Table ST1. Search strategy PubMed...Table ST1. Search strategy PubMed 1 (((Depress* OR Clinical depression OR Mood disorder OR Major depressive episode OR Dysthymi* OR Affective disorder

Table ST1. Search strategy

PubMed

1 (((Depress* OR Clinical depression OR Mood disorder OR Major depressive episode OR Dysthymi* OR Affective disorder OR

Anxiety OR Generalised Anxiety Disorder OR GAD OR Phobia OR Panic disorder OR Panic attack OR Agoraphobia OR Social

anxiety disorder OR Obsessive compulsive disorder OR Obsessive-compulsive disorder OR OCD OR PTSD OR Post traumatic

stress disorder OR Post-traumatic stress disorder OR Alprazolam OR Benzodiazepines OR Anthramycin OR Bromazepam OR

Clonazepam OR Devazepide OR Diazepam OR Flumazenil OR Flunitrazepam OR Flurazepam OR Lorazepam OR Nitrazepam OR

Oxazepam OR Pirenzepine OR Prazepam OR Temazepam OR Chlordiazepoxide OR Clorazepate OR Dipotassium OR Estazolam

OR Medazepam OR Midazolam OR Triazolam OR Psychotropic Drugs OR Antidepressive Agents OR Antidepressive Agents

(Second-Generation) OR Antidepressive Agents (Tricyclic) OR Hallucinogens OR Tranquilizing Agents OR Anti-Anxiety Agents

OR Antimanic Agents OR Antipsychotic Agents))) AND ((Cognit* OR Memory OR Attention OR Reaction time OR Speed of

processing OR processing speed OR Crystallized ability OR Crystallized intelligence OR Fluid ability OR Fluid intelligence OR

General mental ability OR GMA OR Intelligence OR Executive function OR Neuropsychological testing OR Mini mental stat*

exam* OR MMSE OR Dementia OR Alzheimer* OR Mild cognitive impairment OR MCI )) AND (Humans[Mesh])

Ovid PsycINFO

1 (Depress* or Clinical depression or Mood disorder or Major depressive episode or Dysthymi* or Affective disorder or Anxiety or

Generalised Anxiety Disorder or GAD or Phobia or Panic disorder or Panic attack or Agoraphobia or Social anxiety disorder or

Obsessive compulsive disorder or OCD or PTSD or Post traumatic stress disorder or Alprazolam or Benzodiazepines or

Anthramycin or Bromazepam or Clonazepam or Devazepide or Diazepam or Flumazenil or Flunitrazepam or Flurazepam or

Lorazepam or Nitrazepam or Oxazepam or Pirenzepine or Prazepam or Temazepam or Chlordiazepoxide or Clorazepate or

Dipotassium or Estazolam or Medazepam or Midazolam or Triazolam or Psychotropic Drugs or Antidepressive Agents or

Antidepressive Agents Second generation or Antidepressive Agents Tricyclic or Hallucinogens or Tranquilizing Agents or Anti

Anxiety Agents or Antimanic Agents or Antipsychotic Agents).mp. [mp=title, abstract, heading word, table of contents, key

concepts, original title, tests & measures]

2 (Cognit* or Memory or Attention or Reaction time or Speed of processing or processing speed or Crystallized ability or Crystallized

intelligence or Fluid ability or Fluid intelligence or General mental ability or GMA or Intelligence or Executive function or

Neuropsychological testing or Mini mental stat* exam* or MMSE or Dementia or Alzheimer* or Mild cognitive impairment or

MCI).mp. [mp=title, abstract, heading word, table of contents, key concepts, original title, tests & measures]

3 1 and 2

4 limit 3 to (human and English language)

Cochrane library

1 'Depress* OR Clinical depression OR Mood disorder OR Major depressive episode OR Dysthymi* OR Affective disorder OR

