

+

?

?

?

?

?

+

?

+

?

?

+

+

?

+

?

+

+

+

+

?

+

+

+

+

+

?

+

+

?

+

+

?

?

+

?

?

?

?

?

+

?

?

+

+

?

?

+

+

?

?

?

+

?

?

+

+

?

+

?

+

+

+

?

Blin

ding

of p

artic

ipan

ts a

nd p

erso

nnel

(per

form

ance

bia

s)

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

–

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Inco

mpl

ete

outc

ome

data

(attr

ition

bia

s)

+

+

+

+

+

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

?

+

+

?

+

+

?

+

+

+

+

+

?

+

+

+

–

+

+

+

+

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

+

?

+

+

+

+

+

?

+

?

?

+

+

+

+

?

?

+

+

+

+

+

?

+

?

+

?

?

?

–

+

?

+

?

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

+

?

+

+

?

?

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

?

+

?

+

?

+

?

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

+

+

+

Sel

ectiv

e re

porti

ng (r

epor

ting

bias

)

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Supplementary Figure 2 (Continued). Risk of bias summary for supplements and CVD and total mortality. Review authors' judgments about each risk of bias item for each included study

77

Ran

dom

seq

uenc

e ge

nera

tion

(sel

ectio

n bi

as)























CDPRG 1975 ?







CLIPS Group 2007 ?










Daly et al., 2008 +
























Graf et al., 2005 ?


































Law et al., 2006 +






Liem et al., 2003 +

Liem et al., 2004 ?

Li et al., 1993 - NIT 2 ?



Lips et al., 1996 +

Liu et al., 2007 +





Ma et al., 2012 - SIT ?














Ooms et al., 1995 ?

Ott et al., 1989 ?


Pike et al., 1995 +








Reid et al., 2008 +





Sang et al., 2009 ?




Sato et al., 1997 ?

Sato et al., 1999 +

Sato et al., 2005 +





















Wang et al., 2015 +









Allo

catio

n co

ncea

lmen

t (se

lect

ion

bias

)

+

?

?

+

+

+

+

?

?

?

?

?

+

?

+

?

+

+

+

?

?

?

?

+

?

?

?

?

+

+

+

+

+

+

?

?

?

+

?

+

?

+

?

?

?

?

+

+

+

+

+

+

+

+

+

+

?

?

?

+

+

+

?

?

+

?

+

+

–

+

?

+

+

?

?

?

?

?

?

+

+

+

+

+

+

?

+

+

?

?

?

?

+

?

–

–

–

?

?

?

+

?

+

?

?

?

+

+

+

+

+

+

+

+

?

+

+

?

?

?

?

?

+

?

+

?

?

+

+

?

+

?

+

+

+

+

?

+

+

+

+

+

?

+

+

?

+

+

?

?

+

?

?

?

?

?

+

?

?

+

+

?

?

+

+

?

?

?

+

?

?

+

+

?

+

?

+

+

+

?

Blin

ding

of p

artic

ipan

ts a

nd p

erso

nnel

(per

form

ance

bia

s)

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

–

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Inco

mpl

ete

outc

ome

data

(attr

ition

bia

s)

+

+

+

+

+

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

?

+

+

?

+

+

?

+

+

+

+

+

?

+

+

+

–

+

+

+

+

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

+

?

+

+

+

+

+

?

+

?

?

+

+

+

+

?

?

+

+

+

+

+

?

+

?

+

?

?

?

–

+

?

+

?

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

+

?

+

+

?

?

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

?

+

?

+

?

+

?

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

+

+

+

Sel

ectiv

e re

porti

ng (r

epor

ting

bias

)

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Ran

dom

seq

uenc

e ge

nera

tion

(sel

ectio

n bi

as)























CDPRG 1975 ?







CLIPS Group 2007 ?










Daly et al., 2008 +
























Graf et al., 2005 ?


































Law et al., 2006 +






Liem et al., 2003 +

Liem et al., 2004 ?

Li et al., 1993 - NIT 2 ?



Lips et al., 1996 +

Liu et al., 2007 +





Ma et al., 2012 - SIT ?














Ooms et al., 1995 ?

Ott et al., 1989 ?


Pike et al., 1995 +








Reid et al., 2008 +





Sang et al., 2009 ?




Sato et al., 1997 ?

Sato et al., 1999 +

Sato et al., 2005 +





















Wang et al., 2015 +









Allo

catio

n co

ncea

lmen

t (se

lect

ion

bias

)

+

?

?

+

+

+

+

?

?

?

?

?

+

?

+

?

+

+

+

?

?

?

?

+

?

?

?

?

+

+

+

+

+

+

?

?

?

+

?

+

?

+

?

?

?

?

+

+

+

+

+

+

+

+

+

+

?

?

?

+

+

+

?

?

+

?

+

+

–

+

?

+

+

?

?

?

?

?

?

+

+

+

+

+

+

?

+

+

?

?

?

?

+

?

–

–

–

?

?

?

+

?

+

?

?

?

+

+

+

+

+

+

+

+

?

+

+

?

?

?

?

?

+

?

+

?

?

+

+

?

+

?

+

+

+

+

?

+

+

+

+

+

?

+

+

?

+

+

?

?

+

?

?

?

?

?

+

?

?

+

+

?

?

+

+

?

?

?

+

?

?

+

+

?

+

?

+

+

+

?

Blin

ding

of p

artic

ipan

ts a

nd p

erso

nnel

(per

form

ance

bia

s)

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

–

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Inco

mpl

ete

outc

ome

data

(attr

ition

bia

s)

+

+

+

+

+

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

?

+

+

?

+

+

?

+

+

+

+

+

?

+

+

+

–

+

+

+

+

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

+

?

+

+

+

+

+

?

+

?

?

+

+

+

+

?

?

+

+

+

+

+

?

+

?

+

?

?

?

–

+

?

+

?

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

+

?

+

+

?

?

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

?

+

+

?

+

?

+

?

+

?

+

+

+

+

?

+

?

+

+

+

?

+

+

+

+

?

?

+

+

?

?

?

+

+

?

?

+

+

+

Sel

ectiv

e re

porti

ng (r

epor

ting

bias

)

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

?

?

+

+

+

+

+

+

+

+

+

+

+

?

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Random Effects

Comparison RCTs N Events RR (95% CIs) RR (95% CIs) I2 p -value

CVD

Total CVD 6 6,546 1,037 0.95 [0.86, 1.05] 0% 0.31Total CHD 3 434 6 0.97 [0.22, 4.22] 0% 0.97MI 12 11,081 643 0.95 [0.83, 1.10] 0% 0.52Stroke 11 11,173 479 1.12 [0.94, 1.34] 0% 0.20Total CVD mortality 2 3,907 219 0.86 [0.66, 1.10] 0% 0.23Total CHD mortality 2 225 4 0.41 [0.06, 2.72] 0% 0.35MI mortality 4 2,873 94 0.85 [0.57, 1.26] 0% 0.41Stroke mortality 2 2,773 60 1.13 [0.68, 1.87] 0% 0.63

ALL-CAUSE MORTALITY All-cause mortality 43 37,550 5,876 0.99 [0.95, 1.03] 0% 0.58

Favours vitamin D/ Favours control/ Negative association Positive association

Pooled Effect Estimates

0.0 0.5 1.0 1.5 2.0

Supplementary Figure 3. Summary of the pooled effect estimates of RCTs assessing the relationship between vitamin D supplementation and CVD and all-cause mortality risk. RR, risk ratio; CIs, confidence intervals; RTCs, randomized clinical trials; CVD, cardiovascular disease; CHD, coronary heart disease; MI, myocardial infarction; N/A, not applicable. The diamonds represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% CIs, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsGallagher et al., 2001 - STOP IT 4 123 3 123 0.4% 1.33 [0.30, 5.83]Trivedi et al., 2003 477 1,345 503 1,341 96.6% 0.95 [0.86, 1.04]Sanders et al., 2010 - Vital D 17 1,131 13 1,127 1.9% 1.30 [0.64, 2.67]Punthakee et al., 2012 - TIDE 2 607 3 614 0.3% 0.67 [0.11, 4.02]Witham et al., 2013 5 39 4 36 0.6% 1.15 [0.34, 3.96]Wang et al., 2014 - OPERA 0 30 6 30 0.1% 0.08 [0.00, 1.31]

Total (95% CI) 3,275 3,271 100% 0.95 [0.86, 1.05]Total events 505 532Heterogeneity: Tau² = 0.00; Chi² = 4.23, df = 5 (P = 0.52); I² = 0%Test for overall effect: Z = 1.02 (P = 0.31) 0.5 0.7 1 1.5 2.0

Favours vitamin D Favours control

Vitamin D ControlWeight

Risk RatioM-H, Random, 95% CI

Risk RatioM-H, Random, 95% CI in total CVD risk

Study or SubgroupGallagher et al., 2001 - STOP ITTrivedi et al., 2003Sanders et al., 2010 - Vital DPunthakee et al., 2012 - TIDEWitham et al., 2013Wang et al., 2014 - OPERA

Total (95% CI)Total eventsHeterogeneity: Tau² = 0.00; Chi² = 4.23, df = 5 (P = 0.52); I² = 0%Test for overall effect: Z = 1.02 (P = 0.31)

Events4

47717

250

505

Total123

13451131

6073930

3275

Events3

50313

346

532

Total123

13411127

6143630

3271

Weight0.4%

96.6%1.9%0.3%0.6%0.1%

100.0%

M-H, Random, 95% CI1.33 [0.30, 5.83]0.95 [0.86, 1.04]1.30 [0.64, 2.67]0.67 [0.11, 4.02]1.15 [0.34, 3.96]0.08 [0.00, 1.31]

0.95 [0.86, 1.05]

Year200120032010201220132014

Vitamin D Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.5 0.7 1 1.5 2Favours vitamin D Favours control

Supplementary Figure 4. Forest plot of vitamin D supplementation and total CVD risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

78

Subgroup and Study, Year Events Total Events TotalRCTsInkovaara et al., 1983 0 45 1 42 21.4% 0.31 [0.01, 7.44]de Zeeuw et al., 2010 - VITAL 1 95 0 93 21.2% 2.94 [0.12, 71.20]Witham et al., 2013 - VitDISH 2 80 2 79 57.5% 0.99 [0.14, 6.84]

Total (95% CI) 220 214 100% 0.97 [0.22, 4.22]Total events 3 3Heterogeneity: Tau² = 0.00; Chi² = 0.96, df = 2 (P = 0.62); I² = 0%Test for overall effect: Z = 0.04 (P = 0.97) 0.001 0.1 1 10 1000




Risk RatioM-H, Random, 95% CI in total CHD risk

Study or SubgroupInkovaara et al., 1983de Zeeuw et al., 2010 - VITALWitham et al., 2013 - VitDISH


Events012

3

Total459580

220

Events102

3

Total429379

214

Weight21.4%21.2%57.5%

100.0%

M-H, Random, 95% CI0.31 [0.01, 7.44]

2.94 [0.12, 71.20]0.99 [0.14, 6.84]

0.97 [0.22, 4.22]

Year198320102013


0.001 0.1 1 10 1000Favours Vitamin D Favours Control

Supplementary Figure 5. Forest plot of vitamin D supplementation and total CHD risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsAloia et al., 1988 1 12 1 15 0.3% 1.25 [0.09, 17.98]Ott et al., 1989 1 43 0 43 0.2% 3.00 [0.13, 71.65]Komulainen et al., 1999 - OSTPRE 1 112 0 115 0.2% 3.08 [0.13, 74.81]Trivedi et al., 2003 224 1,345 233 1,341 74.2% 0.96 [0.81, 1.13]Coburn et al., 2004 0 27 2 28 0.2% 0.21 [0.01, 4.13]Grant et al., 2005 - RECORD* 78 2649 84 2643 22.5% 0.93 [0.68, 1.25]Lappe et al., 2007† 3 445 2 446 0.7% 1.50 [0.25, 8.95]Prince et al., 2008‡ 2 151 3 151 0.7% 0.67 [0.11, 3.93]de Zeeuw et al., 2010 - VITAL 2 95 0 93 0.2% 4.90 [0.24, 100.62]Cherniack et al., 2011 1 23 1 23 0.3% 1.00 [0.07, 15.04]Punthakee et al., 2012 - TIDE 1 607 1 614 0.3% 1.01 [0.06, 16.14]Wang et al., 2014 - OPERA 0 30 2 30 0.2% 0.20 [0.01, 4.00]

Total (95% CI) 5,539 5,542 100% 0.95 [0.83, 1.10]Total events 314 329Heterogeneity: Tau² = 0.00; Chi² = 4.68, df = 11 (P = 0.95); I² = 0%Test for overall effect: Z = 0.64 (P = 0.52) 0.1 0.2 0.5 1 2 5 10




Risk RatioM-H, Random, 95% CI in MI risk

Study or SubgroupAloia et al., 1988Ott et al., 1989Komulainen et al., 1999 - OSTPRETrivedi et al., 2003Coburn et al., 2004Grant et al., 2005 - RECORDLappe et al., 2007Prince et al., 2008de Zeeuw et al., 2010 - VITALCherniack et al., 2011Punthakee et al., 2012 - TIDEWang et al., 2014 - OPERA


Events111

2240

78322110

314

Total1243

1121345

272649

445151

9523

60730

5539

Events100

2332

84230112

329

Total1543

1151341

282643

446151

9323

61430

5542

Weight0.3%0.2%0.2%

74.2%0.2%

22.5%0.7%0.7%0.2%0.3%0.3%0.2%

100.0%

M-H, Random, 95% CI1.25 [0.09, 17.98]3.00 [0.13, 71.65]3.08 [0.13, 74.81]

0.96 [0.81, 1.13]0.21 [0.01, 4.13]0.93 [0.68, 1.25]1.50 [0.25, 8.95]0.67 [0.11, 3.93]

4.90 [0.24, 100.62]1.00 [0.07, 15.04]1.01 [0.06, 16.14]

0.20 [0.01, 4.00]

0.95 [0.83, 1.10]

Year198819891999200320042005200720082010201120122014


0.1 0.2 0.5 1 2 5 10Favours vitamin D Favours control

Supplementary Figure 6. Forest plot of vitamin D supplementation and MI risk. M-H, Manthel-Haenszel, MI, myocardial infarction. *Lappe et al., 2007 - data taken from meta-analysis; Bolland et al., 2014; †Prince et al., 2008 - reported number of participants with at least one event. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

79

Subgroup and Study, Year Events Total Events TotalRCTsInkovaara et al., 1983 12 45 5 42 3.3% 2.24 [0.86, 5.82]Gallagher et al., 2001 - STOP IT 4 123 3 123 1.4% 1.33 [0.30, 5.83]Trivedi et al., 2003 105 1,345 101 1,341 44.2% 1.04 [0.80, 1.35]Grant et al., 2005 - RECORD* 118 2,649 104 2,643 45.7% 1.13 [0.87, 1.47]Lappe et al., 2007† 6 446 5 445 2.2% 1.20 [0.37, 3.89]Prince et al., 2008‡ 3 151 3 151 1.2% 1.00 [0.21, 4.88]de Zeeuw et al., 2010 - VITAL 1 95 0 93 0.3% 2.94 [0.12, 71.20]Gallagher et al., 2012 - VIDOS 1 20 0 21 0.3% 3.14 [0.14, 72.92]Punthakee et al., 2012 - TIDE 1 607 1 614 0.4% 1.01 [0.06, 16.14]Witham et al., 2013 - VitDISH 3 80 1 79 0.6% 2.96 [0.31, 27.88]Wang et al., 2014 - OPERA 0 30 2 30 0.3% 0.20 [0.01, 4.00]





Risk RatioM-H, Random, 95% CI in stroke risk

Study or SubgroupInkovaara et al., 1983Gallagher et al., 2001 - STOP ITTrivedi et al., 2003Grant et al., 2005 - RECORDLappe et al., 2007Prince et al., 2008de Zeeuw et al., 2010 - VITALGallagher et al., 2012 - VIDOSPunthakee et al., 2012 - TIDEWitham et al., 2013 - VitDISHWang et al., 2014 - OPERA


Events12

4105118

6311130

254

Total45

12313452649

446151

9520

6078030

5591

Events53

101104

5300112

225

Total42

12313412643

445151

9321

6147930

5582

Weight3.3%1.4%

44.2%45.7%

2.2%1.2%0.3%0.3%0.4%0.6%0.3%

100.0%

M-H, Random, 95% CI2.24 [0.86, 5.82]1.33 [0.30, 5.83]1.04 [0.80, 1.35]1.13 [0.87, 1.47]1.20 [0.37, 3.89]1.00 [0.21, 4.88]

2.94 [0.12, 71.20]3.14 [0.14, 72.92]1.01 [0.06, 16.14]2.96 [0.31, 27.88]

0.20 [0.01, 4.00]

1.12 [0.94, 1.34]

Year19832001200320052007200820102012201220132014


0.1 0.2 0.5 1 2 5 10Favours vitamin D Favours control

Supplementary Figure 7. Forest plot of vitamin D supplementation and stroke risk. M-H, Manthel-Haenszel. *Grant et al., 2005 - data taken from meta-analysis Bolland et al., 2014 †Lappe et al., 2007 - data taken from meta-analysis Bolland et al., 2014 ‡Prince et al., 2008 - reported number of participants with at least one event. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsTrivedi et al., 2003 101 1,345 117 1,341 99.4% 0.86 [0.67, 1.11]Punthakee et al., 2012 - TIDE 0 607 1 614 0.6% 0.34 [0.01, 8.26]





Risk RatioM-H, Random, 95% CI in CVD mortality risk

Study or SubgroupTrivedi et al., 2003Punthakee et al., 2012 - TIDE


Events101

0

101

Total1345

607

1952

Events117

1

118

Total1341

614

1955

Weight99.4%

0.6%

100.0%

M-H, Random, 95% CI0.86 [0.67, 1.11]0.34 [0.01, 8.26]

0.86 [0.66, 1.10]

Year20032012


0.1 0.2 0.5 1 2 5 10Favours Vitamin D Favours Control

Supplementary Figure 8. Forest plot of vitamin D supplementation and CVD mortality risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

80

Subgroup and Study, Year Events Total Events TotalRCTsInkovaara et al., 1983 0 45 1 42 35.9% 0.31 [0.01, 7.44]Frazao et al., 2000 1 71 2 67 64.1% 0.47 [0.04, 5.08]





Risk RatioM-H, Random, 95% CI in CHD mortality risk

Study or SubgroupInkovaara et al., 1983Frazao et al., 2000


Events01

1

Total4571

116

Events12

3

Total4267

109

Weight35.9%64.1%

100.0%

M-H, Random, 95% CI0.31 [0.01, 7.44]0.47 [0.04, 5.08]

0.41 [0.06, 2.72]

Year19832000



Supplementary Figure 9. Forest plot of vitamin D supplementation and CHD mortality risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsOtt et al., 1989 0 43 1 43 1.6% 0.33 [0.01, 7.96]Trivedi et al., 2003 42 1,345 49 1,341 95.3% 0.85 [0.57, 1.28]Coburn et al., 2004 0 27 1 28 1.6% 0.35 [0.01, 8.12]Cherniack et al., 2011 1 23 0 23 1.6% 3.00 [0.13, 70.02]

Total (95% CI) 1,438 1,435 100% 0.85 [0.57, 1.26]Total events 43 51Heterogeneity: Tau² = 0.00; Chi² = 1.26, df = 3 (P = 0.74); I² = 0%Test for overall effect: Z = 0.82 (P = 0.41) 0.02 0.1 1 10 50




Risk RatioM-H, Random, 95% CI in MI mortality risk

Study or SubgroupOtt et al., 1989Trivedi et al., 2003Coburn et al., 2004Cherniack et al., 2011


Events0

4201

43

Total43

13452723

1438

Events1

4910

51

Total43

13412823

1435

Weight1.6%

95.3%1.6%1.6%

100.0%

M-H, Random, 95% CI0.33 [0.01, 7.96]0.85 [0.57, 1.28]0.35 [0.01, 8.12]

3.00 [0.13, 70.02]

0.85 [0.57, 1.26]

Year1989200320042011



Supplementary Figure 10. Forest plot of vitamin D supplementation and MI mortality risk. M-H, Manthel-Haenszel, MI, myocardial infarction. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

81

Subgroup and Study, Year Events Total Events TotalRCTsInkovaara et al., 1983 4 45 2 42 9.4% 1.87 [0.36, 9.67]Trivedi et al., 2003 28 1,345 26 1,341 90.6% 1.07 [0.63, 1.82]





Risk RatioM-H, Random, 95% CI in stroke mortality risk

Study or SubgroupInkovaara et al., 1983Trivedi et al., 2003


Events4

28

32

Total45

1345

1390

Events2

26

28

Total42

1341

1383

Weight9.4%

90.6%

100.0%

M-H, Random, 95% CI1.87 [0.36, 9.67]1.07 [0.63, 1.82]

1.13 [0.68, 1.87]


0.1 0.2 0.5 1 2 5 10Favours Vitamin D Favours Control

Supplementary Figure 11. Forest plot of vitamin D supplementation and stroke mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

82

Subgroup and Study, Year Events Total Events TotalRCTsBrohult et al., 1973* 1 25 0 25 0.0% 3.00 [0.13, 70.30]Inkovaara et al., 1983 7 45 5 42 0.1% 1.31 [0.45, 3.80]Corless et al., 1985 8 41 8 41 0.2% 1.00 [0.42, 2.41]Ott et al., 1989 0 43 1 43 0.0% 0.33 [0.01, 7.96]Gallagher et al., 1990† 1 25 0 25 0.0% 3.00 [0.13, 70.30]Grady et al., 1991‡ 1 50 0 48 0.0% 2.88 [0.12, 69.07]Hamdy et al., 1995 4 89 1 87 0.0% 3.91 [0.45, 34.29]Ooms et al., 1995§ 11 177 21 171 0.3% 0.51 [0.25, 1.02]Lips et al., 1996 223 1,291 251 1,287 6.4% 0.89 [0.75, 1.04]Sato et al., 1997 1 45 1 39 0.0% 0.87 [0.06, 13.40]Sato et al., 1999ǁ 1 43 0 43 0.0% 3.00 [0.13, 71.65]Komulainen et al., 1999 - OSTPRE 0 112 1 115 0.0% 0.34 [0.01, 8.31]Frazao et al., 2000 1 71 2 67 0.0% 0.47 [0.04, 5.08]Gallagher et al., 2001 - STOP IT 1 123 1 123 0.0% 1.00 [0.06, 15.81]Meyer et al., 2002 169 569 163 575 5.2% 1.05 [0.87, 1.26]Trivedi et al., 2003 224 1,345 247 1,341 6.3% 0.90 [0.77, 1.07]Cooper et al., 2003¶ 0 93 1 94 0.0% 0.34 [0.01, 8.16]Harwood et al., 2004 - NONOF 7 38 5 37 0.2% 1.36 [0.47, 3.91]Dukas et al., 2004 1 192 1 186 0.0% 0.97 [0.06, 15.37]Sato et al., 2005** 1 48 2 48 0.0% 0.50 [0.05, 5.33]Grant et al., 2005 - RECORD 438 2,649 460 2,643 12.0% 0.95 [0.84, 1.07]Aloia et al., 2005†† 1 104 2 104 0.0% 0.50 [0.05, 5.43]Flicker et al., 2005 76 313 85 312 2.4% 0.89 [0.68, 1.16]Coyne et al., 2006 2 107 1 113 0.0% 2.11 [0.19, 22.96]Law et al., 2006 347 1,762 322 1,955 9.0% 1.20 [1.04, 1.37]Schleithoff et al., 2006 7 61 6 62 0.2% 1.19 [0.42, 3.33]Smith et al., 2007‡‡ 355 4,727 354 4,713 8.5% 1.00 [0.87, 1.15]Lyons et al., 2007 947 1,725 953 1,715 47.0% 0.99 [0.93, 1.05]Prince et al., 2008 0 151 1 151 0.0% 0.33 [0.01, 8.12]Bjorkman et al., 2008 10 73 9 68 0.2% 1.04 [0.45, 2.39]Janssen et al., 2010§§ 1 36 0 34 0.0% 2.84 [0.12, 67.36]Sanders et al., 2010 - Vital D 40 1,131 47 1,127 1.0% 0.85 [0.56, 1.28]de Zeeuw et al., 2010 - VITAL 1 95 0 93 0.0% 2.94 [0.12, 71.20]Cherniack et al., 2011 1 23 0 23 0.0% 3.00 [0.13, 70.02]Grimnes et al., 2011 0 51 1 53 0.0% 0.35 [0.01, 8.31]Punthakee et al., 2012 - TIDE 0 607 2 614 0.0% 0.20 [0.01, 4.21]Alvarez et al., 2012 1 24 1 24 0.0% 1.00 [0.07, 15.08]Lehouck et al., 2012 9 91 6 91 0.2% 1.50 [0.56, 4.04]Glendenning et al., 2012 2 353 0 333 0.0% 4.72 [0.23, 97.90]Witham et al., 2013 1 39 0 36 0.0% 2.77 [0.12, 66.02]Hewitt et al., 2013 1 30 1 30 0.0% 1.00 [0.07, 15.26]Delanaye et al., 2013 6 22 5 21 0.2% 1.15 [0.41, 3.19]Witham et al., 2013 - VitDISH 0 80 1 79 0.0% 0.33 [0.01, 7.96]

Total (95% CI) 18,719 18,831 100% 0.99 [0.95, 1.03]Total events 2,908 2,968Heterogeneity: Tau² = 0.00; Chi² = 28.51, df = 42 (P = 0.94); I² = 0%Test for overall effect: Z = 0.55 (P = 0.58) 0.05 0.2 1 5 20




Risk RatioM-H, Random, 95% CI in all-cause mortality risk

Study or SubgroupBrohult et al., 1973Inkovaara et al., 1983Corless et al., 1985Ott et al., 1989Gallagher et al., 1990Grady et al., 1991Hamdy et al., 1995Ooms et al., 1995Lips et al., 1996Sato et al., 1997Sato et al., 1999Komulainen et al., 1999 - OSTPREFrazao et al., 2000Gallagher et al., 2001 - STOP ITMeyer et al., 2002Trivedi et al., 2003Cooper et al., 2003Harwood et al., 2004 - NONOFDukas et al., 2004Sato et al., 2005Grant et al., 2005 - RECORDAloia et al., 2005Flicker et al., 2005Coyne et al., 2006Law et al., 2006Schleithoff et al., 2006Smith et al., 2007Lyons et al., 2007Prince et al., 2008Bjorkman et al., 2008Janssen et al.,2010Sanders et al., 2010 - Vital Dde Zeeuw et al., 2010 - VITALCherniack et al., 2011Grimnes et al. 2011Punthakee et al., 2012 - TIDEAlvarez et al., 2012Lehouck et al., 2012Glendenning et al., 2012Witham et al., 2013Hewitt et al., 2013Delanaye et al., 2013Witham et al., 2013 - VitDISH


Events1780114

11223

11011

169224

0711

4381

762

3477

355947

010

140

11001921160

2908

Total25454143255089

1771291

4543

11271

123569

13459338

19248

2649104313107

176261

47271725

1517336

1131952351

6072491

35339302280

18719

Events0581001

21251

10121

163247

1512

4602

851

3226

354953

190

4700121600151

2968

Total25424143254887

1711287

3943

11567

123575

13419437

18648

2643104312113

195562

47131715

1516834

1127932353

6142491

33336302179

18831

Weight0.0%0.1%0.2%0.0%0.0%0.0%0.0%0.3%6.4%0.0%0.0%0.0%0.0%0.0%5.2%6.3%0.0%0.2%0.0%0.0%

12.0%0.0%2.4%0.0%9.0%0.2%8.5%

47.0%0.0%0.2%0.0%1.0%0.0%0.0%0.0%0.0%0.0%0.2%0.0%0.0%0.0%0.2%0.0%

100.0%

M-H, Random, 95% CI3.00 [0.13, 70.30]

1.31 [0.45, 3.80]1.00 [0.42, 2.41]0.33 [0.01, 7.96]

3.00 [0.13, 70.30]2.88 [0.12, 69.07]3.91 [0.45, 34.29]

0.51 [0.25, 1.02]0.89 [0.75, 1.04]

0.87 [0.06, 13.40]3.00 [0.13, 71.65]

0.34 [0.01, 8.31]0.47 [0.04, 5.08]

1.00 [0.06, 15.81]1.05 [0.87, 1.26]0.90 [0.77, 1.07]0.34 [0.01, 8.16]1.36 [0.47, 3.91]

0.97 [0.06, 15.37]0.50 [0.05, 5.33]0.95 [0.84, 1.07]0.50 [0.05, 5.43]0.89 [0.68, 1.16]

2.11 [0.19, 22.96]1.20 [1.04, 1.37]1.19 [0.42, 3.33]1.00 [0.87, 1.15]0.99 [0.93, 1.05]0.33 [0.01, 8.12]1.04 [0.45, 2.39]

2.84 [0.12, 67.36]0.85 [0.56, 1.28]

2.94 [0.12, 71.20]3.00 [0.13, 70.02]

0.35 [0.01, 8.31]0.20 [0.01, 4.21]

1.00 [0.07, 15.08]1.50 [0.56, 4.04]

4.72 [0.23, 97.90]2.77 [0.12, 66.02]1.00 [0.07, 15.26]

1.15 [0.41, 3.19]0.33 [0.01, 7.96]

0.99 [0.95, 1.03]

Year1973198319851989199019911995199519961997199919992000200120022003200320042004200520052005200520062006200620072007200820082010201020102011201120122012201220122013201320132013



Supplementary Figure 12. Forest plot of vitamin D supplementation and all-cause mortality risk. M-H, Manthel-Haenszel. *Brohult et al., 1973 - data taken from meta-analysis Bjelakovic et al., 2014 †Gallagher et al., 1990 - data taken from meta-analysis Avenell et al., 2014 ‡Grady et al., 1991 - data taken from meta-analysis Chowdhury et al., 2014 §Ooms et al., 1995 - data taken from meta-analysis Zheng et al., 2013 ǁSato et al., 1999 - data taken from meta-analysis Bjelakovic et al., 2014 ¶Cooper et al., 2003 - data taken from meta-analysis Zheng et al., 2013**Sato et al., 2005 - data taken from meta-analysis Chowdhury et al., 2014, Bjelakovic et al., 2014 ††Aloia et al., 2005 - data taken from meta-analysis Zheng et al., 2013 ‡‡Smith et al., 2007 - data taken from meta-analysis Bjelakovic et al., 2014 §§Jansen et al., 2010 - data taken from meta-analysis Avenell et al., 2014. ǁǁAvenell et al., 2012 - Data provided included results from the RCT and follow-up period combined. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P <

83

0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Supplementary Figure 13. Funnel plot of vitamin D supplementation and CVD and all-cause mortality outcomes: (A) MI risk, (B) stroke risk, and (C) all-cause mortality risk. The vertical line represents the pooled effect estimate expressed as a RR. Dashed lines represent pseudo-95% confidence intervals (CI). The circles represent risk estimates for each study, and the horizontal lines represent standard errors of the RR. We were unable to test for funnel plot asymmetry for other CVD outcomes (<10 RCTs).

Random Effects


ALL-CAUSE MORTALITY All-cause mortality 1 2,297 48 0.99 [0.56, 1.72] N/A 0.96

Favours vitamin A / Favours control/ Negative association Positive association


0.5 1.0 1.5

84

Supplementary Figure 14. Summary of the pooled effect estimates of RCTsassessing the relationship vitamin A supplementation and all cause mortality risk. RR, risk ratio; CIs, confidence intervals; CVD, cardiovascular disease; CHD, coronary heart disease; MI, myocardial infarction; N/A, not applicable or not available. The diamonds represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% CIs, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTs Moon et al., 1997- SKICAP-AK 24 1,157 24 1,140 100.0% 0.99 [0.56, 1.72]

Total (95% CI) 1,157 1,140 100% 0.99 [0.56, 1.72]Total events 24 24 Heterogeneity: Not applicable 0.02 0.1 1 10 50Test for overall effect: Z = 0.05 (P = 0.96) Favours vitamin A Favours control

Risk RatioM-H, Random, 95% CI in all-cause mortality

Vitamin A ControlWeight


Study or SubgroupMoon et al., 1997- SKICAP-AK

Total (95% CI)Total eventsHeterogeneity: Not applicableTest for overall effect: Z = 0.05 (P = 0.96)

Events24

24

Total1157

1157

Events24

24

Total1140

1140

Weight100.0%

100.0%

M-H, Random, 95% CI0.99 [0.56, 1.72]

0.99 [0.56, 1.72]

Year1997

Vitamin A Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.02 0.1 1 10 50Favours Vitamin A Favours Control

Supplementary Figure 15. Forest plot of vitamin A supplementation and all-cause mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

85

Random Effects


CVDTotal CVD 3 70,118 2,912 1.03 [0.96, 1.10] 0% 0.44Total CHD 2 21,841 2,081 1.02 [0.94, 1.10] 0% 0.69MI 3 43,912 1,841 0.99 [0.90, 1.09] 0% 0.89Stroke 3 58,761 2,104 1.06 [0.95, 1.19] 36% 0.29Total CVD mortality 4 33,668 1,191 1.10 [0.98, 1.23] 0% 0.11Total CHD mortality 1 13,670 472 0.99 [0.83, 1.18] N/A 0.89MI mortality 1 8,171 34 0.42 [0.20, 0.87] N/A 0.02Stroke mortality 2 36,690 193 1.33 [0.73, 2.44] 61% 0.35



Favours beta-carotene/ Favours control/ Negative association Positive association

0.0 0.5 1.0 1.5 2.0

Supplementary Figure 16. Summary of the pooled effect estimates of RCT assessing the relationship between beta-carotene supplementation and CVD and all cause mortality risk. RR, risk ratio; CIs, confidence intervals; CVD, cardiovascular disease; CHD, coronary heart disease; MI, myocardial infarction; N/A, not applicable or not available. The diamonds represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% CIs, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsHennekens et al., 1996 - PHS 967 11,036 972 11,035 66.6% 0.99 [0.91, 1.08]Lee et al., 1999 - WHS 74 19,939 65 19,937 4.4% 1.14 [0.82, 1.59]Cook et al., 2007 - WACS 435 4,084 399 4,087 29.0% 1.09 [0.96, 1.24]

Total (95% CI) 35,059 35,059 100% 1.03 [0.96, 1.10]Total events 1,476 1,436 Heterogeneity: Tau² = 0.00; Chi² = 1.76, df = 2 (P = 0.42); I² = 0%Test for overall effect: Z = 0.78 (P = 0.44) 0.7 0.85 1 1.2 1.5

Favours beta-carotene Favours control

Beta-carotene ControlWeight



Study or SubgroupHennekens et al., 1996 - PHSLee et al., 1999 - WHSCook et al., 2007 - WACS


Events96774

435

1476

Total11036199394084

35059

Events97265

399

1436

Total11035199374087

35059

Weight66.6%4.4%

29.0%

100.0%

M-H, Random, 95% CI0.99 [0.91, 1.08]1.14 [0.82, 1.59]1.09 [0.96, 1.24]

1.03 [0.96, 1.10]

Year199619992007

Experimental Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.7 0.85 1 1.2 1.5Favours beta-carotene Favours control

Supplementary Figure 17. Forest plot of beta-carotene supplementation and total CVD risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

86

Subgroup and Study, Year Events Total Events TotalRCTsTornwall et al., 2004 - ATBC 548 6,821 534 6,849 50.8% 1.03 [0.92, 1.16]Cook et al., 2007 - WACS 500 4,084 499 4,087 49.2% 1.00 [0.89, 1.13]






Study or SubgroupTornwall et al., 2004 - ATBCCook et al., 2007 - WACS


Events548500

1048

Total68214084

10905

Events534499

1033

Total68494087

10936

Weight50.8%49.2%

100.0%

M-H, Random, 95% CI1.03 [0.92, 1.16]1.00 [0.89, 1.13]

1.02 [0.94, 1.10]

Year20042007



Supplementary Figure 18. Forest plot of beta-carotene supplementation and total CHD risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsHennekens et al., 1996 - PHS 468 11,036 489 11,035 52.0% 0.96 [0.85, 1.08]Tornwall et al., 2004 - ATBC 314 6,821 296 6,849 33.2% 1.07 [0.91, 1.24]Cook et al., 2007 - WACS 135 4,084 139 4,087 14.8% 0.97 [0.77, 1.23]

Total (95% CI) 21,941 21,971 100% 0.99 [0.91, 1.09]Total events 917 924Heterogeneity: Tau² = 0.00; Chi² = 1.16, df = 2 (P = 0.56); I² = 0%Test for overall effect: Z = 0.13 (P = 0.89) 0.85 0.9 1 1.1 1.2





Study or SubgroupHennekens et al., 1996 - PHSTornwall et al., 2004 - ATBCCook et al., 2007 - WACS


Events468314135

917

Total1103668214084

21941

Events489296139

924

Total1103568494087

21971

Weight52.0%33.2%14.8%

100.0%

M-H, Random, 95% CI0.96 [0.85, 1.08]1.07 [0.91, 1.24]0.97 [0.77, 1.23]

0.99 [0.91, 1.09]

Year199620042007



Supplementary Figure 19. Forest plot of beta-carotene supplementation and MI risk. M-H, Manthel-Haenszel, MI, myocardial infarction. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

87

Subgroup and Study, Year Events Total Events TotalRCTsHennekens et al., 1996 - PHS 367 11,036 382 11,035 36.6% 0.96 [0.83, 1.11]Leppala et al., 2000 - ATBC 554 14,246 503 14,273 44.3% 1.10 [0.98, 1.24]Cook et al., 2007 - WACS 161 4,084 137 4,087 19.2% 1.18 [0.94, 1.47]

Total (95% CI) 29,366 29,395 100% 1.06 [0.95, 1.19]Total events 1,082 1022Heterogeneity: Tau² = 0.00; Chi² = 3.15, df = 2 (P = 0.21); I² = 36%Test for overall effect: Z = 1.07 (P = 0.29) 0.7 0.85 1 1.2 1.5





Study or SubgroupHennekens et al., 1996 - PHSLeppala et al., 2000 - ATBCCook et al., 2007 - WACS


Events367554161

1082

Total11036142464084

29366

Events382503137

1022

Total11035142734087

29395

Weight36.6%44.3%19.2%

100.0%

M-H, Random, 95% CI0.96 [0.83, 1.11]1.10 [0.98, 1.24]1.18 [0.94, 1.47]

1.06 [0.95, 1.19]

Year199620002007



Supplementary Figure 20. Forest plot of beta-carotene supplementation and stroke risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsHennekens et al., 1996 - PHS 338 11,036 313 11,035 54.2% 1.08 [0.93, 1.26]Greenberg et al., 1996 - SCPS 68 913 59 892 11.0% 1.13 [0.80, 1.58]Green et al., 1999 - NSCPT 6 801 12 820 1.3% 0.51 [0.19, 1.36]Cook et al., 2007 - WACS 211 4,084 184 4,087 33.4% 1.15 [0.95, 1.39]

Total (95% CI) 16,834 16,834 100% 1.10 [0.98, 1.23]Total events 623 568Heterogeneity: Tau² = 0.00; Chi² = 2.62, df = 3 (P = 0.45); I² = 0%Test for overall effect: Z = 1.62 (P = 0.11) 0.5 0.7 1 1.5 2





Study or SubgroupHennekens et al., 1996 - PHSGreenberg et al., 1996 - Skin Cancer PreventionGreen et al., 1999 - NSCPTCook et al., 2007 - WACS


Events338686

211

623

Total11036

913801

4084

16834

Events3135912

184

568

Total11035

892820

4087

16834

Weight54.2%11.0%1.3%

33.4%

100.0%

M-H, Random, 95% CI1.08 [0.93, 1.26]1.13 [0.80, 1.58]0.51 [0.19, 1.36]1.15 [0.95, 1.39]

1.10 [0.98, 1.23]

Year1996199619992007


0.5 0.7 1 1.5 2Favours beta-carotene Favours control

Supplementary Figure 21. Forest plot of beta-carotene supplementation and CVD mortality risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

88

Subgroup and Study, Year Events Total Events TotalRCTsTornwall et al., 2004 - ATBC 234 6,821 238 6,849 100% 0.99 [0.83, 1.18]

Total (95% CI) 6,821 6,849 100% 0.99 [0.83, 1.18]Total events 234 238Heterogeneity: Not applicableTest for overall effect: Z = 0.14 (P = 0.89) 0.2 0.5 1 1.5 2.0





Study or SubgroupTornwall et al., 2004 - ATBC


Events234

234

Total6821

6821

Events238

238

Total6849

6849

Weight100.0%

100.0%

M-H, Random, 95% CI0.99 [0.83, 1.18]

0.99 [0.83, 1.18]

Year2004


0.2 0.5 1 2 5Favours beta-carotene Favours control

Supplementary Figure 22. Forest plot of beta-carotene supplementation and CHD mortality risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsCook et al., 2007 - WACS 10 4,084 24 4,087 100.0% 0.42 [0.20, 0.87]

Total (95% CI) 4,084 4,087 100% 0.42 [0.20, 0.87]Total events 10 24Heterogeneity: Not applicable 0.05 0.2 1 5 20 Test for overall effect: Z = 2.33 (P = 0.02) Favours beta-carotene Favours control




Study or SubgroupCook et al., 2007 - WACS


Events10

10

Total4084

4084

Events24

24

Total4087

4087

Weight100.0%

100.0%

M-H, Random, 95% CI0.42 [0.20, 0.87]

0.42 [0.20, 0.87]

Year2007



Supplementary Figure 23. Forest plot of beta-carotene supplementation and MI mortality risk. M-H, Manthel-Haenszel, MI, myocardial infarction. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsLeppala et al., 2000 - ATBC 82 14,246 78 14,273 63.5% 1.05 [0.77, 1.43]Cook et al., 2007 - WACS 22 4,084 11 4,087 36.5% 2.00 [0.97, 4.12]

Total (95% CI) 18,330 18,360 100% 1.33 [0.73, 2.44]Total events 104 89Heterogeneity: Tau² = 0.13; Chi² = 2.57, df = 1 (P = 0.11); I² = 61% 0.2 0.5 1 2 5 Test for overall effect: Z = 0.93 (P = 0.35) Favours beta-carotene Favours control




Study or SubgroupLeppala et al., 2000 - ATBCCook et al., 2007 - WACS


Events8222

104

Total14246

4084

18330

Events7811

89

Total14273

4087

18360

Weight63.5%36.5%

100.0%

M-H, Random, 95% CI1.05 [0.77, 1.43]2.00 [0.97, 4.12]

1.33 [0.73, 2.44]

Year20002007



Supplementary Figure 24. Forest plot of beta-carotene supplementation and stroke mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

89

Subgroup and Study, Year Events Total Events TotalRCTsGreenberg et al., 1996 - Skin Cancer Prevention 146 913 139 892 10.0% 1.03 [0.83, 1.27]Hennekens et al., 1996 - PHS 979 11,036 968 11,035 41.0% 1.01 [0.93, 1.10]Green et al., 1999 - NSCPT 11 820 21 801 1.0% 0.51 [0.25, 1.05]Correa et al., 2000 2 117 0 117 0.1% 5.00 [0.24, 103.03]Virtamo et al., 2003 - ATBC 919 7,282 851 7,287 39.5% 1.08 [0.99, 1.18]Cook 2007 - WACS 124 1,020 124 1,022 8.5% 1.00 [0.79, 1.27]

Total (95% CI) 21,188 21,154 100% 1.03 [0.96, 1.11]Total events 2,181 2,103 Heterogeneity: Tau² = 0.00; Chi² = 5.99, df = 5 (P = 0.31); I² = 17%Test for overall effect: Z = 0.90 (P = 0.37) 0.2 0.5 1 2 5





Study or SubgroupGreenberg et al., 1996 - Skin Cancer PreventionHennekens et al., 1996 - PHSGreen et al., 1999 - NSCPTCorrea et al., 2000Virtamo et al., 2003 - ATBCCook 2007 - WACS


Events146979

112

919124

2181

Total913

11036820117

72821020

21188

Events139968

210

851124

2103

Total892

11035801117

72871022

21154

Weight10.0%41.0%

1.0%0.1%

39.5%8.5%

100.0%

M-H, Random, 95% CI1.03 [0.83, 1.27]1.01 [0.93, 1.10]0.51 [0.25, 1.05]

5.00 [0.24, 103.03]1.08 [0.99, 1.18]1.00 [0.79, 1.27]

1.03 [0.96, 1.11]

Year199619961999200020032007


0.2 0.5 1 2 5Favours [experimental] Favours [control]

Supplementary Figure 25. Forest plot of beta-carotene supplementation and all-cause mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Random Effects


CVDTotal CVD 7 60,826 7,726 0.99 [0.95, 1.04] 0% 0.77

Total CHD 1 13,630 1,045 0.97 [0.86, 1.09] N/A 0.57

MI 6 42,134 1,815 0.98 [0.90, 1.08] 0% 0.73

Stroke 7 60,589 1,986 1.00 [0.92, 1.09] 0% 0.98

Total CVD mortality 7 49,729 2,263 1.02 [0.94, 1.10] 0% 0.65

Total CHD mortality 2 34,166 1,754 1.02 [0.93, 1.13] 10% 0.68

MI mortality 3 947 8 1.51 [0.39, 5.93] 0% 0.55

Stroke mortality 5 56,352 342 1.10 [0.87, 1.39] 7% 0.43


Favours antioxidants/ Favours control/

Negative association Positive association


0.0 0.5 1.0 1.5 2.0

Supplementary Figure 26. Summary of the pooled effect estimates of RCTs assessing the relationship between antioxidants supplementation and CVD and all cause mortality risk. RR, risk ratio; CIs, confidence intervals; RTCs, randomized clinical trials; CVD, cardiovascular disease; CHD, coronary heart disease; MI, myocardial infarction; N/A, not applicable. The diamonds represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% CIs, using the Manthel-Haenszel method with random effects model.

90

Subgroup and Study, Year Events Total Events TotalRCTsBrown et al., 2001 - HATS* 11 42 12 38 0.3% 0.83 [0.42, 1.65]HPS Collaborative Group 2002 2,306 10,269 2,312 10,267 63.7% 1.00 [0.95, 1.05]CLIPS Group 2007 16 185 11 181 0.3% 1.42 [0.68, 2.98]Plummer et al., 2007 1 990 - 990 0.0% 3.00 [0.12, 73.55]Sesso et al., 2008 - PHS II 310 3,656 316 3,653 7.3% 0.98 [0.84, 1.14]Lippman et al., 2009 - SELECT 1,041 8,904 1,050 8,910 25.3% 0.99 [0.92, 1.08]Hercberg et al., 2010 - SU.VI.MAX 137 6,377 143 6,364 3.1% 0.96 [0.76, 1.21]

Total (95% CI) 30,423 30,403 100% 0.99 [0.95, 1.04]Total events 3,822 3,844 Heterogeneity: Tau² = 0.00; Chi² = 1.79, df = 6 (P = 0.94); I² = 0%Test for overall effect: Z = 0.29 (P 0.2 0.5 1 2 5

Favours antioxidant Favours control

Antioxidant Control Risk RatioM-H, Random, 95% CI

Risk RatioM-H, Random, 95%CI in total CVD riskWeight

Study or SubgroupBrown et al., 2001 - HATSHPS Collaborative Group 2002CLIPS Group 2007Plummer et al., 2007Sesso et al., 2008 - PHS IILippman et al., 2009 - SELECTHercberg et al., 2010 - SU.VI.MAX


Events11

230616

1310

1041137

3822

Total42

10269185990

365689046377

30423

Events12

231211

0316

1050143

3844

Total38

10267181990

365389106364

30403

Weight0.3%

63.7%0.3%0.0%7.3%

25.3%3.1%

100.0%

M-H, Random, 95% CI0.83 [0.42, 1.65]1.00 [0.95, 1.05]1.42 [0.68, 2.98]

3.00 [0.12, 73.55]0.98 [0.84, 1.14]0.99 [0.92, 1.08]0.96 [0.76, 1.21]

0.99 [0.95, 1.04]

Year2001200220072007200820092010

Antioxidants Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.2 0.5 1 2 5Favours antioxidants Favours control

Supplementary Figure 27. Forest plot of antioxidants supplementation and total CVD risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. *Brown et al., 2001 - reported number of participants with at least one event. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsTornwall et al., 2004 - ATBC 511 6,781 534 6,849 100.0% 0.97 [0.86, 1.09]

Total (95% CI) 6,781 6,849 100% 0.97 [0.86, 1.09]Total events 511 534 Heterogeneity: Not applicable 0.5 0.7 1 1.5 2Test for overall effect: Z = 0.57 (P = 0.57) Favours antioxidant Favours control

Antioxidant ControlWeight


Risk RatioM-H, Random, 95%CI in total CHD risk

Study or SubgroupTornwall et al., 2004 - ATBC


Events511

511

Total6781

6781

Events534

534

Total6849

6849

Weight100.0%

100.0%

M-H, Random, 95% CI0.97 [0.86, 1.09]

0.97 [0.86, 1.09]

Year2004


0.5 0.7 1 1.5 2Favours antioxidants Favours control

Supplementary Figure 28. Forest plot of antioxidants supplementation and total CHD risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence intervals, using the Manthel-Haenszel method with random effects model.

91

Subgroup and Study, Year Events Total Events TotalRCTsBrown et al., 2001 - HATS* 1 42 4 38 0.2% 0.23 [0.03, 1.94]HPS Collaborative Group 2002 464 10,269 467 10,267 51.6% 0.99 [0.88, 1.13]Waters et al., 2002 - WAVE 3 105 1 108 0.2% 3.09 [0.33, 29.19]Tornwall et al., 2004 - ATBC 289 6,781 296 6,849 32.3% 0.99 [0.84, 1.16]CLIPS Group 2007 9 185 4 181 0.6% 2.20 [0.69, 7.02]Sesso et al., 2008 - PHS II 133 3,656 144 3,653 15.2% 0.92 [0.73, 1.16]

Total (95% CI) 21,038 21,096 100% 0.98 [0.90, 1.08]Total events 899 916 Heterogeneity: Tau² = 0.00; Chi² = 4.97, df = 5 (P = 0.42); I² = 0%Test for overall effect: Z = 0.35 (P = 0.73) 0.2 0.5 1 2 5




Risk RatioM-H, Random, 95%CI in MI risk

Study or SubgroupBrown et al., 2001 - HATSHPS 2002Waters et al., 2002 - WAVETornwall et al., 2004 - ATBCCLIPS 2007Sesso et al., 2008 - PHS II


Events1

4643

2899

133

899

Total42

10269105

6781185

3656

21038

Events4

4671

2964

144

916

Total38

10267108

6849181

3653

21096

Weight0.2%

51.6%0.2%

32.3%0.6%

15.2%

100.0%

M-H, Random, 95% CI0.23 [0.03, 1.94]0.99 [0.88, 1.13]

3.09 [0.33, 29.19]0.99 [0.84, 1.16]2.20 [0.69, 7.02]0.92 [0.73, 1.16]

0.98 [0.90, 1.08]

Year200120022002200420072008



Supplementary Figure 29. Forest plot of antioxidants supplementation and MI risk. M-H, Manthel-Haenszel, MI, myocardial infarction. *Brown et al., 2001 - reported number of participants with at least one event. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsLeppala et al., 2000 - ATBC 258 7,118 252 7,153 25.6% 1.03 [0.87, 1.22]Brown et al., 2001 - HATS* 2 42 2 38 0.2% 0.90 [0.13, 6.11]Waters et al., 2002 - WAVE 1 105 3 108 0.1% 0.34 [0.04, 3.24]HPS Collaborative Group 2002 511 10,269 518 10,267 52.4% 0.99 [0.88, 1.11]CLIPS Group 2007 6 185 5 181 0.5% 1.17 [0.36, 3.78]Sesso et al., 2008 - PHS II 104 3,656 113 3,653 10.8% 0.92 [0.71, 1.20]Lippman et al., 2009 - SELECT 111 8,904 100 8,910 10.3% 1.11 [0.85, 1.45]





Risk RatioM-H, Random, 95%CI in stroke risk

Study or SubgroupLeppala et al., 2000 - ATBCBrown et al., 2001 - HATSWaters et al., 2002 - WAVEHPS Collaborative Group 2002CLIPS Group 2007Sesso et al., 2008 - PHS IILippman et al., 2009 - SELECT


Events258

21

5116

104111

993

Total7118

42105

10269185

36568904

30279

Events252

23

5185

113100

993

Total7153

38108

10267181

36538910

30310

Weight25.6%

0.2%0.1%

52.4%0.5%

10.8%10.3%

100.0%

M-H, Random, 95% CI1.03 [0.87, 1.22]0.90 [0.13, 6.11]0.34 [0.04, 3.24]0.99 [0.88, 1.11]1.17 [0.36, 3.78]0.92 [0.71, 1.20]1.11 [0.85, 1.45]

1.00 [0.92, 1.09]

Year2000200120022002200720082009



Supplementary Figure 30. Forest plot of antioxidants supplementation and stroke risk. M-H, Manthel-Haenszel. *Brown et al., 2001 - reported number of participants with at least one event. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

92

Subgroup and Study, Year Events Total Events TotalRCTsBrown et al., 2001 - HATS - 42 1 38 0.1% 0.30 [0.01, 7.21]You et al., 2001 9 1,706 12 1,705 0.9% 0.75 [0.32, 1.77]Waters et al., 2002 - WAVE 4 105 2 108 0.2% 2.06 [0.38, 10.99]HPS Collaborative Group 2002 878 10,269 840 10,267 77.2% 1.05 [0.95, 1.14]CLIPS Group 2007 6 185 3 181 0.3% 1.96 [0.50, 7.71]Sesso et al., 2008 - PHS II 127 3,656 122 3,653 10.6% 1.04 [0.81, 1.33]Lippman et al., 2009 - SELECT 117 8,904 142 8,910 10.7% 0.82 [0.65, 1.05]





Risk RatioM-H, Random, 95%CI in CVD mortality risk

Study or SubgroupBrown et al., 2001 - HATSYou et al., 2001Waters et al., 2002 - WAVEHPS Collaborative Group 2002CLIPS Group 2007Sesso et al., 2008 - PHS IILippman et al., 2009 - SELECT


Events094

8786

127117

1141

Total42

1706105

10269185

36568904

24867

Events1

122

8403

122142

1122

Total38

1705108

10267181

36538910

24862

Weight0.1%0.9%0.2%

77.2%0.3%

10.6%10.7%

100.0%

M-H, Random, 95% CI0.30 [0.01, 7.21]0.75 [0.32, 1.77]

2.06 [0.38, 10.99]1.05 [0.95, 1.14]1.96 [0.50, 7.71]1.04 [0.81, 1.33]0.82 [0.65, 1.05]

1.02 [0.94, 1.10]

Year2001200120022002200720082009



Supplementary Figure 31. Forest plot of antioxidants supplementation and CVD mortality risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsHPS Collaborative Group 2002 664 10,269 630 10,267 72.0% 1.05 [0.95, 1.17]Tornwall et al., 2004 - ATBC 222 6,781 238 6,849 28.0% 0.94 [0.79, 1.13]

Total (95% CI) 17,050 17,116 100% 1.02 [0.93, 1.13]Total events 886 868 Heterogeneity: Tau² = 0.00; Chi² = 1.11, df = 1 (P = 0.29); I² = 10% 0.5 0.7 1 1.5 2 Test for overall effect: Z = 0.42 (P = 0.68) Favours antioxidant Favours control



Risk RatioM-H, Random, 95%CI in CHD mortality risk

Study or SubgroupHPS 2002Tornwall et al., 2004 - ATBC


Events664222

886

Total10269

6781

17050

Events630238

868

Total10267

6849

17116

Weight72.0%28.0%

100.0%

M-H, Random, 95% CI1.05 [0.95, 1.17]0.94 [0.79, 1.13]

1.02 [0.93, 1.13]

Year20022004


0.5 0.7 1 1.5 2Favours antioxidants Favours control

Supplementary Figure 32. Forest plot of antioxidants supplementation and CHD mortality risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

93

Subgroup and Study, Year Events Total Events TotalRCTsChylack et al.., 2002 - REACT 2 149 1 148 32.6% 1.99 [0.18, 21.67]Mooney et al., 2005 1 142 0 142 18.3% 3.00 [0.12, 73.03]CLIPS Group 2007 2 185 2 181 49.1% 0.98 [0.14, 6.87]

Total (95% CI) 476 471 100% 1.51 [0.39, 5.93]Total events 5 3 Heterogeneity: Tau² = 0.00; Chi² = 0.42, df = 2 (P = 0.81); I² = 0%Test for overall effect: Z = 0.59 (P = 0.55) 0.02 0.1 1 10 50




Risk RatioM-H, Random, 95%CI in MI mortality risk

Study or SubgroupChylack et al.., 2002 - REACTMooney et al., 2005CLIPS Group 2007


Events212

5

Total149142185

476

Events102

3

Total148142181

471

Weight32.6%18.3%49.1%

100.0%

M-H, Random, 95% CI1.99 [0.18, 21.67]3.00 [0.12, 73.03]0.98 [0.14, 6.87]

1.51 [0.39, 5.93]

Year200220052007



Supplementary Figure 33. Forest plot of antioxidants supplementation and MI mortality risk. M-H, Manthel-Haenszel, MI, myocardial infarction. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsLeppala et al., 2000 - ATBC 46 7,118 34 7,153 25.4% 1.36 [0.87, 2.12]HPS Collaborative Group 2002 108 10,269 107 10,267 58.9% 1.01 [0.77, 1.32]CLIPS Group 2007 3 185 0 181 0.6% 6.85 [0.36, 131.67]Lippman et al., 2009 - SELECT 12 8,904 8 8,910 6.8% 1.50 [0.61, 3.67]Ma et al., 2012 10 1,677 14 1,688 8% 0.72 [0.32, 1.61]





Risk RatioM-H, Random, 95%CI in stroke mortality risk

Study or SubgroupLeppala et al., 2000 - ATBCHPS Collaborative Group 2002CLIPS Group 2007Lippman et al., 2009 - SELECTMa et al., 2012


Events46

1083

1210

179

Total7118

10269185

89041677

28153

Events34

10708

14

163

Total7153

10267181

89101688

28199

Weight25.4%58.9%0.6%6.8%8.2%

100.0%

M-H, Random, 95% CI1.36 [0.87, 2.12]1.01 [0.77, 1.32]

6.85 [0.36, 131.67]1.50 [0.61, 3.67]0.72 [0.32, 1.61]

1.10 [0.87, 1.39]

Year20002002200720092012



Supplementary Figure 34. Forest plot of antioxidants supplementation and stroke mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

94

Subgroup and Study, Year Events Total Events TotalRCTsMcKeown-Eyssen et al., 1988 4 96 3 89 0.10% 1.24 [0.28, 5.37]Omenn et al., 1996 - CARET 544 9,420 424 8,894 12.50% 1.21 [1.07, 1.37]Girodon et al., 1997* 7 21 7 20 0.40% 0.95 [0.41, 2.23]Girodon et al., 1999 - MIN.VIT.AOX† 55 181 51 182 2.6% 1.08 [0.79, 1.49]Salonen et al., 2000 - ASAP 1 130 1 130 0.0% 1.00 [0.06, 15.82]Correa et al., 2000 2 121 0 117 0.0% 4.84 [0.23, 99.67]Jacobson et al., 2000 0 57 1 55 0.0% 0.32 [0.01, 7.74]Brown et al., 2001 - HATS 0 42 1 38 0.0% 0.30 [0.01, 7.21]AREDS Research Group 2001 251 2,304 240 2,325 8.1% 1.06 [0.89, 1.25]HPS Collaborative Group 2002 1,446 10,269 1,389 10,267 22.8% 1.04 [0.97, 1.11]Waters et al., 2002 - WAVE 6 105 2 108 0.1% 3.09 [0.64, 14.95]Chylack et al.., 2002 - REACT 9 149 3 148 0.2% 2.98 [0.82, 10.79]Virtamo et al., 2003 - ATBC 932 7,278 851 7,287 18.6% 1.10 [1.01, 1.20]Mooney et al., 2005 1 142 0 142 0.0% 3.00 [0.12, 73.03]CLIPS Group 2007 7 185 4 181 0.2% 1.71 [0.51, 5.75]Plummer et al., 2007 16 990 11 990 0.5% 1.45 [0.68, 3.12]Cook et al., 2007 - WAC 133 1,020 124 1,022 4.8% 1.07 [0.85, 1.35]Lippman et al., 2009 - SELECT 359 8,904 382 8,910 10.4% 0.94 [0.82, 1.08]Hercberg et al., 2010 - SU.VI.MAX 77 6,377 99 6,364 3.1% 0.78 [0.58, 1.04]Ma et al., 2012 - SIT 82 1,706 101 1,705 3.3% 0.81 [0.61, 1.08]Wang et al., 2014 - PHS II 440 3,656 406 3,653 12.1% 1.08 [0.95, 1.23]

Total (95% CI) 53,153 52,627 100.0% 1.06 [1.00, 1.12]Total events 4372 4100Heterogeneity: Tau² = 0.00; Chi² = 24.05, df = 20 (P = 0.24); I² = 17%Test for overall effect: Z = 2.00 (P = 0.05) 0.1 0.2 0.5 1 2 5 10




Risk RatioM-H, Random, 95%CI in all-cause mortality risk

Study or SubgroupMcKeown-Eyssen et al., 1988Omenn et al., 1996 - CARETGirodon et al., 1997Girodon et al., 1999 - MIN.VIT.AOXSalonen et al., 2000 - ASAPCorrea et al., 2000Jacobson et al., 2000Brown et al., 2001 - HATSAREDS Research Group 2001HPS Collaborative Group 2002Waters et al., 2002 - WAVEChylack et al.., 2002 - REACTVirtamo et al., 2003 - ATBCMooney et al., 2005CLIPS Group 2007Plummer et al., 2007Cook et al., 2007 - WACLippman et al., 2009 - SELECTHercberg et al., 2010 - SU.VI.MAXMa et al., 2012 - SITWang et al., 2014 - PHS II


Events4

5447

551200

2511446

69

93217

161333597782

440

4372

Total96

942021

1811301215742

230410269

105149

7278142185990

10208904637717063656

53153

Events3

4247

511011

2401389

23

85104

1112438299

101406

4100

Total89

889420

1821301175538

232510267

108148

7287142181990

10228910636417053653

52627

Weight0.1%

12.5%0.4%2.6%0.0%0.0%0.0%0.0%8.1%

22.8%0.1%0.2%

18.6%0.0%0.2%0.5%4.8%

10.4%3.1%3.3%

12.1%

100.0%

M-H, Random, 95% CI1.24 [0.28, 5.37]1.21 [1.07, 1.37]0.95 [0.41, 2.23]1.08 [0.79, 1.49]

1.00 [0.06, 15.82]4.84 [0.23, 99.67]0.32 [0.01, 7.74]0.30 [0.01, 7.21]1.06 [0.89, 1.25]1.04 [0.97, 1.11]

3.09 [0.64, 14.95]2.98 [0.82, 10.79]1.10 [1.01, 1.20]

3.00 [0.12, 73.03]1.71 [0.51, 5.75]1.45 [0.68, 3.12]1.07 [0.85, 1.35]0.94 [0.82, 1.08]0.78 [0.58, 1.04]0.81 [0.61, 1.08]1.08 [0.95, 1.23]

1.06 [1.00, 1.12]

Year198819961997199920002000200020012001200220022002200320052007200720072009201020122014


0.1 0.2 0.5 1 2 5 10Favours antioxidants Favours control

Supplementary Figure 35. Forest plot of antioxidants supplementation and all-cause mortality risk. M-H, Manthel-Haenszel. *Girodon et al., 1997 - vitamin C, vitamin E, β-carotene, selenium and zinc vs placebo; †Girodon et al., 1999 - vitamin C, vitamin E, β-carotene, selenium and zinc vs placebo. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

95

Supplementary Figure 36. Funnel plot of antioxidants supplementation and all-cause mortality risk. The vertical line represents the pooled effect estimate expressed as a RR. Dashed lines represent pseudo-95% confidence intervals (CI). The circles represent risk estimates for each study, and the horizontal lines represent standard errors of the RR. We were unable to test for other CVD outcomes (<10 RCTs).