Anxiety OR Generalised Anxiety Disorder OR GAD OR Phobia OR Panic disorder OR Panic attack OR Agoraphobia OR Social

anxiety disorder OR Obsessive compulsive disorder OR Obsessive-compulsive disorder OR OCD OR PTSD OR Post traumatic

stress disorder OR Post-traumatic stress disorder OR Alprazolam OR Benzodiazepines OR Anthramycin OR Bromazepam OR

Clonazepam OR Devazepide OR Diazepam OR Flumazenil OR Flunitrazepam OR Flurazepam OR Lorazepam OR Nitrazepam OR

Oxazepam OR Pirenzepine OR Prazepam OR Temazepam OR Chlordiazepoxide OR Clorazepate OR Dipotassium OR Estazolam

OR Medazepam OR Midazolam OR Triazolam OR Psychotropic Drugs OR Antidepressive Agents OR Antidepressive Agents

(Second generation) OR Antidepressive Agents (Tricyclic) OR Hallucinogens OR Tranquilizing Agents OR Anti-Anxiety Agents

OR Antimanic Agents OR Antipsychotic Agents and Cognit* OR Memory OR Attention OR Reaction time OR Speed of processing

OR processing speed OR Crystallized ability OR Crystallized intelligence OR Fluid ability OR Fluid intelligence OR General mental

ability OR GMA OR Intelligence OR Executive function OR Neuropsychological testing OR Mini mental stat* exam* OR MMSE

OR Dementia OR Alzheimer* OR Mild cognitive impairment OR MCI

Table ST2. Quality rating of studies selected for inclusion based on the Newcastle-Ottawa Scale.

Quality Indicators From the Newcastle-Ottawa Scale

Studies 1 2 3 4 5 6 7 8 9

Barnes et al.,

2012

Yes yes yes yes no yes yes yes yes

Becker et al.,

2009

yes yes yes yes yes yes yes yes yes

Blasko et al.,

2010

yes yes yes yes no yes yes yes yes

Chen et al.,

1999

yes yes yes yes yes yes yes yes yes

Chen et al., 2008

(China)

no yes yes yes yes yes yes yes yes

Chen et al., 2008

(UK)