Random Effects

Comparison RCTs N Events RR (95% CIs) RR (95% CIs) I2 p -valueCVDTotal CVD 10 100,475 7,936 0.96 [0.89, 1.03] 50% 0.23

Total CHD 2 21,840 2,052 0.97 [0.90, 1.06] 0% 0.50MI 10 87,049 3,119 0.87 [0.75, 1.01] 64% 0.07Stroke 11 101 1,956 0.98 [0.88, 1.08] 15% 0.67Total CVD mortality 11 102,533 2,887 0.95 [0.89, 1.02] 0% 0.16Total CHD mortality 2 24,993 970 0.90 [0.80, 1.02] 0% 0.10MI mortality 7 66,417 592 0.91 [0.78, 1.07] 0% 0.26Stroke mortality 5 72,317 104 0.97 [0.66, 1.43] 0% 0.88


Favours vitamin E/ Favours control/ Negative association Positive association


0.0 0.5 1.0 1.5 2.0

Supplementary Figure 37. Summary of the pooled effect estimates of RCTs assessing the relationship between vitamin E supplementation and CVD and all-cause mortality risk. RR, risk ratio; CIs, confidence intervals; RTCs, randomized clinical trials; CVD, cardiovascular disease; CHD, coronary heart disease; MI, myocardial infarction; N/A, not applicable. The diamonds represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% CIs, using the Manthel-Haenszel method with random effects model.

96

Subgroup and Study, Year Events Total Events TotalRCTsGISSI-Prevenzione Investigators 1999 571 5,660 584 5,664 15.6% 0.98 [0.88, 1.09]Boaz et al., 2000 - SPACE 15 97 33 99 1.5% 0.46 [0.27, 0.80]de Gaetano et al., 2001 - PPP 56 2,231 53 2,264 3.1% 1.07 [0.74, 1.55]Hodis et al., 2002 - VEAPS 11 177 14 176 0.8% 0.78 [0.36, 1.67]Lonn et al., 2005 - HOPE & HOPE-TOO 1,022 4,761 985 4,780 19.3% 1.04 [0.96, 1.13]Lee et al., 2005 - WHS 482 19,937 517 19,939 14.2% 0.93 [0.82, 1.05]Cook et al., 2007 - WACS 399 4,083 435 4,088 13.6% 0.92 [0.81, 1.04]Sesso et al., 2008 - PHS II 310 3,659 316 3,653 11.6% 0.98 [0.84, 1.14]Milman et al., 2008 16 726 33 708 1.3% 0.47 [0.26, 0.85]Lippman et al., 2009 - SELECT 1,034 8,863 1,050 8,910 19% 0.99 [0.91, 1.07]

Total (95% CI) 50,194 50,281 100% 0.96 [0.89, 1.03]Total events 3,916 4,020 Heterogeneity: Tau² = 0.00; Chi² = 17.84, df = 9 (P = 0.04); I² = 50%Test for overall effect: Z = 1.21 (P = 0.23) 0.5 0.7 1 1.5 2

Favours vitamin E Favours control

Vitamin E ControlWeight


Risk RatioM-H, Random, 95%CI in total CVD risk

Study or SubgroupGISSI-Prevenzione Investigators 1999Boaz et al., 2000 - SPACEde Gaetano et al., 2001 - PPPHodis et al., 2002 - VEAPSLonn et al., 2005 - HOPE & HOPE-TOOLee et al., 2005 - WHSCook et al., 2007 - WACSMilman et al., 2008Sesso et al., 2008 - PHS IILippman et al., 2009 - SELECT


Events571

155611

1022482399

16310

1034

3916

Total5660

972231

1774761

199374083

72636598863

50194

Events584

335314

985517435

33316

1050

4020

Total5664

992264176

478019939

4088708

36538910

50281

Weight15.6%

1.5%3.1%0.8%

19.3%14.2%13.6%

1.3%11.6%19.0%

100.0%

M-H, Random, 95% CI0.98 [0.88, 1.09]0.46 [0.27, 0.80]1.07 [0.74, 1.55]0.78 [0.36, 1.67]1.04 [0.96, 1.13]0.93 [0.82, 1.05]0.92 [0.81, 1.04]0.47 [0.26, 0.85]0.98 [0.84, 1.14]0.99 [0.91, 1.07]

0.96 [0.89, 1.03]

Year1999200020012002200520052007200820082009

Vitamin E Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.5 0.7 1 1.5 2Favours vitamin E Favours control

Supplementary Figure 38. Forest plot of vitamin E supplementation and total CVD risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsVirtamo et al., 1998 - ATBC 519 6,820 534 6,849 50.1% 0.98 [0.87, 1.10]Cook et al., 2007 - WACS 491 4,083 508 4,088 49.9% 0.97 [0.86, 1.09]





Risk RatioM-H, Random, 95%CI in CHD risk

Study or SubgroupVirtamo et al., 1998 - ATBCCook et al., 2007 - WACS


Events519491

1010

Total68204083

10903

Events534508

1042

Total68494088

10937

Weight50.1%49.9%

100.0%

M-H, Random, 95% CI0.98 [0.87, 1.10]0.97 [0.86, 1.09]

0.97 [0.90, 1.06]

Year19982007


0.7 0.85 1 1.2 1.5Favours Vitamin E Favours Control

Supplementary Figure 39. Forest plot of vitamin E supplementation and total CHD risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

97

Subgroup and Study, Year Events Total Events TotalRCTsStephens et al., 1996 - CHAOS 32 1,035 54 967 7.8% 0.55 [0.36, 0.85]Virtamo et al., 1998 - ATBC 307 6,820 296 6,849 17.4% 1.04 [0.89, 1.22]Boaz et al., 2000 - SPACE 5 97 17 99 2.2% 0.30 [0.12, 0.78]de Gaetano et al., 2001 - PPP 22 2,231 25 2,264 5.3% 0.89 [0.51, 1.58]Hodis et al., 2002 - VEAPS 5 177 6 176 1.5% 0.83 [0.26, 2.67]Lee et al., 2005 - WHS 196 19,937 195 19,939 15.7% 1.01 [0.83, 1.22]Lonn et al., 2005 - HOPE & HOPE-TOO 724 4,761 686 4,780 19.7% 1.06 [0.96, 1.17]Cook et al., 2007 - WACS 131 4,083 143 4,088 14.2% 0.92 [0.73, 1.16]Milman et al., 2008 7 726 17 708 2.6% 0.40 [0.17, 0.96]Sesso et al., 2008 - PHS II 107 3,659 144 3,653 13.6% 0.74 [0.58, 0.95]






Study or SubgroupStephens et al., 1996 - CHAOSVirtamo et al., 1998 - ATBCBoaz et al., 2000 - SPACEde Gaetano et al., 2001 - PPPHodis et al., 2002 - VEAPSLee et al., 2005 - WHSLonn et al., 2005 - HOPE & HOPE-TOOCook et al., 2007 - WACSMilman et al., 2008Sesso et al., 2008 - PHS II


Events32

3075

225

196724131

7107

1536

Total10356820

972231

17719937

47614083

7263659

43526

Events54

2961725

6195686143

17144

1583

Total967

684999

2264176

1993947804088

7083653

43523

Weight7.8%

17.4%2.2%5.3%1.5%

15.7%19.7%14.2%

2.6%13.6%

100.0%

M-H, Random, 95% CI0.55 [0.36, 0.85]1.04 [0.89, 1.22]0.30 [0.12, 0.78]0.89 [0.51, 1.58]0.83 [0.26, 2.67]1.01 [0.83, 1.22]1.06 [0.96, 1.17]0.92 [0.73, 1.16]0.40 [0.17, 0.96]0.74 [0.58, 0.95]

0.87 [0.75, 1.01]

Year1996199820002001200220052005200720082008


0.5 0.7 1 1.5 2Favours Vitamin E Favours Control

Supplementary Figure 40. Forest plot of vitamin E supplementation and MI risk. M-H, Manthel-Haenszel, MI, myocardial infarction. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsSteiner et al., 1995 3 52 6 48 0.6% 0.46 [0.12, 1.74]GISSI-Prevenzione Investigators 1999 83 5,660 95 5,664 10.4% 0.87 [0.65, 1.17]Boaz et al., 2000 - SPACE 5 97 6 99 0.8% 0.85 [0.27, 2.70]de Gaetano et al., 2001 - PPP 22 2,231 18 2,264 2.6% 1.24 [0.67, 2.31]Hodis et al., 2002 - VEAPS 0 177 2 176 0.1% 0.20 [0.01, 4.11]Lee et al., 2005 - WHS 241 19,937 246 19,939 22.1% 0.98 [0.82, 1.17]Lonn et al., 2005 - HOPE & HOPE-TOO 270 4,761 246 4,780 23.6% 1.10 [0.93, 1.30]Cook et al., 2007 - WACS 137 4,083 151 4,088 15.5% 0.91 [0.72, 1.14]Milman et al., 2008 6 726 11 708 1.1% 0.53 [0.20, 1.43]Sesso et al., 2008 - PHS II 133 3,659 113 3,653 13.70% 1.18 [0.92, 1.50]Lippman et al., 2009 - SELECT 70 8,863 92 8,910 9.40% 0.76 [0.56, 1.04]






Study or SubgroupSteiner et al., 1995GISSI-Prevenzione Investigators 1999Boaz et al., 2000 - SPACEde Gaetano et al., 2001 - PPPHodis et al., 2002 - VEAPSLee et al., 2005 - WHSLonn et al., 2005 - HOPE & HOPE-TOOCook et al., 2007 - WACSMilman et al., 2008Sesso et al., 2008 - PHS IILippman et al., 2009 - SELECT


Events3

835

220

241270137

6133

70

970

Total52

566097

2231177

1993747614083

72636598863

50246

Events6

956

182

246246151

11113

92

986

Total48

566499

2264176

1993947804088

70836538910

50329

Weight0.6%

10.4%0.8%2.6%0.1%

22.1%23.6%15.5%

1.1%13.7%

9.4%

100.0%

M-H, Random, 95% CI0.46 [0.12, 1.74]0.87 [0.65, 1.17]0.85 [0.27, 2.70]1.24 [0.67, 2.31]0.20 [0.01, 4.11]0.98 [0.82, 1.17]1.10 [0.93, 1.30]0.91 [0.72, 1.14]0.53 [0.20, 1.43]1.18 [0.92, 1.50]0.76 [0.56, 1.04]

0.98 [0.88, 1.08]

Year19951999200020012002200520052007200820082009


0.2 0.5 1 2 5Favours vitamin E Favours control

Supplementary Figure 41. Forest plot of vitamin E supplementation and stroke risk. * M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

98

Subgroup and Study, Year Events Total Events TotalRCTsStephens et al., 1996 - CHAOS 27 1,035 23 967 1.7% 1.10 [0.63, 1.90]GISSI-Prevenzione Investigators 1999 310 5,660 329 5,664 22.1% 0.94 [0.81, 1.10]Boaz et al., 2000 - SPACE 9 97 15 99 0.8% 0.61 [0.28, 1.33]de Gaetano et al., 2001 - PPP 22 2,231 26 2,264 1.6% 0.86 [0.49, 1.51]Lee et al., 2005 - WHS 106 19,937 140 19,939 7.9% 0.76 [0.59, 0.97]Lonn et al., 2005 - HOPE & HOPE-TOO 482 4,761 475 4,780 34.8% 1.02 [0.90, 1.15]Magliano et al., 2006 - MAVET 2 205 4 204 0.2% 0.50 [0.09, 2.69]Cook et al., 2007 - WACS 193 4,083 202 4,088 13.6% 0.96 [0.79, 1.16]Sesso et al., 2008 - PHS II 131 3,659 122 3,653 8.6% 1.07 [0.84, 1.37]Milman et al., 2008 3 726 5 708 0.2% 0.59 [0.14, 2.44]Lippman et al., 2009 - SELECT 119 8,863 142 8,910 8.60% 0.84 [0.66, 1.07]






Study or SubgroupStephens et al., 1996 - CHAOSGISSI-Prevenzione Investigators 1999Boaz et al., 2000 - SPACEde Gaetano et al., 2001 - PPPLee et al., 2005 - WHSLonn et al., 2005 - HOPE & HOPE-TOOMagliano et al., 2006 - MAVETCook et al., 2007 - WACSSesso et al., 2008 - PHS IIMilman et al., 2008Lippman et al., 2009 - SELECT


Events27

3109

22106482

2193131

3119

1404

Total10355660

972231

199374761205

40833659726

8863

51257

Events23

3291526

140475

4202122

5142

1483

Total967

566499

226419939

4780204

40883653

7088910

51276

Weight1.7%

22.1%0.8%1.6%7.9%

34.8%0.2%

13.6%8.6%0.2%8.6%

100.0%

M-H, Random, 95% CI1.10 [0.63, 1.90]0.94 [0.81, 1.10]0.61 [0.28, 1.33]0.86 [0.49, 1.51]0.76 [0.59, 0.97]1.02 [0.90, 1.15]0.50 [0.09, 2.69]0.96 [0.79, 1.16]1.07 [0.84, 1.37]0.59 [0.14, 2.44]0.84 [0.66, 1.07]

0.95 [0.89, 1.02]

Year19961999200020012005200520062007200820082009



Supplementary Figure 42. Forest plot of vitamin E supplementation and CVD mortality risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Risk Ratio Risk RatioSubgroup and Study, Year Events Total Events Total Weight M-H, Random, 95% CI M-H, Random, 95% CI in CHD mortality riskRCTsVirtamo et al., 1998 - ATBC 212 6,820 238 6,849 46.0% 0.89 [0.75, 1.07]GISSI-Prevenzione Investigators 1999 247 5,660 273 5,664 54.0% 0.91 [0.77, 1.07]

Total (95% CI) 12,480 12,513 100% 0.90 [0.80, 1.02]Total events 459 511 Heterogeneity: Tau² = 0.00; Chi² = 0.01, df = 1 (P = 0.92); I² = 0%Test for overall effect: Z = 1.67 (P = 0.10) 0.85 0.9 1 1.1 1.2


Vitamin E Control

Study or SubgroupVirtamo et al., 1998 - ATBCGISSI-Prevenzione Investigators 1999


Events212247

459

Total68205660

12480

Events238273

511

Total68495664

12513

Weight46.0%54.0%

100.0%

M-H, Random, 95% CI0.89 [0.75, 1.07]0.91 [0.77, 1.07]

0.90 [0.80, 1.02]

Year19981999


0.5 0.7 1 1.5 2Favours Vitamin E Favours Control

Supplementary Figure 43. Forest plot of vitamin E supplementation and CHD mortality risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

99

Subgroup and Study, Year Events Total Events TotalRCTsStephens et al., 1996 - CHAOS 18 1,035 13 967 5.1% 1.29 [0.64, 2.63]GISSI-Prevenzione Investigators 1999 228 5,660 251 5,664 82.3% 0.91 [0.76, 1.08]Boaz et al., 2000 - SPACE 2 97 8 99 1.1% 0.26 [0.06, 1.17]de Gaetano et al., 2001 - PPP 3 2,231 7 2,264 1.4% 0.43 [0.11, 1.68]Hodis et al., 2002 - VEAPS 1 177 1 176 0.3% 0.99 [0.06, 15.77]Lee et al., 2005 - WHS 12 19,937 14 19,939 4.3% 0.86 [0.40, 1.85]Cook et al., 2007 - WACS 18 4,083 16 4,088 5.6% 1.13 [0.58, 2.21]

Total (95% CI) 33,220 33,197 100% 0.91 [0.78, 1.07]Total events 282 310 Heterogeneity: Tau² = 0.00; Chi² = 5.19, df = 6 (P = 0.52); I² = 0%Test for overall effect: Z = 1.13 (P = 0.26) 0.1 0.2 0.5 1 2 5 10





Study or SubgroupStephens et al., 1996 - CHAOSGISSI-Prevenzione Investigators 1999Boaz et al., 2000 - SPACEde Gaetano et al., 2001 - PPPHodis et al., 2002 - VEAPSLee et al., 2005 - WHSCook et al., 2007 - WACS


Events18

228231

1218

282

Total10355660

972231

17719937

4083

33220

Events13

251871

1416

310

Total967

566499

2264176

199394088

33197

Weight5.1%

82.3%1.1%1.4%0.3%4.3%5.6%

100.0%

M-H, Random, 95% CI1.29 [0.64, 2.63]0.91 [0.76, 1.08]0.26 [0.06, 1.17]0.43 [0.11, 1.68]

0.99 [0.06, 15.77]0.86 [0.40, 1.85]1.13 [0.58, 2.21]

0.91 [0.78, 1.07]

Year1996199920002001200220052007


0.1 0.2 0.5 1 2 5 10Favours Vitamin E Favours Control

Supplementary Figure 44. Forest plot of vitamin E supplementation and MI mortality risk. M-H, Manthel-Haenszel, MI, myocardial infarction. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsStephens et al., 1996 - CHAOS 1 1,035 1 967 2.0% 0.93 [0.06, 14.92]de Gaetano et al., 2001 - PPP 2 2,231 5 2,264 5.6% 0.41 [0.08, 2.09]Lee et al., 2005 - WHS 21 19,937 24 19,939 43.8% 0.88 [0.49, 1.57]Cook et al., 2007 - WACS 18 4,083 15 4,088 32.1% 1.20 [0.61, 2.38]Lippman et al., 2009 - SELECT 9 8,863 8 8,910 16.6% 1.13 [0.44, 2.93]






Study or SubgroupStephens et al., 1996 - CHAOSde Gaetano et al., 2001 - PPPLee et al., 2005 - WHSCook et al., 2007 - WACSLippman et al., 2009 - SELECT


Events12

2118

9

51

Total10352231

1993740838863

36149

Events15

2415

8

53

Total967

226419939

40888910

36168

Weight2.0%5.6%

43.8%32.1%16.6%

100.0%

M-H, Random, 95% CI0.93 [0.06, 14.92]

0.41 [0.08, 2.09]0.88 [0.49, 1.57]1.20 [0.61, 2.38]1.13 [0.44, 2.93]

0.97 [0.66, 1.43]

Year19962001200520072009



Supplementary Figure 45. Forest plot of vitamin E supplementation and stroke mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

100

Subgroup and Study, Year Events Total Events TotalRCTsde la Maza et al., 1995 5 37 4 37 0.1% 1.25 [0.36, 4.29]Stephens et al., 1996 - CHAOS 36 1,035 26 967 0.6% 1.29 [0.79, 2.13]Sano et al., 1997 - ADCS 12 85 12 84 0.3% 0.99 [0.47, 2.07]Gillilan et al., 1997 2 26 2 26 0.0% 1.00 [0.15, 6.57]Shoulson et al., 1998 - DATATOP 73 399 64 401 1.6% 1.15 [0.84, 1.56]GISSI-Prevenzione Investigators 1999 488 5,660 529 5,664 11.0% 0.92 [0.82, 1.04]Hoffman et al., 1999* 1 27 0 12 0.0% 1.39 [0.06, 31.93]Salonen et al., 2000 - ASAP 3 130 1 130 0.0% 3.00 [0.32, 28.47]Boaz et al., 2000 - SPACE 31 97 29 99 0.9% 1.09 [0.72, 1.66]de Gaetano et al., 2001 - PPP 72 2,231 68 2,264 1.4% 1.07 [0.78, 1.49]Desnuelle et al., 2001 - ALSRT 34 144 35 144 0.9% 0.97 [0.64, 1.47]de Waart et al., 2001 0 109 1 109 0.0% 0.33 [0.01, 8.09]Wluka et al., 2002 1 67 0 69 0.0% 3.09 [0.13, 74.50]Graat et al., 2002 3 164 5 153 0.1% 0.56 [0.14, 2.30]Hodis et al., 2002 - VEAPS 2 177 1 176 0.0% 1.99 [0.18, 21.73]Virtamo et al., 2003 - ATBC 868 7,286 851 7,287 19.3% 1.02 [0.93, 1.11]Collins et al., 2003† 1 26 1 26 0.0% 1.00 [0.07, 15.15]McNeil et al., 2004 - VECAT 11 595 7 598 0.2% 1.58 [0.62, 4.05]Manuel-Y-Keenoy et al., 2004 - DATOR 1 12 - 12 0.0% 3.00 [0.13, 67.06]Meydani et al., 2004 39 311 44 306 0.9% 0.87 [0.58, 1.30]Lonn et al., 2005 - HOPE & HOPE-TOO 799 4,761 801 4,780 19.0% 1.00 [0.92, 1.10]Graf et al., 2005 31 83 28 77 0.9% 1.03 [0.68, 1.54]Peterson et al., 2005 - ADCS 2 5 257 5 259 0.1% 1.01 [0.30, 3.44]Lee et al., 2005 - WHS 636 19,937 615 19,939 12.8% 1.03 [0.93, 1.15]Bairati et al., 2006 65 194 47 190 1.5% 1.35 [0.99, 1.86]Magliano et al., 2006 - MAVET 9 205 17 204 0.2% 0.53 [0.24, 1.15]Cook et al., 2007 - WACS 502 4,083 493 4,088 11.2% 1.02 [0.91, 1.15]Milman et al., 2008 11 726 12 708 0.2% 0.89 [0.40, 2.01]Lippman et al., 2009 - SELECT 358 8,863 382 8,910 7.6% 0.94 [0.82, 1.08]Sanyal et al., 2010 - PIVENS 1 84 0 83 0.0% 2.96 [0.12, 71.75]Manning et al., 2013 0 36 1 40 0% 0.37 [0.02, 8.79]Wang et al., 2014 - PHS II 397 3,659 406 3,653 9% 0.98 [0.86, 1.11]

Total (95% CI) 61,506 61,495 100% 1.00 [0.97, 1.04]Total events 4,497 4,487 Heterogeneity: Tau² = 0.00; Chi² = 17.22, df = 31 (P = 0.98); I² = 0%Test for overall effect: Z = 0.17 (P = 0.87)

0.05 0.2 1 5 20 Favours vitamin E Favours control




Study or Subgroupde la Maza et al., 1995Stephens et al., 1996 - CHAOSSano et al., 1997 - ADCSGillilan et al., 1997Shoulson et al., 1998 - DATATOPGISSI-Prevenzione Investigators 1999Hoffman et al., 1999Salonen et al., 2000 - ASAPBoaz et al., 2000 - SPACEde Gaetano et al., 2001 - PPPDesnuelle et al., 2001 - ALSRTde Waart et al., 2001Wluka et al., 2002Graat et al., 2002Hodis et al., 2002 - VEAPSVirtamo et al., 2003 - ATBCCollins et al., 2003McNeil et al., 2004 - VECATManuel-Y-Keenoy et al., 2004 - DATORMeydani et al., 2004Lonn et al., 2005 - HOPE & HOPE-TOOGraf et al., 2005Peterson et al., 2005 - ADCS 2Lee et al., 2005 - WHSBairati et al., 2006Magliano et al., 2006 - MAVETCook et al., 2007 - WACSMilman et al., 2008Lippman et al., 2009 - SELECTSanyal et al., 2010 - PIVENSManning et al., 2013Wang et al., 2014 - PHS II


Events5

3612

273

48813

317234

0132

8681

111

39799

315

63665

9502

11358

10

397

4497

Total37

10358526

3995660

27130

972231

144109

67164177

728626

59512

3114761

83257

19937194205

4083726

88638436

3659

61506

Events4

2612

264

52901

296835

1051

851170

44801

285

6154717

49312

38201

406

4487

Total37

9678426

4015664

12130

992264

144109

69153176

728726

59812

3064780

77259

19939190204

4088708

89108340

3653

61495

Weight0.1%0.6%0.3%0.0%1.6%

11.0%0.0%0.0%0.9%1.4%0.9%0.0%0.0%0.1%0.0%

19.3%0.0%0.2%0.0%0.9%

19.0%0.9%0.1%

12.8%1.5%0.2%

11.2%0.2%7.6%0.0%0.0%8.9%

100.0%

M-H, Random, 95% CI1.25 [0.36, 4.29]1.29 [0.79, 2.13]0.99 [0.47, 2.07]1.00 [0.15, 6.57]1.15 [0.84, 1.56]0.92 [0.82, 1.04]

1.39 [0.06, 31.93]3.00 [0.32, 28.47]

1.09 [0.72, 1.66]1.07 [0.78, 1.49]0.97 [0.64, 1.47]0.33 [0.01, 8.09]

3.09 [0.13, 74.50]0.56 [0.14, 2.30]

1.99 [0.18, 21.73]1.02 [0.93, 1.11]

1.00 [0.07, 15.15]1.58 [0.62, 4.05]

3.00 [0.13, 67.06]0.87 [0.58, 1.30]1.00 [0.92, 1.10]1.03 [0.68, 1.54]1.01 [0.30, 3.44]1.03 [0.93, 1.15]1.35 [0.99, 1.86]0.53 [0.24, 1.15]1.02 [0.91, 1.15]0.89 [0.40, 2.01]0.94 [0.82, 1.08]

2.96 [0.12, 71.75]0.37 [0.02, 8.79]0.98 [0.86, 1.11]

1.00 [0.97, 1.04]

Year19951996199719971998199919992000200020012001200120022002200220032003200420042004200520052005200520062006200720082009201020132014



Supplementary Figure 46. Forest plot of vitamin E supplementation and all-cause mortality. M-H, Manthel-Haenszel. *Hoffman et al., 1999 - data taken from Curtis et al., 2014; †Collins et al., 2003 - data taken from Curtis et al., 2014. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

101

Supplementary Figure 47. Funnel plot of vitamin E supplementation and and CVD and all-cause mortality outcomes: (A) total CVD risk, (B) MI risk, (C) stroke risk, (D) CVD mortality risk and (E) all-cause mortality risk. The vertical line represents the pooled effect estimate expressed as a RR. Dashed lines represent pseudo-95% confidence intervals (CI). The circles represent risk estimates for each study, and the horizontal lines represent standard errors of the RR. We were unable to test for funnel plot asymmetry for other CVD outcomes (<10 RCTs).

102

Random Effects

Comparison RCTs N Events RR (95% CIs) RR (95% CIs) I2 p -valueCVDTotal CVD 2 15,497 1,459 0.99 [0.90, 1.10] 0% 0.89Total CHD 1 8,171 999 1.04 [0.93, 1.17] N/A 0.49MI 2 15,497 545 0.96 [0.81, 1.14] 6% 0.62Stroke 2 15,497 525 0.92 [0.78, 1.09] 0% 0.33Total CVD mortality 2 15,497 646 1.07 [0.92, 1.25] 0% 0.35MI mortality 1 8,171 34 0.79 [0.40, 1.55] N/A 0.49Stroke mortality 1 8,171 33 0.83 [0.42, 1.65] N/A 0.60

ALL-CAUSE MORTALITY

All-cause mortality 4 16,004 1,819 1.02 [0.94, 1.11] 0% 0.63

Favours vitamin C/ Favours control/ Negative association Positive association


0.0 0.5 1.0 1.5 2.0

Supplementary Figure 48. Summary of the pooled effect estimates of RCTs assessing the relationship between vitamin C supplementation and CVD and all-cause mortality risk. RR, risk ratio; CIs, confidence intervals; RTCs, randomized clinical trials; CVD, cardiovascular disease; CHD, coronary heart disease; MI, myocardial infarction; N/A, not applicable. The diamonds represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% CIs, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsCook et al., 2007 - WACS 419 4,087 415 4,084 57.6% 1.01 [0.89, 1.15]Sesso et al., 2008 - PHS II* 309 3,673 316 3,653 42.4% 0.97 [0.84, 1.13]

Total (95% CI) 7,760 7,737 100% 0.99 [0.90, 1.10]Total events 728 731 Heterogeneity: Tau² = 0.00; Chi² = 0.13, df = 1 (P = 0.72); I² = 0%Test for overall effect: Z = 0.13 (P = 0.89) 0.5 0.7 1 1.5 2

Favours vitamin C Favours control

Vitamin C ControlWeight



Study or SubgroupCook et al., 2007 - WACSSesso et al., 2008 - PHS II


Events419309

728

Total40873673

7760

Events415316

731

Total40843653

7737

Weight57.6%42.4%

100.0%

M-H, Random, 95% CI1.01 [0.89, 1.15]0.97 [0.84, 1.13]

0.99 [0.90, 1.10]

Vitamin C Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.5 0.7 1 1.5 2Favours vitamin C Favours control

Supplementary Figure 49. Forest plot of vitamin C supplementation and total CVD risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. *Sesso et al., 2008 - Data taken from most updated trial data from Wang et al., 2014. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

103

Subgroup and Study, Year Events Total Events TotalRCTsCook et al., 2007 - WACS 510 4,087 489 4,084 100% 1.04 [0.93, 1.17]

Total (95% CI) 4,087 4,084 100% 1.04 [0.93, 1.17]Total events 510 489 Heterogeneity: Not applicable 0.5 0.7 1 1.5 2Test for overall effect: Z = 0.70 (P = 0.49) Favours vitamin C Favours control



Risk RatioM-H, Random, 95% CI in CHD risk



Events510

510

Total4087

4087

Events489

489

Total4084

4084

Weight100.0%

100.0%

M-H, Random, 95% CI1.04 [0.93, 1.17]

1.04 [0.93, 1.17]

Year2007



Supplementary Figure 50. Forest plot of vitamin C supplementation and total CHD risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.









Events140127

267

Total40873673

7760

Events134144

278

Total40843653

7737

Weight50.3%49.7%

100.0%

M-H, Random, 95% CI1.04 [0.83, 1.32]0.88 [0.69, 1.11]

0.96 [0.81, 1.14]

Year20072008



Supplementary Figure 51. Forest plot of vitamin C supplementation and MI risk. M-H, Manthel-Haenszel, MI, myocardial infarction. *Sesso et al., 2008 - Data taken from most updated trial data from Wang et al., 2014. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.


Total (95% CI) 7,760 7,737 100% 0.92 [0.78, 1.09]Total events 252 273Heterogeneity: Tau² = 0.00; Chi² = 0.77, df = 1 (P = 0.38); I² = 0%Test for overall effect: Z = 0.97 (P = 0.33) 0.5 0.7 1 1.5 2







Events138114

252

Total40873673

7760

Events160113

273

Total40843653

7737

Weight56.8%43.2%

100.0%

M-H, Random, 95% CI0.86 [0.69, 1.08]1.00 [0.78, 1.30]

0.92 [0.78, 1.09]

Year20072008



Supplementary Figure 52. Forest plot of vitamin C supplementation and stroke risk. M-H, Manthel-Haenszel. *Sesso et al., 2008 - Data taken from most updated trial data from Wang et al., 2014. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

104









Events206129

335

Total40873673

7760

Events189122

311

Total40843653

7737

Weight61.5%38.5%

100.0%

M-H, Random, 95% CI1.09 [0.90, 1.32]1.05 [0.82, 1.34]

1.07 [0.92, 1.25]

Year20072008



Supplementary Figure 53. Forest plot of vitamin C supplementation and CVD mortality risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. *Sesso et al., 2008 - Data taken from most updated trial data from Wang et al., 2014. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.


Total (95% CI) 4,087 4,084 100% 0.79 [0.40, 1.55]Total events 15 19 Heterogeneity: Not applicable 0.02 0.1 1 10 50Test for overall effect: Z = 0.69 (P = 0.49) Favours vitamin C Favours control






Events15

15

Total4087

4087

Events19

19

Total4084

4084

Weight100.0%

100.0%

M-H, Random, 95% CI0.79 [0.40, 1.55]

0.79 [0.40, 1.55]

Year2007


0.02 0.1 1 10 50Favours vitamin C Favours control

Supplementary Figure 54. Forest plot of vitamin C supplementation and MI mortality risk. M-H, Manthel-Haenszel, MI, myocardial infarction. *Sesso et al., 2008 - Data taken from most updated trial data from Wang et al., 2014. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

105


Total (95% CI) 4,087 4,084 100% 0.83 [0.42, 1.65]Total events 15 18 Heterogeneity: Not applicableTest for overall effect: Z = 0.52 (P = 0.60) 0.02 0.1 1 10 50







Events15

15

Total4087

4087

Events18

18

Total4084

4084

Weight100.0%

100.0%

M-H, Random, 95% CI0.83 [0.42, 1.65]

0.83 [0.42, 1.65]

Year2007


0.02 0.1 1 10 50Favours vitamin C Favours control

Supplementary Figure 55. Forest plot of vitamin C supplementation and stroke mortality risk. M-H, Manthel-Haenszel. *Sesso et al., 2008 - Data taken from most updated trial data from Wang et al., 2014. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsSalonen et al., 2000- ASAP 1 130 1 130 0.1% 1.00 [0.06, 15.82]Correa et al., 2000 2 130 0 117 0.1% 4.50 [0.22, 92.86]Cook et al., 2007 - WACS 504 4,087 491 4,084 55.0% 1.03 [0.91, 1.15]Wang et al., 2014 - PHS II 414 3,673 406 3,653 44.8% 1.01 [0.89, 1.15]

Total (95% CI) 8,020 7,984 100% 1.02 [0.94, 1.11]Total events 921 898 Heterogeneity: Tau² = 0.00; Chi² = 0.94, df = 3 (P = 0.82); I² = 0%

0.1 0.2 0.5 1 2 5 10 Favours vitamin C Favours control




Study or SubgroupSalonen et al., 2000- ASAPCorrea et al., 2000Cook et al., 2007 - WACSWang et al., 2014 - PHS II


Events12

504414

921

Total130130

40873673

8020

Events10

491406

898

Total130117

40843653

7984

Weight0.1%0.1%

55.0%44.8%

100.0%

M-H, Random, 95% CI1.00 [0.06, 15.82]4.50 [0.22, 92.86]

1.03 [0.91, 1.15]1.01 [0.89, 1.15]

1.02 [0.94, 1.11]

Year2000200020072014


0.1 0.2 0.5 1 2 5 10Favours vitamin C Favours control

Supplementary Figure 56. Forest plot of vitamin C supplementation and all-cause mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

106

Random Effects


CVDTotal CVD 2 18,770 2,329 1.04 [0.96, 1.12] 0% 0.34

Total CHD 1 1,004 122 1.06 [0.76, 1.48] N/A 0.73

MI 2 1,085 87 0.93 [0.62, 1.39] 0% 0.72

Stroke 2 18,770 249 0.89 [0.70, 1.14] 0% 0.36

Total CVD mortality 3 18,851 346 1.00 [0.66, 1.51] 48% 1.00

MI mortality 2 1,184 18 1.21 [0.49, 2.99] 0% 0.68

Stroke mortality 1 17,766 17 1.13 [0.44, 2.93] N/A 0.80

ALL-CAUSE MORTALITY

All-cause mortality 4 19,373 992 0.99 [0.88, 1.12] 0% 0.87


Favours selenium/ Favours control/ Negative association Positive association

0.0 0.5 1.0 1.5 2.0

Supplementary Figure 57. Summary of the pooled effect estimates of RCTs assessing the relationship between selenium supplementation and CVD and all cause mortality risk. RR, risk ratio; CIs, confidence intervals; RTCs, randomized clinical trials; CVD, cardiovascular disease; CHD, coronary heart disease; MI, myocardial infarction; N/A, not applicable. The diamonds represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% CIs, using the Manthel-Haenszel method with random effects.

Subgroup and Study, Year Events Total Events TotalRCTsStranges et al., 2006 - NPC 103 504 96 500 9.3% 1.06 [0.83, 1.37]Lippman et al., 2009 - SELECT 1,080 8,856 1,050 8,910 90.7% 1.03 [0.96, 1.12]

Total (95% CI) 9,360 9,410 100% 1.04 [0.96, 1.12]Total events 1,183 1,146Heterogeneity: Tau² = 0.00; Chi² = 0.04, df = 1 (P = 0.83); I² = 0%Test for overall effect: Z = 0.95 (P = 0.34) 0.7 0.85 1 1.2 1.5

Favours selenium Favours control

Selenium ControlWeight



Study or SubgroupStranges et al., 2006 - NPCLippman et al., 2009 - SELECT


Events103

1080

1183

Total504

8856

9360

Events96

1050

1146

Total500

8910

9410

Weight9.3%

90.7%

100.0%

M-H, Random, 95% CI1.06 [0.83, 1.37]1.03 [0.96, 1.12]

1.04 [0.96, 1.12]

Year20062009

Selenium Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.7 0.85 1 1.2 1.5Favours selenium Favours control

Supplementary Figure 58. Forest plot of selenium supplementation and total CVD risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

107

Subgroup and Study, Year Events Total Events TotalRCTsStranges et al., 2006 - NPC 63 504 59 500 100.0% 1.06 [0.76, 1.48]

Total (95% CI) 504 500 100% 1.06 [0.76, 1.48]Total events 63 59Heterogeneity: Not applicableTest for overall effect: Z = 0.34 (P = 0.73) 0.2 0.5 1 2 5





Study or SubgroupStranges et al., 2006 - NPC


Events63

63

Total504

504

Events59

59

Total500

500

Weight100.0%

100.0%

M-H, Random, 95% CI1.06 [0.76, 1.48]

1.06 [0.76, 1.48]

Year2006


0.2 0.5 1 2 5Favours selenium Favours control

Supplementary Figure 59. Forest plot of selenium supplementation and total CHD risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsKorpela et al., 1989 1 40 2 41 2.9% 0.51 [0.05, 5.43]Stranges et al., 2006 - NPC 41 504 43 500 97.1% 0.95 [0.63, 1.42]






Study or SubgroupKorpela et al., 1989Stranges et al., 2006 - NPC


Events1

41

42

Total40

504

544

Events2

43

45

Total41

500

541

Weight2.9%

97.1%

100.0%

M-H, Random, 95% CI0.51 [0.05, 5.43]0.95 [0.63, 1.42]

0.93 [0.62, 1.39]

Year19892006



Supplementary Figure 60. Forest plot of selenium supplementation and MI risk. M-H, Manthel-Haenszel, MI, myocardial infarction. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsStranges et al., 2006 - NPC 35 504 32 500 28.2% 1.09 [0.68, 1.72]Lippman et al., 2009 - SELECT 82 8,856 100 8,910 71.8% 0.82 [0.62, 1.10]






Study or SubgroupStranges et al., 2006 - NPCLippman et al., 2009 - SELECT


Events3582

117

Total504

8856

9360

Events32

100

132

Total500

8910

9410

Weight28.2%71.8%

100.0%

M-H, Random, 95% CI1.09 [0.68, 1.72]0.82 [0.62, 1.10]

0.89 [0.70, 1.14]

Year20062009


0.1 0.2 0.5 1 2 5 10Favours selenium Favours control

Supplementary Figure 61. Forest plot of selenium supplementation and stroke risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

108

Subgroup and Study, Year Events Total Events TotalRCTsKorpela et al., 1989 0 40 4 41 2.0% 0.11 [0.01, 2.05]Stranges et al., 2006 - NPC 40 504 31 500 38.9% 1.28 [0.81, 2.01]Lippman et al., 2009 - SELECT 129 8,856 142 8,910 59.1% 0.91 [0.72, 1.16]






Study or SubgroupKorpela et al., 1989Stranges et al., 2006 - NPCLippman et al., 2009 - SELECT


Events0

40129

169

Total40

5048856

9400

Events4

31142

177

Total41

5008910

9451

Weight2.0%

38.9%59.1%

100.0%

M-H, Random, 95% CI0.11 [0.01, 2.05]1.28 [0.81, 2.01]0.91 [0.72, 1.16]

1.00 [0.66, 1.51]

Year198920062009


0.005 0.1 1 10 200Favours Selenium Favours control

Supplementary Figure 62. Forest plot of selenium supplementation and CVD mortality risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsLimburg et al., 2005 1 90 0 90 8.1% 3.00 [0.12, 72.68]Stranges et al., 2006 - NPC 9 504 8 500 91.9% 1.12 [0.43, 2.87]





Risk RatioM-H, Random, 95% CI in MI mortality

Study or SubgroupLimburg et al., 2005Stranges et al., 2006 - NPC


Events19

10

Total90

504

594

Events08

8

Total90

500

590

Weight8.1%

91.9%

100.0%

M-H, Random, 95% CI3.00 [0.12, 72.68]1.12 [0.43, 2.87]

1.21 [0.49, 2.99]

Year20052006



Supplementary Figure 63. Forest plot of selenium supplementation and MI mortality risk. M-H, Manthel-Haenszel, MI, myocardial infarction. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsLippman et al., 2009 - SELECT 9 8,856 8 8,910 100.0% 1.13 [0.44, 2.93]

Total (95% CI) 8,856 8,910 100% 1.13 [0.44, 2.93]Total events 9 8Heterogeneity: Not applicableTest for overall effect: Z = 0.26 (P = 0.80) 0.01 0.1 1 10 100





Study or SubgroupLippman et al., 2009 - SELECT


Events9

9

Total8856

8856

Events8

8

Total8910

8910

Weight100.0%

100.0%

M-H, Random, 95% CI1.13 [0.44, 2.93]

1.13 [0.44, 2.93]

Year2009



Supplementary Figure 64. Forest plot of selenium supplementation and stroke mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

109

Subgroup and Study, Year Events Total Events TotalRCTsLimburg et al., 2005 1 90 0 90 0.1% 3.00 [0.12, 72.68]Stranges et al., 2006 - NPC 110 504 111 500 26.0% 0.98 [0.78, 1.24]Lippman et al., 2009 - SELECT 378 8856 382 8910 72.9% 1.00 [0.87, 1.14]Marshall et al., 2011 4 212 6 211 0.9% 0.66 [0.19, 2.32]


Favours Selenium Favours control



Risk RatioM-H, Random, 95% CI in all-cause mortality

Study or SubgroupLImburg et al., 2005Stranges et al., 2006 - NPCLippman et al., 2009 - SELECTMarshall et al., 2011


Events1

110378

4

493

Total90

5048856212

9662

Events0

111382

6

499

Total90

5008910211

9711

Weight0.1%

26.0%72.9%0.9%

100.0%

M-H, Random, 95% CI3.00 [0.12, 72.68]0.98 [0.78, 1.24]1.00 [0.87, 1.14]0.66 [0.19, 2.32]

0.99 [0.88, 1.12]

Year2005200620092011


0.2 0.5 1 2 5Favours Selenium Favours control

Supplementary Figure 65. Forest plot of selenium supplementation and all-cause mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Random Effects

Comparison RCTs N Events RR (95% CIs) RR (95% CIs) I2 p -valueCVDTotal CVD 9 39,756 6,888 0.98 [0.93, 1.04] 29% 0.58Total CHD 5 20,886 3,197 1.04 [0.96, 1.14] 12% 0.32MI 13 44,285 2,875 1.00 [0.93, 1.07] 0% 1.00Stroke 12 43,339 2,067 0.90 [0.81, 1.00] 16% 0.04

Total CVD mortality 5 33,693 2,642 0.98 [0.87, 1.11] 52% 0.74Total CHD mortality 3 13,267 964 1.09 [0.97, 1.23] 0% 0.16

MI mortality 2 13,944 482 1.11 [0.93, 1.32] 0% 0.26


ALL-CAUSE MORTALITY


Favours B-complex/ Favours control/ Negative association Positive association


0.5 1.0 1.5

Supplementary Figure 66. Summary of the pooled effect estimates of RCTs assessing the relationship between vitamin B complex supplementation and CVD and all cause mortality risk. RR, risk ratio; CIs, confidence intervals; RTCs, randomized clinical trials; CVD, cardiovascular disease; CHD, coronary heart disease; MI, myocardial infarction; N/A, not applicable. The diamonds represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% CIs, using the Manthel-Haenszel method with random effects model.

110

Subgroup and Study, Year Events Total Events TotalRCTLonn et al., 2006 - HOPE-TOO 519 2,758 547 2,764 17.8% 0.95 [0.85, 1.06]Bonaa et al., 2006 - NORVIT* 201 937 172 943 8.5% 1.18 [0.98, 1.41]Albert et al., 2008 - WAFACS 205 2,721 211 2,721 8.3% 0.97 [0.81, 1.17]Hodis et al., 2009 - BVAIT† 9 254 11 252 0.5% 0.81 [0.34, 1.93]Heinz et al., 2010 83 327 98 323 5.0% 0.84 [0.65, 1.07]VITATOPS Trial Study Group 2010 616 4,089 678 4,075 19.4% 0.91 [0.82, 1.00]Galan et al., 2010 - SU.FOL.OM3 75 1,242 82 1,259 3.5% 0.93 [0.68, 1.26]Armitage et al., 2010 - SEARCH 1,537 6,033 1,493 6,031 29.7% 1.03 [0.97, 1.09]Van Dijk et al., 2015 - B-PROOF 181 1,516 170 1,511 7.50% 1.06 [0.87, 1.29]


Favours B-complex Favours control

B-complex ControlWeight



Study or SubgroupLonn et al., 2006 - HOPE 2Bonaa et al., 2006 - NORVITAlbert et al., 2008 - WAFACSHodis et al., 2009 - BVAITHeinz et al., 2010VITATOPS Trial Study Group 2010Galan et al., 2010 - SU.FOL.OM3Armitage et al., 2010 - SEARCHVan Dijk et al., 2015 - B-PROOF


Events519201205

983

61675

1537181

3426

Total2758

9372721

254327

4089124260331516

19877

Events547172211

1198

67882

1493170

3462

Total2764

9432721

252323

4075125960311511

19879

Weight17.8%

8.5%8.3%0.5%5.0%

19.4%3.5%

29.7%7.5%

100.0%

M-H, Random, 95% CI0.95 [0.85, 1.06]1.18 [0.98, 1.41]0.97 [0.81, 1.17]0.81 [0.34, 1.93]0.84 [0.65, 1.07]0.91 [0.82, 1.00]0.93 [0.68, 1.26]1.03 [0.97, 1.09]1.06 [0.87, 1.29]

0.98 [0.93, 1.04]

Year200620062008200920102010201020102015

B-Complex Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.5 0.7 1 1.5 2Favours B-Complex Favours Control

Supplementary Figure 67. Forest plot of vitamin B complex supplementation and total CVD risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. *Bonaa et al., 2006 - folic acid, B6 and B12 vs placebo. †Hodis et al., 2009 - reported number of participants with at least one event. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTLange et al., 2004 53 316 35 320 4.4% 1.53 [1.03, 2.28]Albert et al., 2008 - WAFACS 283 2,721 280 2,721 23.7% 1.01 [0.86, 1.18]Imasa et al., 2009 15 118 13 125 1.5% 1.22 [0.61, 2.46]Galan et al., 2010 - SU.FOL.OM3 49 1,242 55 1,259 4.9% 0.90 [0.62, 1.32]Armitage et al., 2010 - SEARCH 1,229 6,033 1,185 6,031 65.6% 1.04 [0.97, 1.11]






Study or SubgroupLange et al., 2004Albert et al., 2008 - WAFACSImasa et al., 2009Galan et al., 2010 - SU.FOL.OM3Armitage et al., 2010 - SEARCH


Events53

2831549

1229

1629

Total316

2721118

12426033

10430

Events35

2801355

1185

1568

Total320

2721125

12596031

10456

Weight4.4%

23.7%1.5%4.9%

65.6%

100.0%

M-H, Random, 95% CI1.53 [1.03, 2.28]1.01 [0.86, 1.18]1.22 [0.61, 2.46]0.90 [0.62, 1.32]1.04 [0.97, 1.11]

1.04 [0.96, 1.14]

Year20042008200920102010


0.5 0.7 1 1.5 2Favours B-complex Favours control

Supplementary Figure 68. Forest plot of vitamin B complex supplementation and total CHD risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

111

Subgroup and Study, Year Events Total Events TotalRCTsSchnyder et al., 2002 - The Swiss Heart Study 7 272 12 281 0.6% 0.60 [0.24, 1.51]Lange et al., 2004 3 316 2 320 0.2% 1.52 [0.26, 9.03]Bonaa et al., 2006 - NORVIT* 182 937 153 943 12.7% 1.20 [0.99, 1.45]Lonn et al., 2006 - HOPE-TOO 341 2,758 349 2,764 24.8% 0.98 [0.85, 1.13]Jamison et al., 2007 - HOST 129 1,032 150 1,024 10.1% 0.85 [0.69, 1.06]Albert et al., 2008 - WAFACS 65 2,721 74 2,721 4.5% 0.88 [0.63, 1.22]Ebbing et al., 2008 - WENBIT† 59 772 58 780 4.0% 1.03 [0.73, 1.46]VITATOPS Trial Study Group 2010 118 4,089 114 4,075 7.5% 1.03 [0.80, 1.33]Galan et al., 2010 - SU.FOL.OM3 28 1,242 32 1,259 1.9% 0.89 [0.54, 1.46]Armitage et al., 2010 - SEARCH 431 6,033 429 6,031 29.2% 1.00 [0.88, 1.14]House et al., 2010 - DIVINe 8 119 4 119 0.4% 2.00 [0.62, 6.46]Heinz et al., 2010 20 327 19 323 1.3% 1.04 [0.57, 1.91]Van Dijk et al., 2015 - B-PROOF 45 1,516 43 1,511 2.9% 1.04 [0.69, 1.57]

Total (95% CI) 22,134 22,151 100% 1.00 [0.93, 1.07]Total events 1,436 1,439Heterogeneity: Tau² = 0.00; Chi² = 9.06, df = 12 (P = 0.70); I² = 0%Test for overall effect: Z = 0.00 (P = 1.00) 0.5 0.7 1 1.5 2





Study or SubgroupSchnyder et al., 2002 - The Swiss Heart StudyLange et al., 2004Bonaa et al., 2006 - NORVITLonn et al., 2006 - HOPE 2Jamison et al., 2007 - HOSTAlbert et al., 2008 - WAFACSEbbing et al., 2008 - WENBITVITATOPS Trial Study Group 2010Galan et al., 2010 - SU.FOL.OM3Armitage et al., 2010 - SEARCHHouse et al., 2010 - DIVINeHeinz et al., 2010Van Dijk et al., 2015 - B-PROOF


Events73

182341129

6559

11828

4318

2045

1436

Total272316937

275810322721

772408912426033

119327

1516

22134

Events12

2153349150

7458

11432

4294

1943

1439

Total281320943

276410242721

780407512596031

119323

1511

22151

Weight0.6%0.2%

12.7%24.8%10.1%

4.5%4.0%7.5%1.9%

29.2%0.4%1.3%2.9%

100.0%

M-H, Random, 95% CI0.60 [0.24, 1.51]1.52 [0.26, 9.03]1.20 [0.99, 1.45]0.98 [0.85, 1.13]0.85 [0.69, 1.06]0.88 [0.63, 1.22]1.03 [0.73, 1.46]1.03 [0.80, 1.33]0.89 [0.54, 1.46]1.00 [0.88, 1.14]2.00 [0.62, 6.46]1.04 [0.57, 1.91]1.04 [0.69, 1.57]

1.00 [0.93, 1.07]

Year2002200420062006200720082008201020102010201020102015



Supplementary Figure 69. Forest plot of vitamin B complex supplementation and MI risk. M-H, Manthel-Haenszel, MI, myocardial infarction. *Bonaa et al., 2006 - folic acid, B6 and B12 vs placebo. †Ebbing et al., 2008 - folic acid, B6 and B12 vs placebo. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsBonaa et al., 2006 - NORVIT* 21 937 27 943 3.3% 0.78 [0.45, 1.37]Jamison et al., 2007 - HOST 37 1,032 41 1,024 5.2% 0.90 [0.58, 1.38]Ebbing et al., 2008 - WENBIT† 11 772 19 780 2.0% 0.58 [0.28, 1.22]Albert et al., 2008 - WAFACS 79 2,721 69 2,721 9.1% 1.14 [0.83, 1.57]Saposnik et al., 2009 - HOPE 2 111 2,758 147 2,764 14.2% 0.76 [0.59, 0.96]Imasa et al., 2009 0 118 1 125 0.1% 0.35 [0.01, 8.58]VITATOPS Trial Study Group 2010 360 4,089 388 4,075 28.7% 0.92 [0.81, 1.06]Heinz et al., 2010 11 327 15 323 1.8% 0.72 [0.34, 1.55]Galan et al., 2010 - SU.FOL.OM3 35 1,242 48 1,259 5.4% 0.74 [0.48, 1.13]Armitage et al., 2010 - SEARCH 269 6,033 265 6,031 23.3% 1.01 [0.86, 1.20]House et al., 2010 - DIVINe 6 119 1 119 0.2% 6.00 [0.73, 49.08]Van Dijk et al., 2015 - B-PROOF 46 1,516 60 1,511 7% 0.76 [0.52, 1.11]

Total (95% CI) 21,664 21,675 100% 0.90 [0.81, 1.00]Total events 986 1,081 Heterogeneity: Tau² = 0.01; Chi² = 13.13, df = 11 (P = 0.28); I² = 16%Test for overall effect: Z = 2.01 (P = 0.04) 0.2 0.5 1 2 5





Study or SubgroupBonaa et al., 2006 - NORVITJamison et al., 2007 - HOSTEbbing et al., 2008 - WENBITAlbert et al., 2008 - WAFACSSaposnik et al., 2009 - HOPE 2Imasa et al., 2009VITATOPS Trial Study Group 2010Heinz et al., 2010Galan et al., 2010 - SU.FOL.OM3Armitage et al., 2010 - SEARCHHouse et al., 2010 - DIVINeVan Dijk et al., 2015 - B-PROOF


Events21371179

1110

3601135

2696

46

986

Total937

1032772

27212758

1184089

32712426033

1191516

21664

Events27411969

1471

3881548

2651

60

1081

Total943

1024780

27212764

1254075

32312596031

1191511

21675

Weight3.3%5.2%2.0%9.1%

14.2%0.1%

28.7%1.8%5.4%

23.3%0.2%6.8%

100.0%

M-H, Random, 95% CI0.78 [0.45, 1.37]0.90 [0.58, 1.38]0.58 [0.28, 1.22]1.14 [0.83, 1.57]0.76 [0.59, 0.96]0.35 [0.01, 8.58]0.92 [0.81, 1.06]0.72 [0.34, 1.55]0.74 [0.48, 1.13]1.01 [0.86, 1.20]

6.00 [0.73, 49.08]0.76 [0.52, 1.11]

0.90 [0.81, 1.00]

Year200620072008200820092009201020102010201020102015


0.2 0.5 1 2 5Favours B-Complex Favours Control

Supplementary Figure 70. Forest plot of vitamin B complex supplementation and stroke risk. M-H, Manthel-Haenszel. *Bonaa et al., 2006 - folic acid, B6 and B12 vs placebo. †Ebbing et al., 2008 - folic acid, B6 and B12 vs placebo. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

112

Subgroup and Study, Year Events Total Events TotalRCTsLonn et al., 2006 - HOPE-TOO 276 2,758 291 2,764 25.0% 0.95 [0.81, 1.11]Albert et al., 2008 - WAFACS 96 2,721 94 2,721 12.9% 1.02 [0.77, 1.35]VITATOPS Trial Study Group 2010 328 4,089 380 4,075 27.1% 0.86 [0.75, 0.99]Galan et al., 2010 - SU.FOL.OM3 26 1,242 14 1,259 3.2% 1.88 [0.99, 3.59]Armitage et al., 2010 - SEARCH 578 6,033 559 6,031 31.7% 1.03 [0.93, 1.15]






Study or SubgroupLonn et al., 2006 - HOPE 2Albert et al., 2008 - WAFACSVITATOPS 2010Galan et al., 2010 - SU.FOL.OM3Armitage et al., 2010 - SEARCH


Events276

96328

26578

1304

Total27582721408912426033

16843

Events291

94380

14559

1338

Total27642721407512596031

16850

Weight25.0%12.9%27.1%

3.2%31.7%

100.0%

M-H, Random, 95% CI0.95 [0.81, 1.11]1.02 [0.77, 1.35]0.86 [0.75, 0.99]1.88 [0.99, 3.59]1.03 [0.93, 1.15]

0.98 [0.87, 1.11]

Year20062008201020102010



Supplementary Figure 71. Forest plot of vitamin B complex supplementation and CVD mortality risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsSchnyder et al., 2002 - The Swiss Heart Study 3 272 6 281 0.8% 0.52 [0.13, 2.04]Heinz et al., 2010 37 327 32 323 7.4% 1.14 [0.73, 1.79]Armitage et al., 2010 - SEARCH 463 6,033 423 6,031 91.8% 1.09 [0.96, 1.24]

Total (95% CI) 6,632 6,635 100% 1.09 [0.97, 1.23]Total events 503 461 Heterogeneity: Tau² = 0.00; Chi² = 1.18, df = 2 (P = 0.56); I² = 0%Test for overall effect: Z = 1.41 (P = 0.16)

0.01 0.1 1 10 100 Favours B-complex Favours control



Risk RatioM-H, Random, 95%CI in CHD mortality risk

Study or SubgroupSchnyder et al., 2002 - The Swiss Heart StudyHeinz et al., 2010Armitage et al., 2010 - SEARCH


Events3

37463

503

Total272327

6033

6632

Events6

32423

461

Total281323

6031

6635

Weight0.8%7.4%

91.8%

100.0%

M-H, Random, 95% CI0.52 [0.13, 2.04]1.14 [0.73, 1.79]1.09 [0.96, 1.24]

1.09 [0.97, 1.23]

Year200220102010


0.01 0.1 1 10 100Favours B-complex Favours control

Supplementary Figure 72. Forest plot of vitamin B complex supplementation and CHD mortality risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsBonaa et al., 2006 - NORVIT* 68 937 59 943 27.1% 1.16 [0.83, 1.62]Armitage et al., 2010 - SEARCH 185 6,033 170 6,031 72.9% 1.09 [0.89, 1.34]






Study or SubgroupBonaa et al., 2006 - NORVITArmitage et al., 2010 - SEARCH


Events68

185

253

Total937

6033

6970

Events59

170

229

Total943

6031

6974

Weight27.1%72.9%

100.0%

M-H, Random, 95% CI1.16 [0.83, 1.62]1.09 [0.89, 1.34]

1.11 [0.93, 1.32]

Year20062010


0.5 0.7 1 1.5 2Favours B-complex Favours control

Supplementary Figure 73. Forest plot of vitamin B complex supplementation and MI mortality risk. M-H, Manthel-Haenszel, MI, myocardial infarction. *Bonaa et al., 2006 - folic acid, B6 and B12 vs placebo. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

113

Subgroup and Study, Year Events Total Events TotalRCTsSaposnik et al., 2009 - HOPE 2 27 2,758 30 2,764 31.5% 0.90 [0.54, 1.51]Armitage et al., 2010 - SEARCH 59 6,033 65 6,031 68.5% 0.91 [0.64, 1.29]

Total (95% CI) 8,791 8,795 100% 0.91 [0.68, 1.21]Total events 86 95 Heterogeneity: Tau² = 0.00; Chi² = 0.00, df = 1 (P = 0.98); I² = 0%Test for overall effect: Z = 0.67 (P = 0.50) 0.1 0.2 0.5 1 2 5 10





Study or SubgroupSaposnik et al., 2009 - HOPE 2Armitage et al., 2010 - SEARCH


Events2759

86

Total27586033

8791

Events3065

95

Total27646031

8795

Weight31.5%68.5%

100.0%

M-H, Random, 95% CI0.90 [0.54, 1.51]0.91 [0.64, 1.29]

0.91 [0.68, 1.21]