yes yes yes yes yes yes yes yes yes

Dal Forno et al.,

2005

yes yes yes yes no yes yes yes no

Devanand et al.,

1996

no yes yes yes yes yes yes yes yes

Dotson et al.,

2010

yes yes yes yes yes yes yes yes no

Fuhrer et al.,

2003

yes yes yes yes yes yes yes yes yes

Gatz et al.,

2005

yes yes yes yes yes yes yes yes yes

Geerlings et al.,

2008

yes yes yes yes yes yes yes yes no

Geerlings et al.,

2000

yes yes yes yes yes yes yes yes yes

Goveas et al.,

2011

no yes yes yes no yes yes yes yes

Heser et al.,

2013

yes yes yes yes yes yes yes yes yes

Heun et al., yes yes yes yes yes yes yes yes yes

2006

Irie et al.,

2008

no yes yes yes no yes yes yes yes

Jessen et al.,

2010

yes yes yes yes yes yes yes yes yes

Jessen et al.,

2011

yes yes yes yes yes yes yes yes yes

Jungwirth et al.,

2009

yes yes yes yes no no yes yes yes

Kim et al.,

2010

yes yes yes yes yes yes yes yes yes

Kim et al.,

2011

yes yes yes yes yes yes yes yes yes

Kohler et al.,

2011

yes yes yes yes yes yes yes yes yes

Lenoir et al.,

2011

yes yes yes yes yes yes yes yes yes

Li et al.,

2011

yes yes yes yes yes yes yes yes yes

Luchsinger et al.,

2008

yes yes yes yes yes yes yes yes no

Luppa et al.,

2013

yes yes yes yes yes yes yes yes yes

Palmer et al.,

2007

yes yes yes yes yes yes yes yes yes

Palsson et al.,

1999

yes yes yes yes no no yes yes yes

Saczynski et al.,

2010

yes yes yes yes yes yes yes yes no

Samieri et al.,

2008

yes yes yes yes no yes yes yes yes

Schmand et al.,

1997

yes yes yes yes no no yes yes yes

St John et al.,

2002

yes yes yes yes yes yes yes yes yes

Van der Kommer

et al., 2010

yes yes yes yes no no yes yes yes

Vilalta-Franch et

al., 2012

yes yes yes yes yes yes yes yes yes

Wilson et al.,

2003

yes yes yes yes yes yes yes yes yes

Wilson et al.,

2011

yes yes yes yes yes yes yes yes yes

1. Cohort truly representative of the population 2. Non-exposed participants from same community as exposed participants 3. Ascertainment of

exposure 4. Dementia not present at baseline 5. Study controls for age and sex 6. Study controls for other factors 7. Quality of outcome

assessment 8. Follow-up long enough for dementia to occur 9. Complete follow-up (all participants are accounted for)

Analyses:

AD continuous (all compatible)

Random-Effects Model (k = 10; tau^2 estimator: DL)

logLik deviance AIC BIC AICc

8.1673 22.1823 -12.3346 -11.7294 -10.6203

tau^2 (estimated amount of total heterogeneity): 0.0015 (SE = 0.0014)

tau (square root of estimated tau^2 value): 0.0391

I^2 (total heterogeneity / total variability): 62.06%

H^2 (total variability / sampling variability): 2.64

Test for Heterogeneity:

Q(df = 9) = 23.7240, p-val = 0.0048

Model Results:

estimate se zval pval ci.lb ci.ub

0.0536 0.0190 2.8217 0.0048 0.0164 0.0908 **

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Leave one out sensitivity analysis estimate se zval pval ci.lb ci.ub Q Qp tau2 I2 H2

1 0.0492 0.0189 2.5979 0.0094 0.0121 0.0863 21.8313 0.0052 0.0015 63.3553 2.7289

2 0.054 0.0194 2.7833 0.0054 0.016 0.0919 23.53 0.0027 0.0016 66.0008 2.9412

3 0.0586 0.024 2.437 0.0148 0.0115 0.1057 22.1508 0.0046 0.0022 63.8839 2.7688

4 0.0623 0.0165 3.7839 0.0002 0.0301 0.0946 13.2307 0.1042 0.0007 39.5346 1.6538

5 0.0619 0.0235 2.6365 0.0084 0.0159 0.108 23.7195 0.0026 0.0021 66.2725 2.9649

6 0.0601 0.0226 2.6605 0.0078 0.0158 0.1044 23.7212 0.0026 0.0019 66.2749 2.9651

7 0.0486 0.0184 2.6442 0.0082 0.0126 0.0847 21.0271 0.0071 0.0014 61.9539 2.6284

8 0.0438 0.0166 2.6455 0.0082 0.0114 0.0763 17.505 0.0253 0.001 54.2989 2.1881

9 0.0484 0.0182 2.6532 0.008 0.0126 0.0841 20.7307 0.0079 0.0013 61.4098 2.5913

10 0.061 0.0241 2.5344 0.0113 0.0138 0.1081 23.4956 0.0028 0.0022 65.951 2.9369

Fail-safe N Calculation Using the Rosenthal Approach

Fail-safe N Calculation Using the Rosenthal Approach

Observed Significance Level: <.0001

Target Significance Level: 0.05

Fail-safe N: 101

Trim-and-fill analysis

Estimated number of missing studies on the left side: 5 (SE = 1.9266)

Random-Effects Model (k = 15; tau^2 estimator: DL)

tau^2 (estimated amount of total heterogeneity): 0.0025 (SE = 0.0021)

tau (square root of estimated tau^2 value): 0.0501

I^2 (total heterogeneity / total variability): 64.10%

H^2 (total variability / sampling variability): 2.79

Test for Heterogeneity:

Q(df = 14) = 38.9936, p-val = 0.0004

Model Results:

estimate se zval pval ci.lb ci.ub

0.0334 0.0215 1.5493 0.1213 -0.0088 0.0756

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

AD continuous (CES-D/HAM)

Random-Effects Model (k = 7; tau^2 estimator: DL)

logLik deviance AIC BIC AICc

3.5299 15.2687 -3.0598 -3.4762 0.9402

tau^2 (estimated amount of total heterogeneity): 0.0025 (SE = 0.0029)

tau (square root of estimated tau^2 value): 0.0505

I^2 (total heterogeneity / total variability): 72.87%

H^2 (total variability / sampling variability): 3.69

Test for Heterogeneity:

Q(df = 5) = 18.4272, p-val = 0.0025

Model Results:

estimate se zval pval ci.lb ci.ub

0.0538 0.0297 1.8108 0.0702 -0.0044 0.1121 .

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Leave one out sensitivity analysis

estimate se zval pval ci.lb ci.ub Q Qp tau2 I2 H2

1 0.0661 0.0314 2.1066 0.0352 0.0046 0.1275 21.3265 0.0007 0.0031 76.555 4.2653

2 0.0928 0.0462 2.0118 0.0442 0.0024 0.1833 18.5071 0.0024 0.0056 72.9833 3.7014

3 0.0911 0.0325 2.8064 0.005 0.0275 0.1548 11.1795 0.0479 0.0022 55.2754 2.2359

4 0.0977 0.0453 2.1576 0.031 0.009 0.1865 21.5077 0.0006 0.0054 76.7525 4.3015

5 0.0544 0.03 1.8121 0.07 -

0.0044

0.1133 18.7243 0.0022 0.0026 73.2967 3.7449

6 0.044 0.0269 1.6383 0.1014 -

0.0086

0.0966 15.1112 0.0099 0.0019 66.9119 3.0222

7 0.0538 0.0297 1.8108 0.0702 -

0.0044

0.1121 18.4272 0.0025 0.0025 72.8661 3.6854

Fail-safe N Calculation Using the Rosenthal Approach

Observed Significance Level: <.0001

Target Significance Level: 0.05

Fail-safe N: 43

Trim-and-fill analysis

Estimated number of missing studies on the left side: 3 (SE = 1.6850)

Random-Effects Model (k = 10; tau^2 estimator: DL)

tau^2 (estimated amount of total heterogeneity): 0.0047 (SE = 0.0048)

tau (square root of estimated tau^2 value): 0.0686

I^2 (total heterogeneity / total variability): 74.45%

H^2 (total variability / sampling variability): 3.91

Test for Heterogeneity:

Q(df = 9) = 35.2272, p-val < .0001

Model Results:

estimate se zval pval ci.lb ci.ub

0.0276 0.0339 0.8125 0.4165 -0.0389 0.0940

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

AD Clinical Threshold (All compatible)

Random-Effects Model (k = 10; tau^2 estimator: DL)

logLik deviance AIC BIC AICc

-8.5049 16.0360 21.0097 21.6149 22.7240

tau^2 (estimated amount of total heterogeneity): 0.1929 (SE = 0.1725)

tau (square root of estimated tau^2 value): 0.4392

I^2 (total heterogeneity / total variability): 54.92%

H^2 (total variability / sampling variability): 2.22

Test for Heterogeneity:

Q(df = 9) = 19.9626, p-val = 0.0181

Model Results:

estimate se zval pval ci.lb ci.ub

0.7138 0.1924 3.7096 0.0002 0.3367 1.0910 ***

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Leave one out sensitivity analysis estimate se zval pval ci.lb ci.ub Q Qp tau2 I2 H2