Year20092010


0.1 0.2 0.5 1 2 5 10Favours B-complex Favours control

Supplementary Figure 74. Forest plot of vitamin B complex supplementation and stroke mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsSchnyder et al., 2002 - The Swiss Heart Study 4 272 8 281 0.1% 0.52 [0.16, 1.70]Lange et al., 2004 1 316 1 320 0.0% 1.01 [0.06, 16.12]Lonn et al., 2006 - HOPE-TOO 470 2,758 475 2,764 14.5% 0.99 [0.88, 1.11]Bonaa et al., 2006 - NORVIT* 104 937 89 943 2.7% 1.18 [0.90, 1.54]Jamison et al., 2007 - HOST 448 1,032 436 1,024 19.7% 1.02 [0.92, 1.13]Ebbing et al., 2008 - WENBIT† 35 772 30 780 0.9% 1.18 [0.73, 1.90]Albert et al., 2008 - WAFACS 250 2,721 256 2,721 7.1% 0.98 [0.83, 1.15]Hodis et al., 2009 - BVAIT 0 254 2 252 0.0% 0.20 [0.01, 4.11]Imasa et al., 2009 22 118 20 125 0.6% 1.17 [0.67, 2.02]Heinz et al., 2010 102 327 92 323 3.5% 1.10 [0.86, 1.39]Armitage et al., 2010 - SEARCH 983 6,033 951 6,031 29.3% 1.03 [0.95, 1.12]Galan et al., 2010 - SU.FOL.OM3 72 1,242 45 1,259 1.5% 1.62 [1.13, 2.33]VITATOPS Trial Study Group 2010 614 4,089 633 4,075 18.7% 0.97 [0.87, 1.07]House et al., 2010 - DIVINe 7 119 6 119 0.2% 1.17 [0.40, 3.37]Van Wijngaarden et al., 2014 - B-PROOF 37 1,516 42 1,511 1.00% 0.88 [0.57, 1.36]Wang et al., 2015 4 195 6 195 0% 0.67 [0.19, 2.33]

Total (95% CI) 22,701 22,723 100% 1.02 [0.97, 1.06]Total events 3,153 3,092 Heterogeneity: Tau² = 0.00; Chi² = 13.22, df = 15 (P = 0.58); I² = 0%Test for overall effect: Z = 0.82 (P = 0.41) 0.1 0.2 0.5 1 2 5 10





Study or SubgroupSchnyder et al., 2002 - The Swiss Heart StudyLange et al., 2004Lonn et al., 2006 - HOPE 2Bonaa et al., 2006 - NORVITJamison et al., 2007 - HOSTEbbing et al., 2008 - WENBITAlbert et al., 2008 - WAFACSHodis et al., 2009Imasa et al., 2009Heinz et al., 2010Armitage et al., 2010 - SEARCHGalan et al., 2010 - SU.FOL.OM3VITATOPS Trial Study Group 2010House et al., 2010 - DIVINeVan Wijngaarden et al., 2014 - B-PROOFWang et al., 2015


Events41

470104448

35250

022

102983

72614

737

4

3153

Total272316

2758937

1032772

2721254118327

603312424089

1191516

195

22701

Events81

47589

43630

2562

2092

95145

6336

426

3092

Total281320

2764943

1024780

2721252125323

603112594075

1191511

195

22723

Weight0.1%0.0%

14.5%2.7%

19.7%0.9%7.1%0.0%0.6%3.5%

29.3%1.5%

18.7%0.2%1.0%0.1%

100.0%

M-H, Random, 95% CI0.52 [0.16, 1.70]

1.01 [0.06, 16.12]0.99 [0.88, 1.11]1.18 [0.90, 1.54]1.02 [0.92, 1.13]1.18 [0.73, 1.90]0.98 [0.83, 1.15]0.20 [0.01, 4.11]1.17 [0.67, 2.02]1.10 [0.86, 1.39]1.03 [0.95, 1.12]1.62 [1.13, 2.33]0.97 [0.87, 1.07]1.17 [0.40, 3.37]0.88 [0.57, 1.36]0.67 [0.19, 2.33]

1.02 [0.97, 1.06]

Year2002200420062006200720082008200920092010201020102010201020142015


0.1 0.2 0.5 1 2 5 10Favours B-complex Favours control

Supplementary Figure 75. Forest plot of vitamin B complex supplementation and all-cause mortality risk. M-H, Manthel-Haenszel. *Bonaa et al., 2006 - folic acid, B6 and B12 vs placebo. †Ebbing et al., 2008 - folic acid, B6 and B12 vs placebo. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

114

Supplementary Figure 76. Funnel plot of vitamin B complex supplementation and CVD and all-cause mortality outcomes: (A) MI risk, (B) stroke risk, and (C) all-cause mortality risk. The vertical line represents the pooled effect estimate expressed as a RR. Dashed lines represent pseudo-95% confidence intervals (CI). The circles represent risk estimates for each study, and the horizontal lines represent standard errors of the RR. We were unable to test for funnel plot asymmetry for other CVD outcomes (<10 RCTs).

115

Random Effects

Comparison RCTs N Events RR (95% CIs) RR (95% CIs) I2 p -valueCVD

Total CVD 5 21,567 960 0.83 [0.73, 0.93] 0% <0.01

Total CHD 2 2,197 77 1.47 [0.95, 2.28] 0% 0.08MI 6 24,210 106 1.21 [0.78, 1.88] 12% 0.41Stroke 7 24,525 694 0.80 [0.69, 0.93] 0% <0.01Total CVD mortality 5 22,468 188 0.89 [0.68, 1.17] 0% 0.41MI mortality 2 20,985 13 1.17 [0.39, 3.49] 0% 0.77


ALL-CAUSE MORTALITY All-cause mortality 10 25,580 877 0.87 [0.72, 1.05] 17% 0.14


Favours folic acid/ Favours control/ Negative association Positive association

0.0 0.5 1.0 1.5 2.0

Supplementary Figure 77. Summary of the pooled effect estimates of RCTs assessing the relationship between folic acid supplementation and CVD and all-cause mortality risk. RR, risk ratio; CIs, confidence intervals; RTCs, randomized clinical trials; CVD, cardiovascular disease; CHD, coronary heart disease; MI, myocardial infarction; N/A, not applicable. The diamonds represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% CIs, using the Manthel-Haenszel method with random effects model.

Folic acid Control Weight Risk RatioSubgroup and Study, Year Events Total Events TotalRCTsRighetti et al., 2003 13 51 11 30 3.2% 0.70 [0.36, 1.35]Liem et al., 2004 43 140 45 143 11.9% 0.98 [0.69, 1.38]Zoungas et al., 2006 - ASFAST 46 156 55 159 13.7% 0.85 [0.62, 1.18]Vianna et al., 2007 9 93 9 93 1.9% 1.00 [0.42, 2.41]Huo et al., 2015 - CSPPT 324 10,348 405 10,354 69.4% 0.80 [0.69, 0.92]


Favours folic acid Favours control


Study or SubgroupRighetti et al., 2003Liem et al., 2004Zoungas et al., 2006 - ASFASTVianna et al., 2007Huo et al., 2015 - CSPPT


Events134346

9324

435

Total51

140156

9310348

10788

Events114555

9405

525

Total30

143159

9310354

10779

Weight3.2%

11.9%13.7%

1.9%69.4%

100.0%

M-H, Random, 95% CI0.70 [0.36, 1.35]0.98 [0.69, 1.38]0.85 [0.62, 1.18]1.00 [0.42, 2.41]0.80 [0.69, 0.92]

0.83 [0.73, 0.93]

Year20032004200620072015

Folate Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.5 0.7 1 1.5 2Favours folate Favours control

Supplementary Figure 78. Forest plot of folic acid supplementation and total CVD risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

116

Subgroup and Study, Year Events Total Events TotalRCTsBaker et al., 2002 23 942 12 940 40.1% 1.91 [0.96, 3.82]Zoungas et al., 2006 - ASFAST 23 156 19 159 59.9% 1.23 [0.70, 2.17]

Total (95% CI) 1,098 1,099 100% 1.47 [0.95, 2.28]Total events 46 31 Heterogeneity: Tau² = 0.00; Chi² = 0.93, df = 1 (P = 0.33); I² = 0% 0.01 0.1 1 10 100Test for overall effect: Z = 1.73 (P = 0.08) Favours folic acid Favours control

Folic acid ControlWeight



Study or SubgroupBaker et al., 2002Zoungas et al., 2006 - ASFAST


Events2323

46

Total942156

1098

Events1219

31

Total940159

1099

Weight40.1%59.9%

100.0%

M-H, Random, 95% CI1.91 [0 .96, 3.82]1.23 [0 .70, 2.17]

1.47 [0.95, 2.28]

Year20022006


0.01 0.1 1 10 100Favours folate Favours control

Supplementary Figure 79. Forest plot of folic acid supplementation and total CHD risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsLiem et al., 2003 3 300 4 293 8.3% 0.73 [0.17, 3.24]Liem et al., 2004 8 140 10 143 20.5% 0.82 [0.33, 2.01]Cole et al., 2007 - AFPPS 14 516 8 505 22.1% 1.71 [0.72, 4.05]Logan et al., 2008 - ukCAP 3 470 0 469 2.2% 6.99 [0.36, 134.86]Wu et al., 2009 - NHS/HPFS 6 338 1 334 4.3% 5.93 [0.72, 48.98]Huo et al., 2015 - CSPPT 25 10,348 24 10,354 42.6% 1.04 [0.60, 1.82]






Study or SubgroupLiem et al., 2003Liem et al., 2004Cole et al., 2007 - AFPPSLogan et al., 2008Wu et al., 2009 - NHS/HPFSHuo et al., 2015 - CSPPT


Events38

1436

25

59

Total300140516470338

10348

12112

Events4

10801

24

47

Total293143505469334

10354

12098

Weight8.3%

20.5%22.1%

2.2%4.3%

42.6%

100.0%

M-H, Random, 95% CI0.73 [0.17, 3.24]0.82 [0.33, 2.01]1.71 [0.72, 4.05]

6.99 [0.36, 134.86]5.93 [0.72, 48.98]

1.04 [0.60, 1.82]

1.21 [0.78, 1.88]

Year200320042007200820092015



Supplementary Figure 80. Forest plot of folic acid supplementation and MI risk. Manthel-Haenszel M-H, MI, myocardial infarction. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

117

Subgroup and Study, Year Events Total Events TotalRCTsLiem et al., 2003 4 300 3 293 1.0% 1.30 [0.29, 5.77]Liem et al., 2004 1 140 0 143 0.2% 3.06 [0.13, 74.58]Zoungas et al., 2006 - ASFAST 8 156 18 159 3.4% 0.45 [0.20, 1.01]Cole et al., 2007 - AFPPS 9 516 5 505 1.9% 1.76 [0.59, 5.22]Logan et al., 2008 - ukCAP 1 470 1 469 0.3% 1.00 [0.06, 15.91]Wu et al., 2009 - NHS/HPFS 4 338 3 334 1.0% 1.32 [0.30, 5.84]Huo et al., 2015 - CSPPT 282 10,348 355 10,354 92.3% 0.79 [0.68, 0.93]






Study or SubgroupLiem et al., 2003Liem et al., 2004Zoungas et al., 2006 - ASFASTCole et al., 2007 - AFPPSLogan et al., 2008Wu et al., 2009 - NHS/HPFSHuo et al., 2015 - CSPPT


Events418914

282

309

Total300140156516470338

10348

12268

Events30

18513

355

385

Total293143159505469334

10354

12257

Weight1.0%0.2%3.4%1.9%0.3%1.0%

92.3%

100.0%

M-H, Random, 95% CI1.30 [0.29, 5.77]

3.06 [0.13, 74.58]0.45 [0.20, 1.01]1.76 [0.59, 5.22]

1.00 [0.06, 15.91]1.32 [0.30, 5.84]0.79 [0.68, 0.93]

0.80 [0.69, 0.93]

Year2003200420062007200820092015



Supplementary Figure 81. Forest plot of folic acid supplementation and stroke risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsLiem et al., 2003 7 300 9 293 7.9% 0.76 [0.29, 2.01]Zoungas et al., 2006 - ASFAST 21 156 24 159 25.7% 0.89 [0.52, 1.53]Vianna et al., 2007 17 93 21 93 23.1% 0.81 [0.46, 1.43]Wu et al., 2009 - NHS/HPFS 0 338 3 334 0.9% 0.14 [0.01, 2.72]Huo et al., 2015 - CSPPT 43 10,348 43 10,354 42.4% 1.00 [0.66, 1.53]






Study or SubgroupLiem et al., 2003Zoungas et al., 2006 - ASFASTVianna et al., 2007Wu et al., 2009 - NHS/HPFSHuo et al., 2015 - CSPPT


Events7

2117

043

88

Total300156

93338

10348

11235

Events9

2421

343

100

Total293159

93334

10354

11233

Weight7.9%

25.7%23.1%

0.9%42.4%

100.0%

M-H, Random, 95% CI0.76 [0.29, 2.01]0.89 [0.52, 1.53]0.81 [0.46, 1.43]0.14 [0.01, 2.72]1.00 [0.66, 1.53]

0.89 [0.68, 1.17]

Year20032006200720092015



Supplementary Figure 82. Forest plot of folic acid supplementation and CVD mortality risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with effects model.

118

Subgroup and Study, Year Events Total Events TotalRCTsLiem et al., 2004 2 140 2 143 31.3% 1.02 [0.15, 7.15]Huo et al., 2015 - CSPPT 5 10,348 4 10,354 68.7% 1.25 [0.34, 4.66]






Study or SubgroupLiem et al., 2004Huo et al., 2015 - CSPPT


Events25

7

Total140

10348

10488

Events24

6

Total143

10354

10497

Weight31.3%68.7%

100.0%

M-H, Random, 95% CI1.02 [0.15, 7.15]1.25 [0.34, 4.66]

1.17 [0.39, 3.49]

Year20042015



Supplementary Figure 83. Forest plot of folic acid supplementation and MI mortality risk. M-H, Manthel-Haenszel, MI, myocardial infarction. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsLiem et al., 2004 1 140 0 143 5.5% 3.06 [0.13, 74.58]Huo et al., 2015 - CSPPT 18 10,348 10 10,354 94.5% 1.80 [0.83, 3.90]






Study or SubgroupLiem et al., 2004Huo et al., 2015 - CSPPT


Events1

18

19

Total140

10348

10488

Events0

10

10

Total143

10354

10497

Weight5.5%

94.5%

100.0%

M-H, Random, 95% CI3.06 [0.13, 74.58]

1.80 [0.83, 3.90]

1.85 [0.88, 3.93]

Year20042015



Supplementary Figure 84. Forest plot of folic acid supplementation and stroke mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

119

Subgroup and Study, Year Events Total Events TotalRCTsLiem et al., 2003 12 300 14 293 5.7% 0.84 [0.39, 1.78]Liem et al., 2004 6 140 7 143 3.0% 0.88 [0.30, 2.54]Potena et al., 2005 1 26 0 25 0.4% 2.89 [0.12, 67.75]Zoungas et al., 2006 - ASFAST 45 156 46 159 20.2% 1.00 [0.71, 1.41]Vianna et al., 2007 23 93 30 93 13.3% 0.77 [0.48, 1.22]Durga et al., 2007 - FACIT 8 405 4 413 2.4% 2.04 [0.62, 6.72]Cole et al., 2007 - AFPPS 10 516 19 505 5.7% 0.52 [0.24, 1.10]Logan et al., 2008 - UKCAP 1 470 7 469 0.8% 0.14 [0.02, 1.15]Wu et al., 2009 - NHS/HPFS 7 338 15 334 4.3% 0.46 [0.19, 1.12]Huo et al., 2015 - CSPPT 302 10,348 320 10,354 44.1% 0.94 [0.81, 1.10]






Study or SubgroupLiem et al., 2003Liem et al., 2004Potena et al., 2005Zoungas et al., 2006 - ASFASTVianna et al., 2007Durga et al., 2007 - FACITCole et al., 2007 - AFPPSLogan et al., 2008 - UKCAPWu et al., 2009 - NHS/HPFSHuo et al., 2015 - CSPPT


Events12

61

4523

810

17

302

415

Total300140

26156

93405516470338

10348

12792

Events14

70

4630

419

715

320

462

Total293143

25159

93413505469334

10354

12788

Weight5.7%3.0%0.4%

20.2%13.3%

2.4%5.7%0.8%4.3%

44.1%

100.0%

M-H, Random, 95% CI0.84 [0.39, 1.78]0.88 [0.30, 2.54]

2.89 [0.12, 67.75]1.00 [0.71, 1.41]0.77 [0.48, 1.22]2.04 [0.62, 6.72]0.52 [0.24, 1.10]0.14 [0.02, 1.15]0.46 [0.19, 1.12]0.94 [0.81, 1.10]

0.87 [0.72, 1.05]

Year2003200420052006200720072007200820092015



Supplementary Figure 85. Forest plot of folic acid supplementation and all-cause mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Supplementary Figure 86. Funnel plot of folic acid supplementation and all-cause mortality risk. The vertical line represents the pooled effect estimate expressed as a RR. Dashed lines represent pseudo-95% confidence intervals (CI). The circles represent risk estimates for each study, and the horizontal lines represent standard errors of the RR. We were unable to test for other CVD outcomes (<10 RCTs).

120

Random Effects

ComparisonRCTs/

CohortsN Events RR (95% CIs) RR (95% CIs) I2 p -value

CVD (RCTs)Total CVD 4 33,162 7,045 0.97 [0.93, 1.03] 34% 0.33Total CHD 2 26,615 1,363 0.96 [0.87, 1.07] 0% 0.47MI 5 34,104 1,536 0.90 [0.75, 1.06] 41% 0.20

Stroke 5 34,104 1,453 0.95 [0.72, 1.26] 60% 0.73Total CVD mortality 3 7,489 957 0.95 [0.84, 1.08] 0% 0.46

Total CHD mortality 3 32,995 1,403 0.99 [0.89, 1.10] 0% 0.89

ALL-CAUSE MORTALITY All-cause mortality (RCTs) 4 33,103 2,691 1.04 [0.95, 1.14] 16% 0.36



Favours niacin(B3)/ Favours control/0.5 1.0 1.5

Supplementary Figure 87. Summary of the pooled effect estimates of RCTs assessing the relationship between niacin (B3) supplementation and CVD and all cause mortality risk. RR, risk ratio; CIs, confidence intervals; RTCs, randomized clinical trials; CVD, cardiovascular disease; CHD, coronary heart disease; MI, myocardial infarction; N/A, not applicable. The diamonds represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% CIs, using the Manthel-Haenszel method with random effects model.

121

Subgroup and Study, Year Events Total Events TotalRCTsNo statinsCDPRG 1975 914 1,119 2,333 2,789 58.3% 0.98 [0.95, 1.01]Subtotal (95% CI) 1,119 2,789 58.3% 0.98 [0.95, 1.01]Total events 914 2,333 Heterogeneity: Not applicableTest for overall effect: Z = 1.45 (P = 0.15)

Background statin treatmentTaylor et al., 2004 - ARBITER 2 4 87 11 80 0.2% 0.33 [0.11, 1.01]Boden et al., 2011 - AIM HIGH* 171 1,718 158 1,696 5.7% 1.07 [0.87, 1.31]HPS2-THRIVE Collaborative Group 2014 1,696 12,838 1,758 12,835 35.8% 0.96 [0.91, 1.03]Subtotal (95% CI) 14,643 14,611 41.7% 0.97 [0.81, 1.17]Total events 1,871 1,927 Heterogeneity: Tau² = 0.01; Chi² = 4.46, df = 2 (P = 0.11); I² = 55%Test for overall effect: Z = 0.29 (P = 0.77)

Total (95% CI) 15,762 17,400 100% 0.97 [0.93, 1.03]Total events 2,785 4,260 Heterogeneity: Tau² = 0.00; Chi² = 4.54, df = 3 (P = 0.21); I² = 34%Test for overall effect: Z = 0.98 (P = 0.33) 0.2 0.5 1 2 5 Test for subgroup differences: Chi² = 0.00, df = 1 (P = 0.96), I² = 0% Favours B3 (niacin) Favours control

Vitamin B3 ControlWeight



Study or Subgroup1.1.1 No statinsCDPRG 1975Subtotal (95% CI)Total eventsHeterogeneity: Not applicableTest for overall effect: Z = 1.45 (P = 0.15)

1.1.2 Background statin treatmentTaylor et al., 2004 - ARBITER 2Boden et al., 2011 - AIM HIGHHPS 2 THRIVE 2014Subtotal (95% CI)Total eventsHeterogeneity: Tau² = 0.01; Chi² = 4.46, df = 2 (P = 0.11); I² = 55%Test for overall effect: Z = 0.29 (P = 0.77)

Total (95% CI)Total eventsHeterogeneity: Tau² = 0.00; Chi² = 4.54, df = 3 (P = 0.21); I² = 34%Test for overall effect: Z = 0.98 (P = 0.33)Test for subgroup differences: Chi² = 0.00, df = 1 (P = 0.96), I² = 0%

Events

914

914

4171

1696

1871

2785

Total

11191119

871718

1283814643

15762

Events

2333

2333

11158

1758

1927

4260

Total

27892789

801696

1283514611

17400

Weight

58.3%58.3%

0.2%5.7%

35.8%41.7%

100.0%

M-H, Random, 95% CI

0.98 [0.95, 1.01]0.98 [0.95, 1.01]

0.33 [0.11, 1.01]1.07 [0.87, 1.31]0.96 [0.91, 1.03]0.97 [0.81, 1.17]

0.97 [0.93, 1.03]

Year

1975

200420112014

Vitamin B3 (Niacin) Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.2 0.5 1 2 5Favours B3 (Niacin) Favours Control

Supplementary Figure 88. Forest plot of niacin (B3) supplementation and total CVD risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. *Boden et al., 2011 - reported number of participants with at least one event. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsGuyton et al., 2008 1 670 0 272 0.1% 1.22 [0.05, 29.87]HPS2-THRIVE Collaborative Group 2014 668 12,838 694 12,835 99.9% 0.96 [0.87, 1.07]


Favours B3 (niacin) Favours control




Study or SubgroupGuyton et al., 2008HPS 2 THRIVE 2014


Events1

668

669

Total670

12838

13508

Events0

694

694

Total272

12835

13107

Weight0.1%

99.9%

100.0%

M-H, Random, 95% CI1.22 [0.05, 29.87]

0.96 [0.87, 1.07]

0.96 [0.87, 1.07]

Year20082014


0.05 0.2 1 5 20Favours Vitamin B3 Favours Control

Supplementary Figure 89. Forest plot of niacin (B3) supplementation and total CHD risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

122

Subgroup and Study, Year Events Total Events TotalRCTsNo statinsCDPRG 1975 114 1,119 386 2,789 32.7% 0.74 [0.60, 0.90]Subtotal (95% CI) 1,119 2,789 32.7% 0.74 [0.60, 0.90]Total events 114 386 Heterogeneity: Not applicableTest for overall effect: Z = 3.05 (P = 0.002)

Background statin treatmentTaylor et al., 2004 - ARBITER 2 2 87 2 80 0.8% 0.92 [0.13, 6.38]Guyton et al., 2008 1 670 1 272 0.4% 0.41 [0.03, 6.47]Boden et al., 2011 - AIM HIGH* 104 1,718 93 1,696 23.5% 1.10 [0.84, 1.45]HPS2-THRIVE Collaborative Group 2014 402 12,838 431 12,835 42.7% 0.93 [0.82, 1.07]Subtotal (95% CI) 15,313 14,883 67.3% 0.96 [0.85, 1.08]Total events 509 527Heterogeneity: Tau² = 0.00; Chi² = 1.57, df = 3 (P = 0.67); I² = 0%Test for overall effect: Z = 0.64 (P = 0.53)

Total (95% CI) 16,432 17,672 100% 0.90 [0.75, 1.06]Total events 623 913 Heterogeneity: Tau² = 0.01; Chi² = 6.75, df = 4 (P = 0.15); I² = 41%Test for overall effect: Z = 1.27 (P = 0.20) 0.02 0.1 1 10 50 Test for subgroup differences: Chi² = 5.18, df = 1 (P = 0.02), I² = 80.7% Favours B3 (niacin) Favours control





1.3.2 Background statin treatmentTaylor et al., 2004 - ARBITER 2Guyton et al., 2008Boden et al., 2011 - AIM HIGHHPS 2 THRIVE 2014Subtotal (95% CI)Total eventsHeterogeneity: Tau² = 0.00; Chi² = 1.57, df = 3 (P = 0.67); I² = 0%Test for overall effect: Z = 0.64 (P = 0.53)

Total (95% CI)Total eventsHeterogeneity: Tau² = 0.01; Chi² = 6.75, df = 4 (P = 0.15); I² = 41%Test for overall effect: Z = 1.27 (P = 0.20)Test for subgroup differences: Chi² = 5.18, df = 1 (P = 0.02), I² = 80.7%

Events

114

114

21

104402

509

623

Total

11191119

87670

17181283815313

16432

Events

386

386

21

93431

527

913

Total

27892789

80272

16961283514883

17672

Weight

32.7%32.7%

0.8%0.4%

23.5%42.7%67.3%

100.0%

M-H, Random, 95% CI

0.74 [0.60, 0.90]0.74 [0.60, 0.90]

0.92 [0.13, 6.38]0.41 [0.03, 6.47]1.10 [0.84, 1.45]0.93 [0.82, 1.07]0.96 [0.85, 1.08]

0.90 [0.75, 1.06]

Year

1975

2004200820112014



Supplementary Figure 90. Forest plot of niacin (B3) supplementation and MI risk. M-H, Manthel-Haenszel, MI, myocardial infarction. *Boden et al., 2011 - reported number of participants with at least one event. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

123


Back ground statin treatmentTaylor et al., 2004 - ARBITER 2 0 87 1 80 0.8% 0.31 [0.01, 7.42]Guyton et al., 2008 0 670 1 272 0.8% 0.14 [0.01, 3.32]Boden et al., 2011 - AIM HIGH* 30 1,718 18 1,696 15.8% 1.65 [0.92, 2.94]HPS2-THRIVE Collaborative Group 2014 498 12,838 499 12,835 44.9% 1.00 [0.88, 1.13]Subtotal (95% CI) 15,313 14,883 62.2% 1.10 [0.70, 1.72]Total events 528 519 Heterogeneity: Tau² = 0.08; Chi² = 4.80, df = 3 (P = 0.19); I² = 37%Test for overall effect: Z = 0.42 (P = 0.68)

Total (95% CI) 16,432 17,672 100% 0.95 [0.72, 1.26]Total events 623 830Heterogeneity: Tau² = 0.04; Chi² = 9.89, df = 4 (P = 0.04); I² = 60%Test for overall effect: Z = 0.35 (P = 0.73) 0.05 0.2 1 5 20 Test for subgroup differences: Chi² = 2.09, df = 1 (P = 0.15), I² = 52.2% Favours B3 (niacin) Favours control




Study or Subgroup1.5.1 no statinsCDPRG 1975Subtotal (95% CI)Total eventsHeterogeneity: Not applicableTest for overall effect: Z = 2.44 (P = 0.01)

1.5.2 Back ground statin treatmentTaylor et al., 2004 - ARBITER 2Guyton et al., 2008Boden et al., 2011 - AIM HIGHHPS 2 THRIVE 2014Subtotal (95% CI)Total eventsHeterogeneity: Tau² = 0.08; Chi² = 4.80, df = 3 (P = 0.19); I² = 37%Test for overall effect: Z = 0.42 (P = 0.68)


Events

95

95

00

30498

528

623

Total

11191119

87670

17181283815313

16432

Events

311

311

11

18499

519

830

Total

27892789

80272

16961283514883

17672

Weight

37.8%37.8%

0.8%0.8%

15.8%44.9%62.2%

100.0%

M-H, Random, 95% CI

0.76 [0.61, 0.95]0.76 [0.61, 0.95]

0.31 [0.01, 7.42]0.14 [0.01, 3.32]1.65 [0.92, 2.94]1.00 [0.88, 1.13]1.10 [0.70, 1.72]

0.95 [0.72, 1.26]

Year

1975

2004200820112014



Supplementary Figure 91. Forest plot of niacin (B3) supplementation and stroke risk. M-H, Manthel-Haenszel. *Boden et al., 2011 - reported number of participants with at least one event. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.