1 0.0492 0.0189 2.5979 0.0094 0.0121 0.0863 21.8313 0.0052 0.0015 63.3553 2.7289

2 0.054 0.0194 2.7833 0.0054 0.016 0.0919 23.53 0.0027 0.0016 66.0008 2.9412

3 0.0586 0.024 2.437 0.0148 0.0115 0.1057 22.1508 0.0046 0.0022 63.8839 2.7688

4 0.0623 0.0165 3.7839 0.0002 0.0301 0.0946 13.2307 0.1042 0.0007 39.5346 1.6538

5 0.0619 0.0235 2.6365 0.0084 0.0159 0.108 23.7195 0.0026 0.0021 66.2725 2.9649

6 0.0601 0.0226 2.6605 0.0078 0.0158 0.1044 23.7212 0.0026 0.0019 66.2749 2.9651

7 0.0486 0.0184 2.6442 0.0082 0.0126 0.0847 21.0271 0.0071 0.0014 61.9539 2.6284

8 0.0438 0.0166 2.6455 0.0082 0.0114 0.0763 17.505 0.0253 0.001 54.2989 2.1881

9 0.0484 0.0182 2.6532 0.008 0.0126 0.0841 20.7307 0.0079 0.0013 61.4098 2.5913

10 0.061 0.0241 2.5344 0.0113 0.0138 0.1081 23.4956 0.0028 0.0022 65.951 2.9369

Fail-safe N Calculation Using the Rosenthal Approach

Observed Significance Level: <.0001

Target Significance Level: 0.05

Fail-safe N: 102

Trim-and-fill analysis

Estimated number of missing studies on the left side: 3 (SE = 2.1498)

Random-Effects Model (k = 13; tau^2 estimator: DL)

tau^2 (estimated amount of total heterogeneity): 0.2369 (SE = 0.1753)

tau (square root of estimated tau^2 value): 0.4867

I^2 (total heterogeneity / total variability): 58.01%

H^2 (total variability / sampling variability): 2.38

Test for Heterogeneity:

Q(df = 12) = 28.5795, p-val = 0.0045

Model Results:

estimate se zval pval ci.lb ci.ub

0.5181 0.1823 2.8415 0.0045 0.1607 0.8755 **

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’

AD Clinical Threshold (CES-D > 20)

Random-Effects Model (k = 5; tau^2 estimator: DL) logLik deviance AIC BIC AICc

-5.3513 9.5265 14.7026 13.9215 20.7026

tau^2 (estimated amount of total heterogeneity): 0.4698 (SE = 0.4821)

tau (square root of estimated tau^2 value): 0.6854

I^2 (total heterogeneity / total variability): 70.96%

H^2 (total variability / sampling variability): 3.44

Test for Heterogeneity:

Q(df = 4) = 13.7733, p-val = 0.0081

Model Results:

estimate se zval pval ci.lb ci.ub

0.6757 0.3682 1.8351 0.0665 -0.0460 1.3973 .

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Leave one out sensitivity analysis

estimate se zval pval ci.lb ci.ub Q Qp tau2 I2 H2

1 0.6953 0.4513 1.5407 0.1234 -0.1892 1.5798 13.7733 0.0032 0.6264 78.2188 4.5911

2 0.6327 0.4892 1.2933 0.1959 -0.3262 1.5915 13.1567 0.0043 0.726 77.1979 4.3856

3 0.4993 0.414 1.2061 0.2278 -0.3121 1.3107 9.6725 0.0216 0.4626 68.9843 3.2242

4 1.0257 0.225 4.5583 0 0.5847 1.4667 1.3841 0.7093 0 0 1

5 0.5639 0.4242 1.3295 0.1837 -0.2674 1.3952 12.527 0.0058 0.5404 76.0517 4.1757

Fail-safe N Calculation Using the Rosenthal Approach

Observed Significance Level: 0.0004

Target Significance Level: 0.05

Fail-safe N: 16

Trim-and-fill analysis

Estimated number of missing studies on the left side: 1 (SE = 1.7009)