Background statin treatmentTaylor et al., 2004 - ARBITER 2 1 87 2 80 0.3% 0.46 [0.04, 4.97]Boden et al., 2011 - AIM HIGH 45 1,718 38 1,696 8.7% 1.17 [0.76, 1.79]Subtotal (95% CI) 1,805 1,776 9.0% 1.14 [0.75, 1.73]Total events 46 40Heterogeneity: Tau² = 0.00; Chi² = 0.57, df = 1 (P = 0.45); I² = 0%Test for overall effect: Z = 0.59 (P = 0.55)

Total (95% CI) 2,924 4,565 100% 0.95 [0.84, 1.08]Total events 284 673 Heterogeneity: Tau² = 0.00; Chi² = 1.31, df = 2 (P = 0.52); I² = 0%Test for overall effect: Z = 0.74 (P = 0.46) 0.05 0.2 1 5 20 Test for subgroup differences: Chi² = 0.73, df = 1 (P = 0.39), I² = 0% Favours B3 (niacin) Favours control





1.6.2 Background statin treatmentTaylor et al., 2004 - ARBITER 2Boden et al., 2011 - AIM HIGHSubtotal (95% CI)Total eventsHeterogeneity: Tau² = 0.00; Chi² = 0.57, df = 1 (P = 0.45); I² = 0%Test for overall effect: Z = 0.59 (P = 0.55)

Total (95% CI)Total eventsHeterogeneity: Tau² = 0.00; Chi² = 1.31, df = 2 (P = 0.52); I² = 0%Test for overall effect: Z = 0.74 (P = 0.46)Test for subgroup differences: Chi² = 0.73, df = 1 (P = 0.39), I² = 0%

Events

238

238

145

46

284

Total

11191119

8717181805

2924

Events

633

633

238

40

673

Total

27892789

8016961776

4565

Weight

91.0%91.0%

0.3%8.7%9.0%

100.0%

M-H, Random, 95% CI

0.94 [0.82, 1.07]0.94 [0.82, 1.07]

0.46 [0.04, 4.97]1.17 [0.76, 1.79]1.14 [0.75, 1.73]

0.95 [0.84, 1.08]

Year

1975

20042011



Supplementary Figure 92. Forest plot of niacin (B3) supplementation and CVD mortality risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

124


Background statin treatmentBoden et al., 2011 - AIM HIGH 38 1,718 34 1,696 5.2% 1.10 [0.70, 1.74]HPS2-THRIVE Collaborative Group 2014 302 12,838 291 12,835 43.3% 1.04 [0.88, 1.22]Subtotal (95% CI) 14,556 14,531 48.5% 1.04 [0.90, 1.21]Total events 340 325Heterogeneity: Tau² = 0.00; Chi² = 0.06, df = 1 (P = 0.80); I² = 0%Test for overall effect: Z = 0.57 (P = 0.57)

Total (95% CI) 15,675 17,320 100% 0.99 [0.89, 1.10]Total events 543 860 Heterogeneity: Tau² = 0.00; Chi² = 0.93, df = 2 (P = 0.63); I² = 0%Test for overall effect: Z = 0.14 (P = 0.89) 0.7 0.85 1 1.2 1.5Test for subgroup differences: Chi² = 0.86, df = 1 (P = 0.35), I² = 0% Favours B3 (niacin) Favours control





1.7.2 Background statin treatmentBoden et al., 2011 - AIM HIGHHPS 2 THRIVE 2014Subtotal (95% CI)Total eventsHeterogeneity: Tau² = 0.00; Ch i² = 0.06, df = 1 (P = 0.80); I² = 0%Test for overall effect: Z = 0.57 (P = 0.57)

Total (95% CI)Total eventsHeterogeneity: Tau² = 0.00; Ch i² = 0.93, df = 2 (P = 0.63); I² = 0%Test for overall effect: Z = 0.14 (P = 0.89)Test for subgroup differences: Chi² = 0.86, df = 1 (P = 0.35), I² = 0%

Events

203

203

38302

340

543

Total

11191119

17181283814556

15675

Events

535

535

34291

325

860

Total

27892789

16961283514531

17320

Weight

51.5%51.5%

5.2%43.3%48.5%

100.0%

M-H, Random, 95% CI

0.95 [0.82, 1.09]0.95 [0.82, 1.09]

1.10 [0.70, 1.74]1.04 [0.88, 1.22]1.04 [0.90, 1.21]

0.99 [0.89, 1.10]

Year

1975

20112014


0.7 0.85 1 1.2 1.5Favours B3 (Niacin) Favours Control

Supplementary Figure 93. Forest plot of niacin (B3) supplementation and CHD mortality risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalNo statinsCDPRG 1975 273 1,119 709 2,789 38.8% 0.96 [0.85, 1.08]Subtotal (95% CI) 1,119 2,789 38.8% 0.96 [0.85, 1.08]Total events 273 709 Heterogeneity: Not applicableTest for overall effect: Z = 0.67 (P = 0.51)

Background statin treatmentSang et al., 2009* - 52 1 56 0.1% 0.36 [0.01, 8.61]Boden et al., 2011 - AIM HIGH 96 1,718 82 1,696 8.9% 1.16 [0.87, 1.54]HPS2-THRIVE Collaborative Group 2014 798 12,838 732 12,835 52.3% 1.09 [0.99, 1.20]Subtotal (95% CI) 14,608 14,587 61.2% 1.10 [1.00, 1.20]Total events 894 815Heterogeneity: Tau² = 0.00; Chi² = 0.62, df = 2 (P = 0.73); I² = 0%Test for overall effect: Z = 1.94 (P = 0.05)

Total (95% CI) 15,727 17,376 100.0% 1.04 [0.95, 1.14]Total events 1,167 1,524 Heterogeneity: Tau² = 0.00; Chi² = 3.55, df = 3 (P = 0.31); I² = 16%Test for overall effect: Z = 0.91 (P = 0.36) 0.2 0.5 1 2 5 Test for subgroup differences: Chi² = 2.90, df = 1 (P = 0.09), I² = 65.6% Favours B3 (niacin) Favours control





1.5.2 Background statin treatmentSang et al., 2009Boden et al., 2011 - AIM HIGHHPS2-THRIVE Collaborative Group 2014Subtotal (95% CI)Total eventsHeterogeneity: Tau² = 0.00; Chi² = 0.62, df = 2 (P = 0.73); I² = 0%Test for overall effect: Z = 1.94 (P = 0.05)


Events

273

273

096

798

894

1167

Total

11191119

521718

1283814608

15727

Events

709

709

182

732

815

1524

Total

27892789

561696

1283514587

17376

Weight

38.8%38.8%

0.1%8.9%

52.3%61.2%

100.0%

M-H, Random, 95% CI

0.96 [0.85, 1.08]0.96 [0.85, 1.08]

0.36 [0.01, 8.61]1.16 [0.87, 1.54]1.09 [0.99, 1.20]1.10 [1.00, 1.20]

1.04 [0.95, 1.14]

Year

1975

200920112014



Supplementary Figure 94. Forest plot of niacin (B3) supplementation and all-cause mortality risk. M-H, Manthel-Haenszel. *Sang et al., 2009 - Data taken from meta-analysis Keene et al., 2014. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

125

Random Effects

Comparison RCTs/Cohorts

N Events RR (95% CIs) RR (95% CIs) I2 p -value

CVD (RCTs)

Total CVD 1 1,877 347 1.03 [0.85, 1.24] N/A 0.78MI 2 3,429 427 1.04 [0.87, 1.23] 0% 0.69

Stroke 2 3,429 88 0.92 [0.61, 1.40] 0% 0.70

MI mortality 1 1,877 120 1.04 [0.74, 1.48] N/A 0.81

ALL-CAUSE MORTALITY

All-cause mortality (RCTs) 2 3,429 239 1.02 [0.80, 1.30] 0% 0.88


Favours vitamin B6/ Favours control/ Negative association Positive association

0.5 1.0 1.5

Supplementary Figure 95. Summary of the pooled effect estimates of RCTs studies assessing the relationship between vitamin B6 supplementation and CVD and all cause mortality risk. RR, risk ratio; CIs, confidence intervals; RTCs, randomized clinical trials; CVD, cardiovascular disease; MI, myocardial infarction; N/A, not applicable. The diamonds represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% CIs, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsBonaa et al., 2006 - NORVIT 175 934 172 943 100.0% 1.03 [0.85, 1.24]

Total (95% CI) 934 943 100% 1.03 [0.85, 1.24]Total events 175 172 Heterogeneity: Not applicableTest for overall effect: Z = 0.28 (P = 0.78) 0.2 0.5 1 2 5

Favours vitamin B6 Favours control




Study or SubgroupBonaa et al., 2006 - NORVIT


Events175

175

Total934

934

Events172

172

Total943

943

Weight100.0%

100.0%

M-H, Random, 95% CI1.03 [0.85, 1.24]

1.03 [0.85, 1.24]

Year2006

Vitamin B6 Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.2 0.5 1 2 5Favours vitamin B6 Favours control

Supplementary Figure 96. Forest plot of vitamin B6 supplementation and total CVD risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

126

Subgroup and Study, Year Events Total Events TotalRCTsBonaa et al., 2006 - NORVIT 161 934 153 943 75.6% 1.06 [0.87, 1.30]Ebbing et al., 2008 - WENBIT 55 772 58 780 24.4% 0.96 [0.67, 1.37]

Total (95% CI) 1,706 1,723 100% 1.04 [0.87, 1.23]Total events 216 211 Heterogeneity: Tau² = 0.00; Chi² = 0.25, df = 1 (P = 0.62); I² = 0%Test for overall effect: Z = 0.39 (P = 0.69) 0.7 0.85 1 1.2 1.5





Study or SubgroupBonaa et al., 2006 - NORVITEbbing et al., 2008 - WENBIT


Events16155

216

Total934772

1706

Events15358

211

Total943780

1723

Weight75.6%24.4%

100.0%

M-H, Random, 95% CI1.06 [0.87, 1.30]0.96 [0.67, 1.37]

1.04 [0.87, 1.23]

Year20062008


0.7 0.85 1 1.2 1.5Favours vitamin B6 Favours control

Supplementary Figure 97. Forest plot of vitamin B6 supplementation and MI risk. M-H, Manthel-Haenszel, MI, myocardial infarction. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.









Events2220

42

Total934772

1706

Events2719

46

Total943780

1723

Weight55.5%44.5%

100.0%

M-H, Random, 95% CI0.82 [0.47, 1.43]1.06 [0.57, 1.98]

0.92 [0.61, 1.40]

Year20062008


0.05 0.2 1 5 20Favours vitamin B6 Favours control

Supplementary Figure 98. Forest plot of vitamin B6 supplementation and stroke risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsBonaa et al., 2006 - NORVIT 61 934 59 943 100% 1.04 [0.74, 1.48]

Total (95% CI) 934 943 100% 1.04 [0.74, 1.48]Total events 61 59 Heterogeneity: Not applicableTest for overall effect: Z = 0.24 (P = 0.81) 0.5 0.7 1 1.5 2





Study or SubgroupBonaa et al., 2006 - NORVIT


Events61

61

Total934

934

Events59

59

Total943

943

Weight100.0%

100.0%

M-H, Random, 95% CI1.04 [0.74, 1.48]

1.04 [0.74, 1.48]

Year2006


0.5 0.7 1 1.5 2Favours vitamin B6 Favours control

Supplementary Figure 99. Forest plot of vitamin B6 supplementation and MI mortality risk. M-H, Manthel-Haenszel, MI, myocardial infarction. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

127


Total (95% CI) 1,706 1,723 100% 1.02 [0.80, 1.30]

Total events 120 119 Heterogeneity: Tau² = 0.00; Chi² = 0.12, df = 1 (P = 0.73); I² = 0%Test for overall effect: Z = 0.16 (P = 0.88) 0.5 0.7 1 1.5 2


Vitamin B6 Control Risk RatioM-H, Random, 95% CI

Risk RatioM-H, Random, 95% CI in all-cause mortality riskWeight



Events9228

120

Total934772

1706

Events8930

119

Total943780

1723

Weight76.9%23.1%

100.0%

M-H, Random, 95% CI1.04 [0.79, 1.38]0.94 [0.57, 1.56]

1.02 [0.80, 1.30]

Year20062008


0.5 0.7 1 1.5 2Favours vitamin B6 Favours control

Supplementary Figure 100. Forest plot of vitamin B6 supplementation and all-cause mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Random Effects


CVD

Total CVD 3 3,328 364 1.43 [0.79, 2.59]80% 0.23

Total CHD 2 3,171 166 1.16 [0.87, 1.56] 0% 0.32MI 4 5,387 231 1.69 [0.94, 3.04] 69% 0.08Stroke 3 3,861 178 1.29 [0.96, 1.72] 0% 0.09Total CVD mortality 2 2,931 47 1.24 [0.27, 5.65] 53% 0.78Total CHD mortality 1 5,292 338 1.15 [0.94, 1.41] N/A 0.18MI mortality 1 1,460 22 1.44 [0.62, 3.36] N/A 0.39Stroke mortality 1 1,460 14 0.75 [0.26, 2.15] N/A 0.59


Favours calcium/ Favours control/ Negative association Positive association


0.0 0.5 1.0 1.5 2.0

Supplementary Figure 101. Summary of the pooled effect estimates of RCTs studies assessing the relationship between calcium supplementation and CVD and all cause mortality risk. RR, risk ratio; CIs, confidence intervals; RTCs, randomized clinical trials; CVD, cardiovascular disease; CHD, coronary heart disease; MI, myocardial infarction; N/A, not applicable. The diamonds represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% CIs, using the Manthel-Haenszel method with random effects model.

128

Subgroup and Study, Year Events Total Events TotalRCTsBolland et al., 2008 101 732 54 739 47.1% 1.89 [1.38, 2.59]Chailurkit et al., 2010 2 201 0 196 3.6% 4.88 [0.24, 100.93]Lewis et al., 2011 - CAIFOS* 104 730 103 730 49.4% 1.01 [0.78, 1.30]


Favours calcium Favours control

Calcium ControlWeight



Study or SubgroupBolland et al., 2008Chailurkit et al., 2010Lewis et al., 2011 - CAIFOS


Events101

2104

207

Total732201730

1663

Events540

103

157

Total739196730

1665

Weight47.1%3.6%

49.4%

100.0%

M-H, Random, 95% CI1.89 [1.38, 2.59]

4.88 [0.24, 100.93]1.01 [0.78, 1.30]

1.43 [0.79, 2.59]

Year200820102011

Calcium Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.01 0.1 1 10 100Favours calcium Favours control

Supplementary Figure 102. Forest plot of calcium supplementation and total CVD risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. *Lewis et al., 2011 - reported number of participants with at least one event. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsBaron et al., 1999 50 464 46 466 60.0% 1.09 [0.75, 1.60]Grant et al., 2005 - RECORD* 39 1,113 31 1,128 40.0% 1.28 [0.80, 2.03]






Study or SubgroupBaron et al., 1999Grant et al., 2005 - RECORD


Events5039

89

Total464

1113

1577

Events4631

77

Total466

1128

1594

Weight60.0%40.0%

100.0%

M-H, Random, 95% CI1.09 [0.75, 1.60]1.28 [0.80, 2.03]

1.16 [0.87, 1.56]

Year19992005



Supplementary Figure 103. Forest plot of calcium supplementation and total CHD risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. *Grant et al., 2005 - data taken from Lewis et al, 2015 (meta), however, data include women only; The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

129

Subgroup and Study, Year Events Total Events TotalRCTsGrant et al., 2005 - RECORD* 26 1,113 13 1,128 27.6% 2.03 [1.05, 3.92]Reid et al., 2008† 2 108 - 107 3.5% 4.95 [0.24, 101.99]Bolland et al., 2008 45 732 19 739 31.6% 2.39 [1.41, 4.05]Lewis et al., 2011 - CAIFOS‡ 63 730 63 730 37.3% 1.00 [0.72, 1.40]






Study or SubgroupGrant et al., 2005 - RECORDReid et al., 2008Bolland et al., 2008Lewis et al., 2011 - CAIFOS


Events262

4563

136

Total1113108732730

2683

Events130

1963

95

Total1128107739730

2704

Weight27.6%3.5%

31.6%37.3%

100.0%

M-H, Random, 95% CI2.03 [1.05, 3.92]

4.95 [0.24, 101.99]2.39 [1.41, 4.05]1.00 [0.72, 1.40]

1.69 [0.94, 3.04]

Year2005200820082011



Supplementary Figure 104. Forest plot of calcium supplementation and MI risk. M-H, Manthel-Haenszel, MI, Myocardial infarction. *Grant et al., 2005 - data taken from Lewis et al, 2015 (meta), however, data include women only; †Reid et al., 2008 - highest dose of calcium (1200mg/day) used as intervention group; ‡Lewis et al., 2011 - reported number of participants with at least one event. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsBaron et al., 1999 12 464 11 466 12.9% 1.10 [0.49, 2.46]Bolland et al., 2008 52 732 34 739 47.6% 1.54 [1.01, 2.35]Lewis et al., 2011 - CAIFOS* 36 730 33 730 39.5% 1.09 [0.69, 1.73]






Study or SubgroupBaron et al., 1999Bolland et al., 2008Lewis et al., 2011 - CAIFOS


Events125236

100

Total464732730

1926

Events113433

78

Total466739730

1935

Weight12.9%47.6%39.5%

100.0%

M-H, Random, 95% CI1.10 [0.49, 2.46]1.54 [1.01, 2.35]1.09 [0.69, 1.73]

1.29 [0.96, 1.72]

Year199920082011



Supplementary Figure 105. Forest plot of calcium supplementation and stroke risk. M-H, Manthel-Haenszel. *Lewis et al., 2011 - reported number of participants with at least one event; The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

130

Subgroup and Study, Year Events Total Events TotalRCTsBolland et al., 2008 4 732 1 739 29.9% 4.04 [0.45, 36.04]Lewis et al., 2011 - CAIFOS* 18 730 24 730 70.1% 0.75 [0.41, 1.37]

Total (95% CI) 1,462 1,469 100% 1.24 [0.27, 5.65]Total events 22 25 Heterogeneity: Tau² = 0.76; Chi² = 2.13, df = 1 (P = 0.14); I² = 53% 0.002 0.1 1 10 500Test for overall effect: Z = 0.28 (P = 0.78) Favours calcium Favours control




Study or SubgroupBolland et al., 2008Lewis et al., 2011 - CAIFOS


Events4

18

22

Total732730

1462

Events1

24

25

Total739730

1469

Weight29.9%70.1%

100.0%

M-H, Random, 95% CI4.04 [0.45, 36.04]

0.75 [0.41, 1.37]

1.24 [0.27, 5.65]

Year20082011



Supplementary Figure 106. Forest plot of calcium supplementation and CVD mortality risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. *Lewis et al., 2011 - reported number of participants with at least one event. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsLewis et al., 2011 - CAIFOS* 13 730 9 730 100% 1.44 [0.62, 3.36]






Study or SubgroupLewis et al., 2011 - CAIFOS


Events13

13

Total730

730

Events9

9

Total730

730

Weight100.0%

100.0%

M-H, Random, 95% CI1.44 [0.62, 3.36]

1.44 [0.62, 3.36]

Year2011



Supplementary Figure 107. Forest plot of calcium supplementation and MI mortality risk. M-H, Manthel-Haenszel, MI, myocardial infarction. *Lewis et al., 2011 - reported number of participants with at least one event. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

131

Subgroup and Study, Year Events Total Events TotalRCTsLewis et al., 2011 - CAIFOS* 6 730 8 730 100.0% 0.75 [0.26, 2.15]






Study or SubgroupLewis et al., 2011 - CAIFOS


Events6

6

Total730

730

Events8

8

Total730

730

Weight100.0%

100.0%

M-H, Random, 95% CI0.75 [0.26, 2.15]

0.75 [0.26, 2.15]

Year2011

Calclium Control Risk Ratio Risk RatioM-H, Random, 95% CI


Supplementary Figure 108. Forest plot of calcium supplementation and stroke mortality risk. M-H, Manthel-Haenszel. *Lewis et al., 2011 - reported number of participants with at least one event. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsBaron et al., 1999 25 464 22 466 3.9% 1.14 [0.65, 1.99]Grant et al., 2005 - RECORD 464 2,617 434 2,675 84.9% 1.09 [0.97, 1.23]Reid et al., 2008* 1 108 1 107 0.2% 0.99 [0.06, 15.64]Bolland et al., 2008 38 732 30 739 5.5% 1.28 [0.80, 2.04]Chailurkit et al., 2010 1 201 1 196 0.2% 0.98 [0.06, 15.48]Lewis et al., 2011 - CAIFOS 29 730 38 730 5.4% 0.76 [0.48, 1.22]






Study or SubgroupBaron et al., 1999Grant et al., 2005 - RECORDReid et al., 2008Bolland et al., 2008Chailurkit et al., 2010Lewis et al., 2011 - CAIFOS


Events25

4641

381

29

558

Total464

2617108732201730

4852

Events22

4341

301

38

526

Total466

2675107739196730

4913

Weight3.9%

84.9%0.2%5.5%0.2%5.4%

100.0%

M-H, Random, 95% CI1.14 [0.65, 1.99]1.09 [0.97, 1.23]

0.99 [0.06, 15.64]1.28 [0.80, 2.04]

0.98 [0.06, 15.48]0.76 [0.48, 1.22]

1.08 [0.97, 1.21]

Year199920052008200820102011



Supplementary Figure 109. Forest plot of calcium supplementation and all-cause mortality risk. M-H, Manthel-Haenszel. *Reid et al., 2008 - highest dose of calcium (1200mg/day) used as intervention group. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

132

Random Effects


CVD

Total CVD 1 459 43 0.49 [0.28, 0.86] N/A 0.01

MI 1 459 5 0.34 [0.06, 2.01] N/A 0.23

Total CVD mortality 2 760 31 0.80 [0.40, 1.58] 0% 0.51

ALL-CAUSE MORTALITY

All-cause mortality 2 762 35 0.79 [0.42, 1.51] 0% 0.48


Favours iron/ Favours control/


0.0 0.5 1.0 1.5 2.0

Supplementary Figure 110. Summary of the pooled effect estimates of RCTs and studies assessing the relationship between iron supplementation and CVD and all cause mortality risk. RR, risk ratio; CIs, confidence intervals; RTCs, randomized clinical trials; CVD, cardiovascular disease; MI, myocardial infarction; N/A, not applicable. The diamonds represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% CIs, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsAnker et al., 2009 - FAIR-HF* 21 304 22 155 100% 0.49 [0.28, 0.86]

Total (95% CI) 304 155 100% 0.49 [0.28, 0.86]Total events 21 22Heterogeneity: Not applicableTest for overall effect: Z = 2.49 (P = 0.01) 0.01 0.1 1 10 100

Favours iron Favours control

Iron ControlWeight



Study or SubgroupAnker et al., 2009 - FAIR-HF


Events21

21

Total304

304

Events22

22

Total155

155

Weight100.0%

100.0%

M-H, Random, 95% CI0.49 [0.28, 0.86]

0.49 [0.28, 0.86]

Year2009

Iron Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.01 0.1 1 10 100Favours iron Favours control

Supplementary Figure 111. Forest plot of iron supplementation and total CVD risk. M-H, Manthel-Haenszel. CVD, cardiovascular disease. *Anker et al., 2009 - To treat iron deficiency in heart failure patients. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

133

Subgroup and Study, Year Events Total Events TotalRCTsAnker et al., 2009 - FAIR-HF* 2 304 3 155 100% 0.34 [0.06, 2.01]



Iron ControlWeight



Study or SubgroupAnker et al., 2009 - FAIR-HF


Events2

2

Total304

304

Events3

3

Total155

155

Weight100.0%

100.0%

M-H, Random, 95% CI0.34 [0.06, 2.01]

0.34 [0.06, 2.01]

Year2009



Supplementary Figure 112. Forest plot of iron supplementation and MI risk. M-H, Manthel-Haenszel. MI, myocardial infarction. *Anker et al., 2009 - To treat iron deficiency in heart failure patients. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsAnker et al., 2009 - FAIR-HF* 4 304 4 155 25% 0.51 [0.13, 2.01]Ponikowski et al., 2015 - CONFIRM HF 11 150 12 151 75% 0.92 [0.42, 2.03]



Iron ControlWeight



Study or SubgroupAnker et al., 2009 - FAIR-HFPonikowski et al., 2015 - CONFIRM HF


Events4

11

15

Total304150

454

Events4

12

16

Total155151

306

Weight24.7%75.3%

100.0%

M-H, Random, 95% CI0.51 [0.13, 2.01]0.92 [0.42, 2.03]

0.80 [0.40, 1.58]

Year20092015



Supplementary Figure 113. Forest plot of iron supplementation and CVD mortality risk. M-H, Manthel-Haenszel. CVD, cardiovascular disease. *Anker et al., 2009 - To treat iron deficiency in heart failure patients. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsAnker et al., 2009 - FAIR-HF* 5 304 4 155 24.3% 0.64 [0.17, 2.34]Ponikowski et al. 2015 - CONFIRM-HF* 12 152 14 151 75.7% 0.85 [0.41, 1.78]



Iron Control Risk RatioM-H, Random, 95% CI

Risk RatioM-H, Random, 95%CI in all-cause mortality riskWeight

Study or SubgroupAnker et al., 2009 - FAIR-HFPonikowski et al. 2015 - CONFIRM-HF


Events5

12

17

Total304152

456

Events4

14

18

Total155151

306

Weight24.3%75.7%

100.0%

M-H, Random, 95% CI0.64 [0.17, 2.34]0.85 [0.41, 1.78]

0.79 [0.42, 1.51]

Year20092015



Supplementary Figure 114. Forest plot of iron supplementation and all-cause mortality risk. M-H, Manthel-Haenszel. *Anker et al., 2009 - To treat iron deficiency in heart failure patients; *Ponikowski et al., 2015 - To treat iron deficiency in heart failure patients. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

134

Random Effects


CVD Total CVD 2 16,349 1,941 0.95 [0.77, 1.17] 62% 0.61MI 3 16,414 772 0.95 [0.82, 1.09] 0% 0.43Stroke 2 16,349 666 0.86 [0.46, 1.62] 59% 0.64Total CVD mortality 3 17,351 984 0.94 [0.83, 1.06] 0% 0.30MI mortality 1 14,641 70 0.63 [0.39, 1.02] N/A 0.06


ALL-CAUSE MORTALITY


Favours multivitamins/ Favours control/ Negative association Positive association


0.0 0.5 1.0 1.5 2.0

Supplementary Figure 115. Summary of the pooled effect estimates of RCTs assessing the relationship between multivitamins supplementation and CVD and all cause mortality risk. RR, risk ratio; CIs, confidence intervals; RTCs, randomized clinical trials; CVD, cardiovascular disease; CHD, coronary heart disease; MI, myocardial infarction; N/A, not applicable. The diamonds represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% CIs, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsSesso et al., 2012 - PHS II 876 7,317 856 7,324 65.0% 1.02 [0.94, 1.12]Lamas et al., 2013 - TACT 94 853 115 855 35.0% 0.82 [0.63, 1.06]

Total (95% CI) 8,170 8,179 100% 0.95 [0.77, 1.17]Total events 970 971 Heterogeneity: Tau² = 0.02; Chi² = 2.62, df = 1 (P = 0.11); I² = 62% 0.5 0.7 1 1.5 2Test for overall effect: Z = 0.51 (P = 0.61) Favours multivitamin Favours control

Multivitamins ControlWeight



Study or SubgroupSesso et al., 2012 - PHS IILamas et al., 2013 - TACT


Events876

94

970

Total7317

853

8170

Events856115

971

Total7324

855

8179

Weight65.0%35.0%

100.0%

M-H, Random, 95% CI1.02 [0.94, 1.12]0.82 [0.63, 1.06]

0.95 [0.77, 1.17]

Year20122013

Multivitamins Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.5 0.7 1 1.5 2Favours multivitamins Favours control

Supplementary Figure 116. Forest plot of multivitamins supplementation and total CVD risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

135

Subgroup and Study, Year Events Total Events TotalRCTsBogden et al., 1994 0 33 1 32 0.2% 0.32 [0.01, 7.66]Sesso et al., 2012 - PHS II 317 7,317 335 7,324 84.1% 0.95 [0.82, 1.10]Lamas et al., 2013 - TACT 58 853 61 855 15.8% 0.95 [0.67, 1.35]


Favours multivitamin Favours control




Study or SubgroupBogden et al., 1994Sesso et al., 2012 - PHS IILamas et al., 2013 - TACT


Events0

31758

375

Total33

7317853

8203

Events1

33561

397

Total32

7324855

8211

Weight0.2%

84.1%15.8%

100.0%

M-H, Random, 95% CI0.32 [0.01, 7.66]0.95 [0.82, 1.10]0.95 [0.67, 1.35]

0.95 [0.82, 1.09]

Year199420122013


0.2 0.5 1 2 5Favours multivitamins Favours control

Supplementary Figure 117. Forest plot of multivitamins supplementation and MI risk. M-H, Manthel-Haenszel, MI, myocardial infarction. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsSesso et al., 2012 - PHS II 332 7,317 311 7,324 69.1% 1.07 [0.92, 1.24]Lamas et al., 2013 - TACT 8 853 15 855 30.9% 0.53 [0.23, 1.25]






Study or SubgroupSesso et al., 2012 - PHS IILamas et al., 2013 - TACT


Events332

8

340

Total7317

853

8170

Events311

15

326

Total7324

855

8179

Weight69.1%30.9%

100.0%

M-H, Random, 95% CI1.07 [0.92, 1.24]0.53 [0.23, 1.25]

0.86 [0.46, 1.62]

Year20122013



Supplementary Figure 118. Forest plot of multivitamins supplementation and stroke risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsCTNS 2008 23 501 31 501 5.4% 0.74 [0.44, 1.25]Sesso et al., 2012 - PHS II 408 7,317 421 7,324 84.5% 0.97 [0.85, 1.11]Lamas et al., 2013 - TACT 45 853 56 855 10.2% 0.81 [0.55, 1.18]






Study or SubgroupCTNS 2008Sesso et al., 2012 - PHS IILamas et al., 2013 - TACT


Events23

40845

476

Total501

7317853

8671

Events31

42156

508

Total501

7324855

8680

Weight5.4%

84.5%10.2%

100.0%

M-H, Random, 95% CI0.74 [0.44, 1.25]0.97 [0.85, 1.11]0.81 [0.55, 1.18]

0.94 [0.83, 1.06]

Year200820122013


0.5 0.7 1 1.5 2Favours multivitamins Favours control

Supplementary Figure 119. Forest plot of multivitamins supplementation and CVD mortality risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered

136

to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsSesso et al., 2012 - PHS II 27 7,317 43 7,324 100% 0.63 [0.39, 1.02]

Total (95% CI) 7,317 7,324 100% 0.63 [0.39, 1.02]Total events 27 43 Heterogeneity: Not applicableTest for overall effect: Z = 1.90 (P = 0.06) 0.05 0.2 1 5 20





Study or SubgroupSesso et al., 2012 - PHS II


Events27

27

Total7317

7317

Events43

43

Total7324

7324

Weight100.0%

100.0%

M-H, Random, 95% CI0.63 [0.39, 1.02]

0.63 [0.39, 1.02]

Year2012



Supplementary Figure 120. Forest plot of multivitamins supplementation and MI mortality risk. M-H, Manthel-Haenszel, MI, myocardial infarction. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsLi et al., 1993 - NIT2 22 1,657 35 1,661 43.7% 0.63 [0.37, 1.07]Sesso et al., 2012 - PHS II 89 7,317 76 7,324 56.3% 1.17 [0.86, 1.59]






Study or SubgroupLi et al., 1993 - NIT2Sesso et al., 2012 - PHS II


Events2289

111

Total16577317

8974

Events3576

111

Total16617324

8985

Weight43.7%56.3%

100.0%

M-H, Random, 95% CI0.63 [0.37, 1.07]1.17 [0.86, 1.59]

0.89 [0.49, 1.63]

Year19932012



Supplementary Figure 121. Forest plot of multivitamins supplementation and stroke mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsChandra et al., 1992 0 48 2 48 0.0% 0.20 [0.01, 4.06]Li et al., 1993 - NIT2 157 1,657 167 1,661 8.1% 0.94 [0.77, 1.16]Bogden et al., 1994 0 33 1 32 0.0% 0.32 [0.01, 7.66]Pike et al., 1995 1 24 0 23 0.0% 2.88 [0.12, 67.29]Graat et al., 2002 0 163 5 153 0.0% 0.09 [0.00, 1.53]Avenell et al., 2005 - MAVIS 8 456 4 454 0% 1.99 [0.60, 6.57]Liu et al., 2007 96 375 97 373 5.90% 0.98 [0.77, 1.26]CTNS 2008 77 510 81 510 4% 0.95 [0.71, 1.27]Sesso et al., 2012 - PHS II 1,345 7,317 1,412 7,324 77% 0.95 [0.89, 1.02]Lamas et al., 2013 - TACT 87 853 93 855 5% 0.94 [0.71, 1.24]






Study or SubgroupChandra et al., 1992Li et al., 1993 - NIT2Bogden et al., 1994Pike et al., 1995Graat et al., 2002Avenell et al., 2005 - MAVISLiu et al., 2007CTNS 2008Sesso et al., 2012 - PHS IILamas et al., 2013 - TACT


Events0

1570108

9677

134587

1771

Total48

16573324

163456375510

7317853

11436

Events2

1671054

9781

141293

1862

Total48

16613223

153454373510

7324855

11433

Weight0.0%8.1%0.0%0.0%0.0%0.2%5.9%4.2%

76.8%4.5%

100.0%

M-H, Random, 95% CI0.20 [0.01, 4.06]0.94 [0.77, 1.16]0.32 [0.01, 7.66]

2.88 [0.12, 67.29]0.09 [0.00, 1.53]1.99 [0.60, 6.57]0.98 [0.77, 1.26]0.95 [0.71, 1.27]0.95 [0.89, 1.02]0.94 [0.71, 1.24]

0.95 [0.90, 1.01]

Year1992199319941995200220052007200820122013



Supplementary Figure 122. Forest plot of multivitamins supplementation and all-cause mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-

137

study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects effects model.

Supplementary Figure 123. Funnel plot of multivitamin supplementation and all-cause mortality risk. The vertical line represents the pooled effect estimate expressed as a RR. Dashed lines represent pseudo-95% confidence intervals (CI). The circles represent risk estimates for each study, and the horizontal lines represent standard errors of the RR. We were unable to test for funnel plot asymmetry for other CVD outcomes (<10 RCTs).

138

Random effects

ComparisonRCTs/

CohortsN Events RR (95% CIs) RR (95% CIs) I2 p -value

CVD Total CVD 2 15,494 1,672 1.03 [0.94, 1.12] 0% 0.58Total CHD 2 15,390 445 0.73 [0.20, 2.68] 49% 0.63MI 5 19,097 453 1.14 [0.95, 1.37] 0% 0.15Stroke 7 19,227 484 1.17 [0.98, 1.39] 0% 0.09Total CVD mortality 1 36,282 470 0.92 [0.77, 1.10] N/A 0.38Total CHD mortality 2 36,370 262 0.94 [0.54, 1.64] 10% 0.83MI mortality 1 231 1 2.97 [0.12, 72.26] N/A 0.50Stroke mortality 2 36,370 118 0.90[0.63, 1.29] 0% 0.56



Fa vours cal cium Favours control/ & vitamin D/

Negative as socia tion Pos iti ve association

0.0 1.0 2.0 3.0

Supplementary Figure 124. Summary of the pooled effect estimates of RCTs and assessing the relationship between calcium and vitamin D supplementation and CVD and all cause mortality risk. RR, risk ratio; CIs, confidence intervals; RTCs, randomized clinical trials; CVD, cardiovascular disease; CHD, coronary heart disease; MI, myocardial infarction; N/A, not applicable. The diamonds represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% CIs, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsBrazier et al., 2005 6 95 5 97 0.6% 1.23 [0.39, 3.88]Prentice et al., 2013 - WHI CaD* 848 7,718 813 7,584 99.4% 1.02 [0.94, 1.12]


Favours Ca+VitD Favours control

Calcium & Vitamin D ControlWeight



Study or SubgroupBrazier et al., 2005Prentice et al., 2013 - WHI CaD


Events6

848

854

Total95

7718

7813

Events5

813

818

Total97

7584

7681

Weight0.6%

99.4%

100.0%

M-H, Random, 95% CI1.23 [0.39, 3.88]1.02 [0.94, 1.12]

1.03 [0.94, 1.12]

Year20052013

Calcium & Vitamin D Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.2 0.5 1 2 5Favours Ca/VitD Favours control

Supplementary Figure 125. Forest plot of calcium and vitamin D supplementation and total CVD risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. *Prentice et al., 2013 - removed users of personal calcium or vitamin D supplement use. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

139

Subgroup and Study, Year Events Total Events TotalRCTsInkovaara et al., 1983 1 46 4 42 24.9% 0.23 [0.03, 1.96]Prentice et al., 2013 - WHI CaD* 229 7,718 211 7,584 75.1% 1.07 [0.89, 1.28]






Study or SubgroupInkovaara et al., 1983Prentice et al., 2013 - WHI CaD


Events1

229

230

Total46

7718

7764

Events4

211

215

Total42

7584

7626

Weight24.9%75.1%

100.0%

M-H, Random, 95% CI0.23 [0.03, 1.96]1.07 [0.89, 1.28]

0.73 [0.20, 2.68]

Year19832013

Calcium and Vitamin D Control Risk Ratio Risk RatioM-H, Random, 95% CI


Supplementary Figure 126. Forest plot of calcium and vitamin D supplementation and total CHD risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. *Prentice et al., 2013 - removed users of personal calcium or vitamin D supplement use. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsKomulainen et al., 1999 - OSTPRE 1 116 1 115 0.4% 0.99 [0.06, 15.66]Grant et al., 2005 - RECORD 44 1,306 39 1,332 18.5% 1.15 [0.75, 1.76]Brazier et al., 2005 3 95 0 97 0.4% 7.15 [0.37, 136.50]Lappe et al., 2007 3 446 2 288 1.0% 0.97 [0.16, 5.76]Prentice et al., 2013 - WHI CaD* 193 7,718 167 7,584 79.6% 1.14 [0.93, 1.39]

Total (95% CI) 9,681 9,416 100.0% 1.14 [0.95, 1.37]Total events 244 209Heterogeneity: Tau² = 0.00; Chi² = 1.53, df = 4 (P = 0.82); I² = 0% 0.1 0.2 0.5 1 2 5 10Test for overall effect: Z = 1.44 (P = 0.15) Favours Ca+VitD Favours control




Study or SubgroupKomulainen et al., 1999 - OSTREGrant et al., 2005 - RECORDBrazier et al., 2005Lappe et al., 2007Prentice et al., 2013 - WHI CaD


Events1

4433

193

244

Total116

130695

4467718

9681

Events1

3902

167

209

Total115

133297

2887584

9416

Weight0.4%

18.5%0.4%1.0%

79.6%

100.0%

M-H, Random, 95% CI0.99 [0.06, 15.66]1.15 [0.75, 1.76]

7.15 [0.37, 136.50]0.97 [0.16, 5.76]1.14 [0.93, 1.39]

1.14 [0.95, 1.37]

Year19992005200520072013


0.1 0.2 0.5 1 2 5 10Favours Ca/VitD Favours control

Supplementary Figure 127. Forest plot of calcium and vitamin D supplementation and MI risk. M-H, Manthel-Haenszel, MI, myocardial infarction. *Prentice et al., 2013 - removed users of personal calcium or vitamin D supplement use. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

140

Subgroup and Study, Year Events Total Events TotalRCTsInkovaara et al., 1983 9 46 5 42 3.0% 1.64 [0.60, 4.51]Komulainen et al., 1999 - OSTPRE 2 116 1 115 0.5% 1.98 [0.18, 21.56]Brazier et al., 2005 1 95 1 97 0.4% 1.02 [0.06, 16.09]Grant et al., 2005 - RECORD 60 1,306 48 1,332 22.5% 1.27 [0.88, 1.85]Lappe et al., 2007 6 446 4 288 2.0% 0.97 [0.28, 3.40]Gallagher et al., 2012 1 21 0 21 0.3% 3.00 [0.13, 69.70]Prentice et al., 2013 - WHI CaD* 184 7,718 162 7,584 71.3% 1.12 [0.91, 1.38]






Study or SubgroupInkovaara et al., 1983Komulainen et al., 1999 - OSTREBrazier et al., 2005Grant et al., 2005 - RECORDLappe et al., 2007Gallagher et al., 2012Prentice et al., 2013 - WHI CaD


Events921

6061

184

263

Total46

11695

130644621

7718

9748

Events511

4840

162

221

Total42

11597

133228821

7584

9479

Weight3.0%0.5%0.4%

22.5%2.0%0.3%

71.3%

100.0%

M-H, Random, 95% CI1.64 [0.60, 4.51]

1.98 [0.18, 21.56]1.02 [0.06, 16.09]1.27 [0.88, 1.85]0.97 [0.28, 3.40]

3.00 [0.13, 69.70]1.12 [0.91, 1.38]

1.17 [0.98, 1.39]

Year1983199920052005200720122013



Supplementary Figure 128. Forest plot of calcium and vitamin D supplementation and stroke risk. M-H, Manthel-Haenszel. *Prentice et al., 2013 - removed users of personal calcium or vitamin D supplement use. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsLaCroix et al., 2009 - WHI CaD 226 18,176 244 18,106 100.0% 0.92 [0.77, 1.10]

Total (95% CI) 18,176 18,106 100% 0.92 [0.77, 1.10]Total events 226 244 Heterogeneity: Not applicable 0.2 0.5 1 2 5Test for overall effect: Z = 0.88 (P = 0.38) Favours Ca+VitD Favours control




Study or SubgroupLaCroix et al., 2009 - WHI CaD


Events226

226

Total18176

18176

Events244

244

Total18106

18106

Weight100.0%

100.0%

M-H, Random, 95% CI0.92 [0.77, 1.10]

0.92 [0.77, 1.10]

Year2009

Vit D+Ca Control Risk Ratio Risk RatioM-H, Random, 95% CI

0.2 0.5 1 2 5Favours Vit D+Ca Favours Control

Supplementary Figure 129. Forest plot of calcium and vitamin D supplementation and CVD mortality risk. M-H, Manthel-Haenszel, CVD, cardiovascular disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

141

Subgroup and Study, Year Events Total Events TotalRCTsInkovaara et al., 1983 1 46 3 42 5.9% 0.30 [0.03, 2.81]

LaCroix et al., 2009 - WHI CaD 130 18,176 128 18,106 94.1% 1.01 [0.79, 1.29]






Study or SubgroupInkovaara et al., 1983LaCroix et al., 2009 - WHI CaD


Events1

130

131

Total46

18176

18222

Events3

128

131

Total42

18106

18148

Weight5.9%

94.1%

100.0%

M-H, Random, 95% CI0.30 [0.03, 2.81]1.01 [0.79, 1.29]

0.94 [0.54, 1.64]

Year19832009



Supplementary Figure 130. Forest plot of calcium and vitamin D supplementation and CHD mortality risk. M-H, Manthel-Haenszel, CHD, coronary heart disease. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsKomulainen et al., 1999 - OSTPRE 1 116 0 115 100.0% 2.97 [0.12, 72.26]

Total (95% CI) 116 115 100% 2.97 [0.12, 72.26]

Total events 1 0

Heterogeneity: Not applicable 0.001 0.1 1 10 1000Test for overall effect: Z = 0.67 (P = 0.50) Favours Ca+VitD Favours control




Study or SubgroupKomulainen et al., 1999 - OSTRE


Events1

1

Total116

116

Events0

0

Total115

115

Weight100.0%

100.0%

M-H, Random, 95% CI2.97 [0.12, 72.26]

2.97 [0.12, 72.26]

Year1999


0.001 0.1 1 10 1000Favours Vit D and calcium Favours Control

Supplementary Figure 131. Forest plot of calcium and vitamin D supplementation and MI mortality risk. M-H, Manthel-Haenszel, MI, myocardial infarction. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

Subgroup and Study, Year Events Total Events TotalRCTsInkovaara et al., 1983 2 46 2 42 3.5% 0.91 [0.13, 6.20]LaCroix et al., 2009 - WHI CaD 54 18,176 60 18,106 96.5% 0.90 [0.62, 1.29]






Study or SubgroupInkovaara et al., 1983LaCroix et al., 2009 - WHI CaD


Events2

54

56

Total46

18176

18222

Events2

60

62

Total42

18106

18148

Weight3.5%

96.5%

100.0%

M-H, Random, 95% CI0.91 [0.13, 6.20]0.90 [0.62, 1.29]

0.90 [0.63, 1.29]

Year19832009



Supplementary Figure 132. Forest plot of calcium and vitamin D supplementation and stroke mortality risk. M-H, Manthel-Haenszel. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

142

Subgroup and Study, Year Events Total Events TotalRCTsInkovaara et al., 1983 7 46 5 42 0.3% 1.28 [0.44, 3.72]Chapuy et al., 1992 258 1,634 274 1,636 14.8% 0.94 [0.81, 1.10]Dawson-Hughes et al., 1997* 2 187 2 202 0.1% 1.08 [0.15, 7.59]Baeksgaard et al., 1998 0 80 1 80 0.0% 0.33 [0.01, 8.06]Komulainen et al., 1999 1 116 1 115 0.0% 0.99 [0.06, 15.66]Krieg et al., 1999 21 124 26 124 1.3% 0.81 [0.48, 1.36]Chapuy et al., 2002 - Decalyos II 67 393 43 190 3.1% 0.75 [0.54, 1.06]Harwood et al., 2004 - NoNOF 6 39 5 37 0.3% 1.14 [0.38, 3.41]Meier et al., 2004 0 30 1 25 0.0% 0.28 [0.01, 6.58]Larsen et al., 2004† 832 4,957 839 4,648 47.0% 0.93 [0.85, 1.01]Grant et al., 2005 - RECORD 221 1306 217 1332 12.2% 1.04 [0.88, 1.23]Porthouse et al., 2005 57 1,321 68 1,993 3.0% 1.26 [0.90, 1.79]Brazier et al., 2005 3 95 1 97 0.1% 3.06 [0.32, 28.93]Daly et al., 2006 1 85 0 82 0.0% 2.90 [0.12, 70.07]Bolton-Smith et al., 2007‡ 0 62 1 61 0.0% 0.33 [0.01, 7.90]Zhu et al., 2008 - CAIFOS 0 39 2 41 0.0% 0.21 [0.01, 4.24]Salovaara et al., 2010 - OSTRE-FPS 15 1,586 13 1,609 0.7% 1.17 [0.56, 2.45]Prentice et al., 2013 - WHI CaD§ 331 7,718 338 7,584 16.3% 0.96 [0.83, 1.12]Gupta et al., 2016 4 25 11 28 0.4% 0.41 [0.15, 1.12]Lappe et al., 2017 7 1,156 9 1,147 0.4% 0.77 [0.29, 2.07]



ControlWeight

Risk RatioM-H, Random, 95%

CIRisk Ratio

M-H, Random, 95% CI in all-cause mortality risk Calcium & Vitamin D

Study or SubgroupInkovaara et al., 1983Chapuy et al., 1992Dawson-Hughes et al., 1997Baeksgaard et al., 1998Komulainen et al., 1999Krieg et al., 1999Chapuy et al., 2002 - Decalyos IIHarwood et al., 2004 - NoNOFMeier et al., 2004Larsen et al., 2004Grant et al., 2005 - RECORDPorthouse et al., 2005Brazier et al., 2005Daly et al., 2006Bolton-Smith et al., 2007Zhu et al., 2008 - CAIFOSSalovaara et al., 2010 - OSTRE-FPSPrentice et al., 2013 - WHI CaDGupta et al., 2016Lappe et al., 2017


Events7

258201

2167

60

832221

573100

15331

47

1833

Total46

1634187

80116124393

3930

495713061321

95856239

15867718

251156

20999

Events5

274211

2643

51

839217

681012

13338

119

1857

Total42

1636202

80115124190

3725

464813321993

97826141

16097584

281147

21073

Weight0.3%

14.8%0.1%0.0%0.0%1.3%3.1%0.3%0.0%

47.0%12.2%

3.0%0.1%0.0%0.0%0.0%0.7%

16.3%0.4%0.4%

100.0%

M-H, Random, 95% CI1.28 [0.44, 3.72]0.94 [0.81, 1.10]1.08 [0.15, 7.59]0.33 [0.01, 8.06]

0.99 [0.06, 15.66]0.81 [0.48, 1.36]0.75 [0.54, 1.06]1.14 [0.38, 3.41]0.28 [0.01, 6.58]0.93 [0.85, 1.01]1.04 [0.88, 1.23]1.26 [0.90, 1.79]

3.06 [0.32, 28.93]2.90 [0.12, 70.07]

0.33 [0.01, 7.90]0.21 [0.01, 4.24]1.17 [0.56, 2.45]0.96 [0.83, 1.12]0.41 [0.15, 1.12]0.77 [0.29, 2.07]

0.95 [0.89, 1.01]

Year19831992199719981999199920022004200420042005200520052006200720082010201320162017

Calcium and Vitamin D Control Risk Ratio Risk RatioM-H, Random, 95% CI


Supplementary Figure 133. Forest plot of calcium and vitamin D supplementation and all-cause mortality risk. M-H, Manthel-Haenszel. *Dawson-Hughes et al., 1997 - Data taken from meta-analysis Bolland et al., 2014, Bjelakovic et al., 2014; †Larsen et al., 2004 - Data taken from meta-analysis Bjelakovic et al., 2014; ‡Bolton-Smith et al., 2007 - Data taken from meta-analysis Bjelakovic et al., 2014. The diamond represents the pooled risk estimate. Inter-study heterogeneity was tested using the Cochran Q statistic (Chi2) at a significance level of P < 0.10, and quantified by the I2 statistic. An I2 value ≥ 50% is considered to indicate substantial heterogeneity. All results are presented as risk ratios with 95% Confidence Intervals, using the Manthel-Haenszel method with random effects model.

143

Supplementary Figure 134. Funnel plot of vitamin D & calcium supplementation and all-cause mortality risk. The vertical line represents the pooled effect estimate expressed as a RR. Dashed lines represent pseudo-95% confidence intervals (CI). The circles represent risk estimates for each study, and the horizontal lines represent standard errors of the RR. We were unable to test for funnel plot asymmetry for other CVD outcomes (<10 RCTs).

144

References

1. Anand SS, Hawkes C, de Souza RJ et al. Food Consumption and its Impact on Cardiovascular Disease: Importance of Solutions Focused on the Globalized Food System: A Report From the Workshop Convened by the World Heart Federation. Journal of the American College of Cardiology 2015;66:1590-614.

2. Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III). Jama 2001;285:2486-97.

3. Grundy SM, Cleeman JI, Merz CN et al. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines. Circulation 2004;110:227-39.

4. Stone NJ, Robinson JG, Lichtenstein AH et al. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation 2014;129:S1-45.

5. Anderson TJ, Gregoire J, Hegele RA et al. 2012 update of the Canadian Cardiovascular Society guidelines for the diagnosis and treatment of dyslipidemia for the prevention of cardiovascular disease in the adult. The Canadian journal of cardiology 2013;29:151-67.

6. Anderson TJ, Gregoire J, Pearson GJ et al. 2016 Canadian Cardiovascular Society Guidelines for the Management of Dyslipidemia for the Prevention of Cardiovascular Disease in the Adult. The Canadian journal of cardiology 2016;32:1263-1282.

7. Reiner Z, Catapano AL, De Backer G et al. ESC/EAS Guidelines for the management of dyslipidaemias: the Task Force for the management of dyslipidaemias of the European Society of Cardiology (ESC) and the European Atherosclerosis Society (EAS). European heart journal 2011;32:1769-818.

8. Stroes ES, Thompson PD, Corsini A et al. Statin-associated muscle symptoms: impact on statin therapy-European Atherosclerosis Society Consensus Panel Statement on Assessment, Aetiology and Management. European heart journal 2015;36:1012-22.

9. British Heart Foundation. Healthy Eating. https://www.bhf.org.uk/heart-health/preventing-heart-disease/healthy-eating. Accessed December 6, 2016.

10. Moyer VA. Vitamin, mineral, and multivitamin supplements for the primary prevention of cardiovascular disease and cancer: U.S. Preventive services Task Force recommendation statement. Annals of internal medicine 2014;160:558-64.

11. World Cancer Research Fund International. Our cancer prevention recommendations. http://www.wcrf.org/int/research-we-fund/our-cancer-prevention-recommendations. Accessed December 6, 2016.

12. Kushi LH, Doyle C, McCullough M et al. American Cancer Society Guidelines on nutrition and physical activity for cancer prevention: reducing the risk of cancer with healthy food choices and physical activity. CA: a cancer journal for clinicians 2012;62:30-67.

13. Canadian Cancer Society. Nutrition and Fitness (Healthy Eating VDhwccep-a-sl-wn-a-froAD.

14. Norat T, Scoccianti C, Boutron-Ruault MC et al. European Code against Cancer 4th Edition: Diet and cancer. Cancer epidemiology 2015;39 Suppl 1:S56-66.

15. Cancer Research UK. Diet and Cancer. http://www.cancerresearchuk.org/about-cancer/causes-of-cancer/diet-and-cancer. Accessed December 7, 2016.

16. Brohult J, Jonson B. Effects of large doses of calciferol on patients with rheumatoid arthritis. A double-blind clinical trial. Scandinavian journal of rheumatology 1973;2:173-6.

17. Clofibrate and niacin in coronary heart disease. Jama 1975;231:360-81.

145

http://www.cancerresearchuk.org/about-cancer/causes-of-cancer/diet-and-cancer

http://www.cancerresearchuk.org/about-cancer/causes-of-cancer/diet-and-cancer

http://www.wcrf.org/int/research-we-fund/our-cancer-prevention-recommendations

https://www.bhf.org.uk/heart-health/preventing-heart-disease/healthy-eating

https://www.bhf.org.uk/heart-health/preventing-heart-disease/healthy-eating

18. Gillilan RE, Mondell B, Warbasse JR. Quantitative evaluation of vitamin E in the treatment of angina pectoris. Am Heart J 1977;93:444-9.

19. Inkovaara J, Gothoni G, Halttula R, Heikinheimo R, Tokola O. Calcium, vitamin D and anabolic steroid in treatment of aged bones: double-blind placebo-controlled long-term clinical trial. Age and ageing 1983;12:124-30.

20. Corless D, Dawson E, Fraser F et al. Do vitamin D supplements improve the physical capabilities of elderly hospital patients? Age and ageing 1985;14:76-84.

21. Aloia JF, Vaswani A, Yeh JK, Ellis K, Yasumura S, Cohn SH. Calcitriol in the treatment of postmenopausal osteoporosis. The American journal of medicine 1988;84:401-8.

22. Korpela H, Kumpulainen J, Jussila E et al. Effect of selenium supplementation after acute myocardial infarction. Research communications in chemical pathology and pharmacology 1989;65:249-52.

23. McKeown-Eyssen G, Holloway C, Jazmaji V, Bright-See E, Dion P, Bruce WR. A randomized trial of vitamins C and E in the prevention of recurrence of colorectal polyps. Cancer Res 1988;48:4701-5.

24. Ott SM, Chesnut CH, 3rd. Calcitriol treatment is not effective in postmenopausal osteoporosis. Annals of internal medicine 1989;110:267-74.

25. Gallagher JC, Goldgar D. Treatment of postmenopausal osteoporosis with high doses of synthetic calcitriol. A randomized controlled study. Annals of internal medicine 1990;113:649-55.

26. Grady D, Halloran B, Cummings S et al. 1,25-Dihydroxyvitamin D3 and muscle strength in the elderly: a randomized controlled trial. The Journal of clinical endocrinology and metabolism 1991;73:1111-7.

27. Chapuy MC, Arlot ME, Duboeuf F et al. Vitamin D3 and calcium to prevent hip fractures in the elderly women. The New England journal of medicine 1992;327:1637-42.

28. Li JY, Taylor PR, Li B et al. Nutrition intervention trials in Linxian, China: multiple vitamin/mineral supplementation, cancer incidence, and disease-specific mortality among adults with esophageal dysplasia. Journal of the National Cancer Institute 1993;85:1492-8.

29. Bogden JD, Bendich A, Kemp FW et al. Daily micronutrient supplements enhance delayed-hypersensitivity skin test responses in older people. The American journal of clinical nutrition 1994;60:437-47.

30. de la Maza MP, Petermann M, Bunout D, Hirsch S. Effects of long-term vitamin E supplementation in alcoholic cirrhotics. J Am Coll Nutr 1995;14:192-6.

31. Hamdy NA, Kanis JA, Beneton MN et al. Effect of alfacalcidol on natural course of renal bone disease in mild to moderate renal failure. BMJ (Clinical research ed) 1995;310:358-63.

32. Ooms ME, Roos JC, Bezemer PD, van der Vijgh WJ, Bouter LM, Lips P. Prevention of bone loss by vitamin D supplementation in elderly women: a randomized double-blind trial. The Journal of clinical endocrinology and metabolism 1995;80:1052-8.

33. Pike J, Chandra RK. Effect of vitamin and trace element supplementation on immune indices in healthy elderly. International journal for vitamin and nutrition research Internationale Zeitschrift fur Vitamin- und Ernahrungsforschung Journal international de vitaminologie et de nutrition 1995;65:117-21.

34. Steiner M, Glantz M, Lekos A. Vitamin E plus aspirin compared with aspirin alone in patients with transient ischemic attacks. Am J Clin Nutr 1995;62:1381s-1384s.

35. Greenberg ER, Baron JA, Karagas MR et al. Mortality associated with low plasma concentration of beta carotene and the effect of oral supplementation. Jama 1996;275:699-703.

146

36. Hennekens CH, Buring JE, Manson JE et al. Lack of effect of long-term supplementation with beta carotene on the incidence of malignant neoplasms and cardiovascular disease. The New England journal of medicine 1996;334:1145-9.

37. Lips P, Graafmans WC, Ooms ME, Bezemer PD, Bouter LM. Vitamin D supplementation and fracture incidence in elderly persons. A randomized, placebo-controlled clinical trial. Annals of internal medicine 1996;124:400-6.

38. Omenn GS, Goodman GE, Thornquist MD et al. Effects of a combination of beta carotene and vitamin A on lung cancer and cardiovascular disease. The New England journal of medicine 1996;334:1150-5.

39. Stephens NG, Parsons A, Schofield PM, Kelly F, Cheeseman K, Mitchinson MJ. Randomised controlled trial of vitamin E in patients with coronary disease: Cambridge Heart Antioxidant Study (CHAOS). Lancet (London, England) 1996;347:781-6.

40. Dawson-Hughes B, Harris SS, Krall EA, Dallal GE. Effect of calcium and vitamin D supplementation on bone density in men and women 65 years of age or older. The New England journal of medicine 1997;337:670-6.

41. Girodon F, Lombard M, Galan P et al. Effect of micronutrient supplementation on infection in institutionalized elderly subjects: a controlled trial. Annals of nutrition & metabolism 1997;41:98-107.

42. Moon TE, Levine N, Cartmel B et al. Effect of retinol in preventing squamous cell skin cancer in moderate-risk subjects: a randomized, double-blind, controlled trial. Southwest Skin Cancer Prevention Study Group. Cancer epidemiology, biomarkers & prevention : a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology 1997;6:949-56.

43. Sano M, Ernesto C, Thomas RG et al. A controlled trial of selegiline, alpha-tocopherol, or both as treatment for Alzheimer's disease. The Alzheimer's Disease Cooperative Study. The New England journal of medicine 1997;336:1216-22.

44. Sato Y, Maruoka H, Oizumi K. Amelioration of hemiplegia-associated osteopenia more than 4 years after stroke by 1 alpha-hydroxyvitamin D3 and calcium supplementation. Stroke 1997;28:736-9.

45. Baeksgaard L, Andersen KP, Hyldstrup L. Calcium and vitamin D supplementation increases spinal BMD in healthy, postmenopausal women. Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 1998;8:255-60.

46. Mortality in DATATOP: a multicenter trial in early Parkinson's disease. Parkinson Study Group. Annals of neurology 1998;43:318-25.

47. Virtamo J, Rapola JM, Ripatti S et al. Effect of vitamin E and beta carotene on the incidence of primary nonfatal myocardial infarction and fatal coronary heart disease. Archives of internal medicine 1998;158:668-75.

48. Baron JA, Beach M, Mandel JS et al. Calcium supplements for the prevention of colorectal adenomas. Calcium Polyp Prevention Study Group. The New England journal of medicine 1999;340:101-7.

49. Hoffman RM, Garewal HS. Alpha-tocopherol supplementation for men with existing coronary artery disease: a feasibility study. Preventive medicine 1999;29:112-8.

50. Girodon F, Galan P, Monget AL et al. Impact of trace elements and vitamin supplementation on immunity and infections in institutionalized elderly patients: a randomized controlled trial. MIN. VIT. AOX. geriatric network. Archives of internal medicine 1999;159:748-54.

51. Dietary supplementation with n-3 polyunsaturated fatty acids and vitamin E after myocardial infarction: results of the GISSI-Prevenzione trial. Gruppo Italiano per lo Studio della Sopravvivenza nell'Infarto miocardico. Lancet (London, England) 1999;354:447-55.

147

52. Green A, Williams G, Neale R et al. Daily sunscreen application and betacarotene supplementation in prevention of basal-cell and squamous-cell carcinomas of the skin: a randomised controlled trial. Lancet (London, England) 1999;354:723-9.

53. Komulainen M, Kroger H, Tuppurainen MT et al. Prevention of femoral and lumbar bone loss with hormone replacement therapy and vitamin D3 in early postmenopausal women: a population-based 5-year randomized trial. The Journal of clinical endocrinology and metabolism 1999;84:546-52.

54. Krieg MA, Jacquet AF, Bremgartner M, Cuttelod S, Thiebaud D, Burckhardt P. Effect of supplementation with vitamin D3 and calcium on quantitative ultrasound of bone in elderly institutionalized women: a longitudinal study. Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 1999;9:483-8.

55. Sato Y, Manabe S, Kuno H, Oizumi K. Amelioration of osteopenia and hypovitaminosis D by 1alpha-hydroxyvitamin D3 in elderly patients with Parkinson's disease. Journal of neurology, neurosurgery, and psychiatry 1999;66:64-8.

56. Boaz M, Smetana S, Weinstein T et al. Secondary prevention with antioxidants of cardiovascular disease in endstage renal disease (SPACE): randomised placebo-controlled trial. Lancet (London, England) 2000;356:1213-8.

57. Correa P, Fontham ET, Bravo JC et al. Chemoprevention of gastric dysplasia: randomized trial of antioxidant supplements and anti-helicobacter pylori therapy. Journal of the National Cancer Institute 2000;92:1881-8.

58. Frazao JM, Elangovan L, Maung HM et al. Intermittent doxercalciferol (1alpha-hydroxyvitamin D(2)) therapy for secondary hyperparathyroidism. American journal of kidney diseases : the official journal of the National Kidney Foundation 2000;36:550-61.

59. Lee IM, Cook NR, Manson JE, Buring JE, Hennekens CH. beta-Carotene Supplementation and Incidence of Cancer and Cardiovascular Disease: the Women's Health Study. Journal of the National Cancer Institute 1999;91:2102-2106.

60. Leppala JM, Virtamo J, Fogelholm R et al. Controlled trial of alpha-tocopherol and beta-carotene supplements on stroke incidence and mortality in male smokers. Arteriosclerosis, thrombosis, and vascular biology 2000;20:230-5.

61. Jacobson JS, Begg MD, Wang LW et al. Effects of a 6-month vitamin intervention on DNA damage in heavy smokers. Cancer Epidemiol Biomarkers Prev 2000;9:1303-11.

62. Salonen JT, Nyyssonen K, Salonen R et al. Antioxidant Supplementation in Atherosclerosis Prevention (ASAP) study: a randomized trial of the effect of vitamins E and C on 3-year progression of carotid atherosclerosis. Journal of internal medicine 2000;248:377-86.

63. A randomized, placebo-controlled, clinical trial of high-dose supplementation with vitamins C and E and beta carotene for age-related cataract and vision loss: AREDS report no. 9. Archives of ophthalmology (Chicago, Ill : 1960) 2001;119:1439-52.

64. Brown BG, Zhao XQ, Chait A et al. Simvastatin and niacin, antioxidant vitamins, or the combination for the prevention of coronary disease. The New England journal of medicine 2001;345:1583-92.

65. de Gaetano G. Low-dose aspirin and vitamin E in people at cardiovascular risk: a randomised trial in general practice. Collaborative Group of the Primary Prevention Project. Lancet (London, England) 2001;357:89-95.

66. Desnuelle C, Dib M, Garrel C, Favier A. A double-blind, placebo-controlled randomized clinical trial of alpha-tocopherol (vitamin E) in the treatment of amyotrophic lateral sclerosis. ALS riluzole-tocopherol Study Group. Amyotrophic lateral sclerosis and other motor neuron disorders : official publication of the World Federation of Neurology, Research Group on Motor Neuron Diseases 2001;2:9-18.

148

67. de Waart FG, Kok FJ, Smilde TJ, Hijmans A, Wollersheim H, Stalenhoef AF. Effect of glutathione S-transferase M1 genotype on progression of atherosclerosis in lifelong male smokers. Atherosclerosis 2001;158:227-31.

68. Gallagher JC, Fowler SE, Detter JR, Sherman SS. Combination treatment with estrogen and calcitriol in the prevention of age-related bone loss. The Journal of clinical endocrinology and metabolism 2001;86:3618-28.

69. You WC, Chang YS, Heinrich J et al. An intervention trial to inhibit the progression of precancerous gastric lesions: compliance, serum micronutrients and S-allyl cysteine levels, and toxicity. European journal of cancer prevention : the official journal of the European Cancer Prevention Organisation (ECP) 2001;10:257-63.

70. Baker F PD, Blackwood S, Hunt J, Erskine M, Dyas M, Ashby M, Siva A, Brown MJ. . Blinded comparison of folic acid and placebo in patients with ischaemic heart disease: an outcome trial. . Circulation 2002;106.

71. Chapuy MC, Pamphile R, Paris E et al. Combined calcium and vitamin D3 supplementation in elderly women: confirmation of reversal of secondary hyperparathyroidism and hip fracture risk: the Decalyos II study. Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 2002;13:257-64.

72. Chylack LT, Jr., Brown NP, Bron A et al. The Roche European American Cataract Trial (REACT): a randomized clinical trial to investigate the efficacy of an oral antioxidant micronutrient mixture to slow progression of age-related cataract. Ophthalmic epidemiology 2002;9:49-80.

73. Graat JM, Schouten EG, Kok FJ. Effect of daily vitamin E and multivitamin-mineral supplementation on acute respiratory tract infections in elderly persons: a randomized controlled trial. Jama 2002;288:715-21.

74. Hodis HN, Mack WJ, LaBree L et al. Alpha-tocopherol supplementation in healthy individuals reduces low-density lipoprotein oxidation but not atherosclerosis: the Vitamin E Atherosclerosis Prevention Study (VEAPS). Circulation 2002;106:1453-9.

75. MRC/BHF Heart Protection Study of antioxidant vitamin supplementation in 20,536 high-risk individuals: a randomised placebo-controlled trial. Lancet (London, England) 2002;360:23-33.

76. Meyer HE, Smedshaug GB, Kvaavik E, Falch JA, Tverdal A, Pedersen JI. Can vitamin D supplementation reduce the risk of fracture in the elderly? A randomized controlled trial. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2002;17:709-15.

77. Schnyder G, Roffi M, Flammer Y, Pin R, Hess OM. Effect of homocysteine-lowering therapy with folic acid, vitamin B12, and vitamin B6 on clinical outcome after percutaneous coronary intervention: the Swiss Heart study: a randomized controlled trial. Jama 2002;288:973-9.

78. Waters DD, Alderman EL, Hsia J et al. Effects of hormone replacement therapy and antioxidant vitamin supplements on coronary atherosclerosis in postmenopausal women: a randomized controlled trial. Jama 2002;288:2432-40.

79. Wluka AE, Stuckey S, Brand C, Cicuttini FM. Supplementary vitamin E does not affect the loss of cartilage volume in knee osteoarthritis: a 2 year double blind randomized placebo controlled study. The Journal of rheumatology 2002;29:2585-91.

80. Collins EG, Edwin Langbein W, Orebaugh C et al. PoleStriding exercise and vitamin E for management of peripheral vascular disease. Medicine and science in sports and exercise 2003;35:384-93.

149

81. Cooper L, Clifton-Bligh PB, Nery ML et al. Vitamin D supplementation and bone mineral density in early postmenopausal women. The American journal of clinical nutrition 2003;77:1324-9.

82. Liem A, Reynierse-Buitenwerf GH, Zwinderman AH, Jukema JW, van Veldhuisen DJ. Secondary prevention with folic acid: effects on clinical outcomes. Journal of the American College of Cardiology 2003;41:2105-13.

83. Righetti M, Ferrario GM, Milani S et al. Effects of folic acid treatment on homocysteine levels and vascular disease in hemodialysis patients. Medical science monitor : international medical journal of experimental and clinical research 2003;9:Pi19-24.

84. Trivedi DP, Doll R, Khaw KT. Effect of four monthly oral vitamin D3 (cholecalciferol) supplementation on fractures and mortality in men and women living in the community: randomised double blind controlled trial. BMJ (Clinical research ed) 2003;326:469.

85. Virtamo J, Pietinen P, Huttunen JK et al. Incidence of cancer and mortality following alpha-tocopherol and beta-carotene supplementation: a postintervention follow-up. Jama 2003;290:476-85.

86. Dukas L, Bischoff HA, Lindpaintner LS et al. Alfacalcidol reduces the number of fallers in a community-dwelling elderly population with a minimum calcium intake of more than 500 mg daily. Journal of the American Geriatrics Society 2004;52:230-6.

87. Harwood RH, Sahota O, Gaynor K, Masud T, Hosking DJ. A randomised, controlled comparison of different calcium and vitamin D supplementation regimens in elderly women after hip fracture: The Nottingham Neck of Femur (NONOF) Study. Age and ageing 2004;33:45-51.

88. Coburn JW, Maung HM, Elangovan L et al. Doxercalciferol safely suppresses PTH levels in patients with secondary hyperparathyroidism associated with chronic kidney disease stages 3 and 4. American journal of kidney diseases : the official journal of the National Kidney Foundation 2004;43:877-90.

89. Lange H, Suryapranata H, De Luca G et al. Folate therapy and in-stent restenosis after coronary stenting. The New England journal of medicine 2004;350:2673-81.

90. Larsen ER, Mosekilde L, Foldspang A. Vitamin D and calcium supplementation prevents osteoporotic fractures in elderly community dwelling residents: a pragmatic population-based 3-year intervention study. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2004;19:370-8.

91. Liem AH, van Boven AJ, Veeger NJ et al. Efficacy of folic acid when added to statin therapy in patients with hypercholesterolemia following acute myocardial infarction: a randomised pilot trial. International journal of cardiology 2004;93:175-9.

92. Taylor AJ, Sullenberger LE, Lee HJ, Lee JK, Grace KA. Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol (ARBITER) 2: a double-blind, placebo-controlled study of extended-release niacin on atherosclerosis progression in secondary prevention patients treated with statins. Circulation 2004;110:3512-7.

93. Tornwall ME, Virtamo J, Korhonen PA et al. Effect of alpha-tocopherol and beta-carotene supplementation on coronary heart disease during the 6-year post-trial follow-up in the ATBC study. European heart journal 2004;25:1171-8.

94. Manuel-y-Keenoy B, Vinckx M, Vertommen J, van Gaal L, de Leeuw I. Impact of vitamin E supplementation on lipoprotein peroxidation and composition in type 1 diabetic patients treated with atorvastatin. Atherosclerosis 2004;175:369-376.

95. McNeil JJ, Robman L, Tikellis G, Sinclair MI, McCarty CA, Taylor HR. Vitamin E supplementation and cataract: randomized controlled trial. Ophthalmology 2004;111:75-84.

96. Meier C, Woitge HW, Witte K, Lemmer B, Seibel MJ. Supplementation with oral vitamin D3 and calcium during winter prevents seasonal bone loss: a randomized controlled

150

open-label prospective trial. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2004;19:1221-30.

97. Meydani SN, Leka LS, Fine BC et al. Vitamin E and respiratory tract infections in elderly nursing home residents: a randomized controlled trial. Jama 2004;292:828-36.

98. Aloia JF, Talwar SA, Pollack S, Yeh J. A randomized controlled trial of vitamin D3 supplementation in African American women. Archives of internal medicine 2005;165:1618-23.

99. Avenell A, Campbell MK, Cook JA et al. Effect of multivitamin and multimineral supplements on morbidity from infections in older people (MAVIS trial): pragmatic, randomised, double blind, placebo controlled trial. BMJ (Clinical research ed) 2005;331:324-9.

100. Brazier M, Grados F, Kamel S et al. Clinical and laboratory safety of one year's use of a combination calcium + vitamin D tablet in ambulatory elderly women with vitamin D insufficiency: results of a multicenter, randomized, double-blind, placebo-controlled study. Clinical therapeutics 2005;27:1885-93.

101. Flicker L, MacInnis RJ, Stein MS et al. Should older people in residential care receive vitamin D to prevent falls? Results of a randomized trial. Journal of the American Geriatrics Society 2005;53:1881-8.

102. Graf M, Ecker D, Horowski R et al. High dose vitamin E therapy in amyotrophic lateral sclerosis as add-on therapy to riluzole: results of a placebo-controlled double-blind study. Journal of neural transmission (Vienna, Austria : 1996) 2005;112:649-60.

103. Grant AM, Avenell A, Campbell MK et al. Oral vitamin D3 and calcium for secondary prevention of low-trauma fractures in elderly people (Randomised Evaluation of Calcium Or vitamin D, RECORD): a randomised placebo-controlled trial. Lancet (London, England) 2005;365:1621-8.

104. Lee IM, Cook NR, Gaziano JM et al. Vitamin E in the primary prevention of cardiovascular disease and cancer: the Women's Health Study: a randomized controlled trial. Jama 2005;294:56-65.

105. Limburg PJ, Wei W, Ahnen DJ et al. Randomized, placebo-controlled, esophageal squamous cell cancer chemoprevention trial of selenomethionine and celecoxib. Gastroenterology 2005;129:863-73.

106. Lonn E, Bosch J, Yusuf S et al. Effects of long-term vitamin E supplementation on cardiovascular events and cancer: a randomized controlled trial. Jama 2005;293:1338-47.

107. Mooney LA, Madsen AM, Tang D et al. Antioxidant vitamin supplementation reduces benzo(a)pyrene-DNA adducts and potential cancer risk in female smokers. Cancer epidemiology, biomarkers & prevention : a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology 2005;14:237-42.

108. Petersen RC, Thomas RG, Grundman M et al. Vitamin E and donepezil for the treatment of mild cognitive impairment. The New England journal of medicine 2005;352:2379-88.

109. Porthouse J, Cockayne S, King C et al. Randomised controlled trial of calcium and supplementation with cholecalciferol (vitamin D3) for prevention of fractures in primary care. BMJ (Clinical research ed) 2005;330:1003.

110. Potena L, Grigioni F, Magnani G et al. Homocysteine-lowering therapy and early progression of transplant vasculopathy: a prospective, randomized, IVUS-based study. American journal of transplantation : official journal of the American Society of Transplantation and the American Society of Transplant Surgeons 2005;5:2258-64.

111. Sato Y, Iwamoto J, Kanoko T, Satoh K. Low-dose vitamin D prevents muscular atrophy and reduces falls and hip fractures in women after stroke: a randomized controlled trial. Cerebrovascular diseases (Basel, Switzerland) 2005;20:187-92.

151

112. Bairati I, Meyer F, Jobin E et al. Antioxidant vitamins supplementation and mortality: a randomized trial in head and neck cancer patients. International journal of cancer 2006;119:2221-4.

113. Bonaa KH, Njolstad I, Ueland PM et al. Homocysteine lowering and cardiovascular events after acute myocardial infarction. The New England journal of medicine 2006;354:1578-88.

114. Coyne D, Acharya M, Qiu P et al. Paricalcitol capsule for the treatment of secondary hyperparathyroidism in stages 3 and 4 CKD. American journal of kidney diseases : the official journal of the National Kidney Foundation 2006;47:263-76.

115. Daly RM, Brown M, Bass S, Kukuljan S, Nowson C. Calcium- and vitamin D3-fortified milk reduces bone loss at clinically relevant skeletal sites in older men: a 2-year randomized controlled trial. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2006;21:397-405.

116. Law M, Withers H, Morris J, Anderson F. Vitamin D supplementation and the prevention of fractures and falls: results of a randomised trial in elderly people in residential accommodation. Age and ageing 2006;35:482-6.

117. Lonn E, Yusuf S, Arnold MJ et al. Homocysteine lowering with folic acid and B vitamins in vascular disease. The New England journal of medicine 2006;354:1567-77.

118. Magliano D, McNeil J, Branley P et al. The Melbourne Atherosclerosis Vitamin E Trial (MAVET): a study of high dose vitamin E in smokers. European journal of cardiovascular prevention and rehabilitation : official journal of the European Society of Cardiology, Working Groups on Epidemiology & Prevention and Cardiac Rehabilitation and Exercise Physiology 2006;13:341-7.

119. Schleithoff SS, Zittermann A, Tenderich G, Berthold HK, Stehle P, Koerfer R. Vitamin D supplementation improves cytokine profiles in patients with congestive heart failure: a double-blind, randomized, placebo-controlled trial. The American journal of clinical nutrition 2006;83:754-9.

120. Stranges S, Marshall JR, Trevisan M et al. Effects of selenium supplementation on cardiovascular disease incidence and mortality: secondary analyses in a randomized clinical trial. American journal of epidemiology 2006;163:694-9.

121. Zoungas S, McGrath BP, Branley P et al. Cardiovascular morbidity and mortality in the Atherosclerosis and Folic Acid Supplementation Trial (ASFAST) in chronic renal failure: a multicenter, randomized, controlled trial. Journal of the American College of Cardiology 2006;47:1108-16.

122. Bolton-Smith C, McMurdo ME, Paterson CR et al. Two-year randomized controlled trial of vitamin K1 (phylloquinone) and vitamin D3 plus calcium on the bone health of older women. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2007;22:509-19.

123. Catalano M, Born G, Peto R. Prevention of serious vascular events by aspirin amongst patients with peripheral arterial disease: randomized, double-blind trial. Journal of internal medicine 2007;261:276-84.

124. Cole BF, Baron JA, Sandler RS et al. Folic acid for the prevention of colorectal adenomas: a randomized clinical trial. Jama 2007;297:2351-9.

125. Cook NR, Albert CM, Gaziano JM et al. A randomized factorial trial of vitamins C and E and beta carotene in the secondary prevention of cardiovascular events in women: results from the Women's Antioxidant Cardiovascular Study. Archives of internal medicine 2007;167:1610-8.

126. Durga J, van Boxtel MP, Schouten EG et al. Effect of 3-year folic acid supplementation on cognitive function in older adults in the FACIT trial: a randomised, double blind, controlled trial. Lancet (London, England) 2007;369:208-16.

152

127. Jamison RL, Hartigan P, Kaufman JS et al. Effect of homocysteine lowering on mortality and vascular disease in advanced chronic kidney disease and end-stage renal disease: a randomized controlled trial. Jama 2007;298:1163-70.

128. Lappe JM, Travers-Gustafson D, Davies KM, Recker RR, Heaney RP. Vitamin D and calcium supplementation reduces cancer risk: results of a randomized trial. The American journal of clinical nutrition 2007;85:1586-91.

129. Liu BA, McGeer A, McArthur MA et al. Effect of multivitamin and mineral supplementation on episodes of infection in nursing home residents: a randomized, placebo-controlled study. Journal of the American Geriatrics Society 2007;55:35-42.

130. Lyons RA, Johansen A, Brophy S et al. Preventing fractures among older people living in institutional care: a pragmatic randomised double blind placebo controlled trial of vitamin D supplementation. Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 2007;18:811-8.

131. Plummer M, Vivas J, Lopez G et al. Chemoprevention of precancerous gastric lesions with antioxidant vitamin supplementation: a randomized trial in a high-risk population. Journal of the National Cancer Institute 2007;99:137-46.

132. Smith H, Anderson F, Raphael H, Maslin P, Crozier S, Cooper C. Effect of annual intramuscular vitamin D on fracture risk in elderly men and women--a population-based, randomized, double-blind, placebo-controlled trial. Rheumatology (Oxford, England) 2007;46:1852-7.

133. Vianna AC, Mocelin AJ, Matsuo T et al. Uremic hyperhomocysteinemia: a randomized trial of folate treatment for the prevention of cardiovascular events. Hemodialysis international International Symposium on Home Hemodialysis 2007;11:210-6.

134. Albert CM, Cook NR, Gaziano JM et al. Effect of folic acid and B vitamins on risk of cardiovascular events and total mortality among women at high risk for cardiovascular disease: a randomized trial. Jama 2008;299:2027-36.

135. Bjorkman M, Sorva A, Risteli J, Tilvis R. Vitamin D supplementation has minor effects on parathyroid hormone and bone turnover markers in vitamin D-deficient bedridden older patients. Age and ageing 2008;37:25-31.

136. Bolland MJ, Barber PA, Doughty RN et al. Vascular events in healthy older women receiving calcium supplementation: randomised controlled trial. BMJ (Clinical research ed) 2008;336:262-6.

137. Maraini G, Williams SL, Sperduto RD et al. A randomized, double-masked, placebo-controlled clinical trial of multivitamin supplementation for age-related lens opacities. Clinical trial of nutritional supplements and age-related cataract report no. 3. Ophthalmology 2008;115:599-607.e1.

138. Ebbing M, Bleie O, Ueland PM et al. Mortality and cardiovascular events in patients treated with homocysteine-lowering B vitamins after coronary angiography: a randomized controlled trial. Jama 2008;300:795-804.

139. Guyton JR, Brown BG, Fazio S, Polis A, Tomassini JE, Tershakovec AM. Lipid-altering efficacy and safety of ezetimibe/simvastatin coadministered with extended-release niacin in patients with type IIa or type IIb hyperlipidemia. Journal of the American College of Cardiology 2008;51:1564-72.

140. Logan RF, Grainge MJ, Shepherd VC, Armitage NC, Muir KR. Aspirin and folic acid for the prevention of recurrent colorectal adenomas. Gastroenterology 2008;134:29-38.

141. Milman U, Blum S, Shapira C et al. Vitamin E supplementation reduces cardiovascular events in a subgroup of middle-aged individuals with both type 2 diabetes mellitus and the haptoglobin 2-2 genotype: a prospective double-blinded clinical trial. Arteriosclerosis, thrombosis, and vascular biology 2008;28:341-7.

153

142. Prince RL, Austin N, Devine A, Dick IM, Bruce D, Zhu K. Effects of ergocalciferol added to calcium on the risk of falls in elderly high-risk women. Archives of internal medicine 2008;168:103-8.

143. Reid IR, Ames R, Mason B et al. Randomized controlled trial of calcium supplementation in healthy, nonosteoporotic, older men. Archives of internal medicine 2008;168:2276-2282.

144. Sesso HD, Buring JE, Christen WG et al. Vitamins E and C in the prevention of cardiovascular disease in men: the Physicians' Health Study II randomized controlled trial. Jama 2008;300:2123-33.

145. Zhu K, Devine A, Dick IM, Wilson SG, Prince RL. Effects of calcium and vitamin D supplementation on hip bone mineral density and calcium-related analytes in elderly ambulatory Australian women: a five-year randomized controlled trial. The Journal of clinical endocrinology and metabolism 2008;93:743-9.

146. Anker SD, Comin Colet J, Filippatos G et al. Ferric carboxymaltose in patients with heart failure and iron deficiency. The New England journal of medicine 2009;361:2436-48.

147. Hodis HN, Mack WJ, Dustin L et al. High-dose B vitamin supplementation and progression of subclinical atherosclerosis: a randomized controlled trial. Stroke 2009;40:730-6.

148. Imasa MS, Gomez NT, Nevado JB, Jr. Folic acid-based intervention in non-ST elevation acute coronary syndromes. Asian cardiovascular & thoracic annals 2009;17:13-21.

149. LaCroix AZ, Kotchen J, Anderson G et al. Calcium plus vitamin D supplementation and mortality in postmenopausal women: the Women's Health Initiative calcium-vitamin D randomized controlled trial. The journals of gerontology Series A, Biological sciences and medical sciences 2009;64:559-67.

150. Lippman SM, Klein EA, Goodman PJ et al. Effect of selenium and vitamin E on risk of prostate cancer and other cancers: the Selenium and Vitamin E Cancer Prevention Trial (SELECT). Jama 2009;301:39-51.

151. Saposnik G, Ray JG, Sheridan P, McQueen M, Lonn E. Homocysteine-lowering therapy and stroke risk, severity, and disability: additional findings from the HOPE 2 trial. Stroke 2009;40:1365-72.

152. Sang ZC, Wang F, Zhou Q et al. Combined use of extended-release niacin and atorvastatin: safety and effects on lipid modification. Chinese medical journal 2009;122:1615-20.

153. Wu K, Platz EA, Willett WC et al. A randomized trial on folic acid supplementation and risk of recurrent colorectal adenoma. The American journal of clinical nutrition 2009;90:1623-31.

154. Armitage JM, Bowman L, Clarke RJ et al. Effects of homocysteine-lowering with folic acid plus vitamin B12 vs placebo on mortality and major morbidity in myocardial infarction survivors: a randomized trial. Jama 2010;303:2486-94.

155. Chailurkit LO, Saetung S, Thakkinstian A, Ongphiphadhanakul B, Rajatanavin R. Discrepant influence of vitamin D status on parathyroid hormone and bone mass after two years of calcium supplementation. Clinical endocrinology 2010;73:167-72.

156. de Zeeuw D, Agarwal R, Amdahl M et al. Selective vitamin D receptor activation with paricalcitol for reduction of albuminuria in patients with type 2 diabetes (VITAL study): a randomised controlled trial. Lancet (London, England) 2010;376:1543-51.

157. Galan P, Kesse-Guyot E, Czernichow S, Briancon S, Blacher J, Hercberg S. Effects of B vitamins and omega 3 fatty acids on cardiovascular diseases: a randomised placebo controlled trial. BMJ (Clinical research ed) 2010;341:c6273.

158. Heinz J, Kropf S, Domrose U et al. B vitamins and the risk of total mortality and cardiovascular disease in end-stage renal disease: results of a randomized controlled trial. Circulation 2010;121:1432-8.

154

159. Hercberg S, Kesse-Guyot E, Druesne-Pecollo N et al. Incidence of cancers, ischemic cardiovascular diseases and mortality during 5-year follow-up after stopping antioxidant vitamins and minerals supplements: a postintervention follow-up in the SU.VI.MAX Study. International journal of cancer 2010;127:1875-81.

160. House AA, Eliasziw M, Cattran DC et al. Effect of B-vitamin therapy on progression of diabetic nephropathy: a randomized controlled trial. Jama 2010;303:1603-9.

161. Janssen HC, Samson MM, Verhaar HJ. Muscle strength and mobility in vitamin D-insufficient female geriatric patients: a randomized controlled trial on vitamin D and calcium supplementation. Aging clinical and experimental research 2010;22:78-84.

162. Salovaara K, Tuppurainen M, Karkkainen M et al. Effect of vitamin D(3) and calcium on fracture risk in 65- to 71-year-old women: a population-based 3-year randomized, controlled trial--the OSTPRE-FPS. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2010;25:1487-95.

163. Sanders KM, Stuart AL, Williamson EJ et al. Annual high-dose oral vitamin D and falls and fractures in older women: a randomized controlled trial. Jama 2010;303:1815-22.

164. Sanyal AJ, Chalasani N, Kowdley KV et al. Pioglitazone, vitamin E, or placebo for nonalcoholic steatohepatitis. The New England journal of medicine 2010;362:1675-85.

165. B vitamins in patients with recent transient ischaemic attack or stroke in the VITAmins TO Prevent Stroke (VITATOPS) trial: a randomised, double-blind, parallel, placebo-controlled trial. The Lancet Neurology 2010;9:855-65.

166. Boden WE, Probstfield JL, Anderson T et al. Niacin in patients with low HDL cholesterol levels receiving intensive statin therapy. The New England journal of medicine 2011;365:2255-67.

167. Cherniack EP, Florez HJ, Hollis BW, Roos BA, Troen BR, Levis S. The response of elderly veterans to daily vitamin D3 supplementation of 2,000 IU: a pilot efficacy study. Journal of the American Geriatrics Society 2011;59:286-90.

168. Grimnes G, Figenschau Y, Almas B, Jorde R. Vitamin D, insulin secretion, sensitivity, and lipids: results from a case-control study and a randomized controlled trial using hyperglycemic clamp technique. Diabetes 2011;60:2748-57.

169. Lewis JR, Calver J, Zhu K, Flicker L, Prince RL. Calcium supplementation and the risks of atherosclerotic vascular disease in older women: results of a 5-year RCT and a 4.5-year follow-up. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2011;26:35-41.

170. Gallagher JC, Sai A, Templin T, 2nd, Smith L. Dose response to vitamin D supplementation in postmenopausal women: a randomized trial. Annals of internal medicine 2012;156:425-37.

171. Marshall JR, Tangen CM, Sakr WA et al. Phase III trial of selenium to prevent prostate cancer in men with high-grade prostatic intraepithelial neoplasia: SWOG S9917. Cancer prevention research (Philadelphia, Pa) 2011;4:1761-9.

172. Alvarez JA, Law J, Coakley KE et al. High-dose cholecalciferol reduces parathyroid hormone in patients with early chronic kidney disease: a pilot, randomized, double-blind, placebo-controlled trial. The American journal of clinical nutrition 2012;96:672-9.

173. Glendenning P, Zhu K, Inderjeeth C, Howat P, Lewis JR, Prince RL. Effects of three-monthly oral 150,000 IU cholecalciferol supplementation on falls, mobility, and muscle strength in older postmenopausal women: a randomized controlled trial. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2012;27:170-6.

174. Lehouck A, Mathieu C, Carremans C et al. High doses of vitamin D to reduce exacerbations in chronic obstructive pulmonary disease: a randomized trial. Annals of internal medicine 2012;156:105-14.

155

175. Ma JL, Zhang L, Brown LM et al. Fifteen-year effects of Helicobacter pylori, garlic, and vitamin treatments on gastric cancer incidence and mortality. Journal of the National Cancer Institute 2012;104:488-92.

176. Punthakee Z, Bosch J, Dagenais G et al. Design, history and results of the Thiazolidinedione Intervention with vitamin D Evaluation (TIDE) randomised controlled trial. Diabetologia 2012;55:36-45.

177. Sesso HD, Christen WG, Bubes V et al. Multivitamins in the prevention of cardiovascular disease in men: the Physicians' Health Study II randomized controlled trial. Jama 2012;308:1751-60.

178. Delanaye P, Weekers L, Warling X et al. Cholecalciferol in haemodialysis patients: a randomized, double-blind, proof-of-concept and safety study. Nephrology, dialysis, transplantation : official publication of the European Dialysis and Transplant Association - European Renal Association 2013;28:1779-86.

179. Hewitt NA, O'Connor AA, O'Shaughnessy DV, Elder GJ. Effects of cholecalciferol on functional, biochemical, vascular, and quality of life outcomes in hemodialysis patients. Clinical journal of the American Society of Nephrology : CJASN 2013;8:1143-9.

180. Lamas GA, Boineau R, Goertz C et al. Oral high-dose multivitamins and minerals after myocardial infarction: a randomized trial. Annals of internal medicine 2013;159:797-805.

181. Manning PJ, Sutherland WH, Williams SM et al. The effect of lipoic acid and vitamin E therapies in individuals with the metabolic syndrome. Nutrition, metabolism, and cardiovascular diseases : NMCD 2013;23:543-9.

182. Prentice RL, Pettinger MB, Jackson RD et al. Health risks and benefits from calcium and vitamin D supplementation: Women's Health Initiative clinical trial and cohort study. Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 2013;24:567-80.

183. Witham MD, Price RJ, Struthers AD et al. Cholecalciferol treatment to reduce blood pressure in older patients with isolated systolic hypertension: the VitDISH randomized controlled trial. JAMA internal medicine 2013;173:1672-9.

184. Witham MD, Dove FJ, Khan F, Lang CC, Belch JJ, Struthers AD. Effects of vitamin D supplementation on markers of vascular function after myocardial infarction--a randomised controlled trial. Int J Cardiol 2013;167:745-9.

185. Landray MJ, Haynes R, Hopewell JC et al. Effects of extended-release niacin with laropiprant in high-risk patients. The New England journal of medicine 2014;371:203-12.

186. Wang AY, Fang F, Chan J et al. Effect of paricalcitol on left ventricular mass and function in CKD--the OPERA trial. Journal of the American Society of Nephrology : JASN 2014;25:175-86.

187. van Wijngaarden JP, Swart KM, Enneman AW et al. Effect of daily vitamin B-12 and folic acid supplementation on fracture incidence in elderly individuals with an elevated plasma homocysteine concentration: B-PROOF, a randomized controlled trial. The American journal of clinical nutrition 2014;100:1578-86.

188. Wang L, Sesso HD, Glynn RJ et al. Vitamin E and C supplementation and risk of cancer in men: posttrial follow-up in the Physicians' Health Study II randomized trial. The American journal of clinical nutrition 2014;100:915-23.

189. Huo Y, Li J, Qin X et al. Efficacy of folic acid therapy in primary prevention of stroke among adults with hypertension in China: the CSPPT randomized clinical trial. Jama 2015;313:1325-35.

190. Ponikowski P, van Veldhuisen DJ, Comin-Colet J et al. Beneficial effects of long-term intravenous iron therapy with ferric carboxymaltose in patients with symptomatic heart failure and iron deficiencydagger. European heart journal 2015;36:657-68.

156

191. Wang LL, H. Zhou, Y. Jin, L. Liu, J. Low-dose B vitamins supplementation ameliorates cardiovascular risk: a double-blind randomized controlled trial in healthy Chinese elderly. Eur J Nutr 2015;54:455-64.

192. van Dijk SC, Enneman AW, Swart KM et al. Effects of 2-year vitamin B12 and folic acid supplementation in hyperhomocysteinemic elderly on arterial stiffness and cardiovascular outcomes within the B-PROOF trial. Journal of hypertension 2015;33:1897-906; discussion 1906.

193. Gupta A, Prabhakar S, Modi M et al. Effect of Vitamin D and calcium supplementation on ischaemic stroke outcome: a randomised controlled open-label trial. International journal of clinical practice 2016;70:764-70.

194. Lappe J, Watson P, Travers-Gustafson D et al. Effect of Vitamin D and Calcium Supplementation on Cancer Incidence in Older Women: A Randomized Clinical Trial. Jama 2017;317:1234-1243.

157

ars.els-cdn.com · Web viewFunnel plot of vitamin D supplementation and CVD and all-cause mortality risk.

Documents