Random-Effects Model (k = 6; tau^2 estimator: DL)

tau^2 (estimated amount of total heterogeneity): 0.4876 (SE = 0.4359)

tau (square root of estimated tau^2 value): 0.6983

I^2 (total heterogeneity / total variability): 72.65%

H^2 (total variability / sampling variability): 3.66

Test for Heterogeneity:

Q(df = 5) = 18.2798, p-val = 0.0026

Model Results:

estimate se zval pval ci.lb ci.ub

0.5005 0.3382 1.4801 0.1389 -0.1623 1.1633

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

AD Clinical Threshold (CES-D > 16)

Random-Effects Model (k = 6; tau^2 estimator: DL) logLik deviance AIC BIC AICc

-0.8119 2.5581 5.6239 5.2074 9.6239

tau^2 (estimated amount of total heterogeneity): 0 (SE = 0.0724)

tau (square root of estimated tau^2 value): 0

I^2 (total heterogeneity / total variability): 0.00%

H^2 (total variability / sampling variability): 1.00

Test for Heterogeneity:

Q(df = 5) = 2.5581, p-val = 0.7677

Model Results:

estimate se zval pval ci.lb ci.ub

0.4585 0.1189 3.8561 0.0001 0.2255 0.6916 ***

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Leave one out sensitivity analysis estimate se zval pval ci.lb ci.ub Q Qp tau2 I2 H2

1 0.5958 0.1827 3.2608 0.0011 0.2377 0.954 1.5787 0.8126 0 0 1

2 0.4684 0.1222 3.8315 0.0001 0.2288 0.708 2.4371 0.6559 0 0 1

3 0.4247 0.1225 3.4667 0.0005 0.1846 0.6648 1.2333 0.8726 0 0 1

4 0.4669 0.1221 3.8221 0.0001 0.2275 0.7063 2.469 0.6502 0 0 1

5 0.4335 0.1233 3.5142 0.0004 0.1917 0.6752 1.9731 0.7407 0 0 1

6 0.4337 0.132 3.2844 0.001 0.1749 0.6924 2.3703 0.668 0 0 1

Fail-safe N Calculation Using the Rosenthal Approach

Fail-safe N Calculation Using the Rosenthal Approach

Observed Significance Level: <.0001

Target Significance Level: 0.05

Fail-safe N: 26

Estimated number of missing studies on the left side: 0 (SE = 1.6829)

Estimated number of missing studies on the left side: 2 (SE = 1.7837)

Random-Effects Model (k = 8; tau^2 estimator: DL)

tau^2 (estimated amount of total heterogeneity): 0 (SE = 0.0676)

tau (square root of estimated tau^2 value): 0

I^2 (total heterogeneity / total variability): 0.00%

H^2 (total variability / sampling variability): 1.00

Test for Heterogeneity:

Q(df = 7) = 5.1651, p-val = 0.6398

Model Results:

estimate se zval pval ci.lb ci.ub

0.3952 0.1119 3.5302 0.0004 0.1758 0.6146 ***

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Dementia continuous (all compatible)

Random-Effects Model (k = 10; tau^2 estimator: DL) logLik deviance AIC BIC AICc

7.1882 27.5279 -10.3764 -9.7712 -8.6621

tau^2 (estimated amount of total heterogeneity): 0.0008 (SE = 0.0007)

tau (square root of estimated tau^2 value): 0.0284

I^2 (total heterogeneity / total variability): 63.06%

H^2 (total variability / sampling variability): 2.71

Test for Heterogeneity:

Q(df = 9) = 24.3653, p-val = 0.0038

Model Results:

estimate se zval pval ci.lb ci.ub

0.0508 0.0135 3.7672 0.0002 0.0244 0.0772 ***

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Leave one out sensitivity analysis estimate se zval pval ci.lb ci.ub Q Qp tau2 I2 H2

1 0.0492 0.0189 2.5979 0.0094 0.0121 0.0863 21.8313 0.0052 0.0015 63.3553 2.7289

2 0.054 0.0194 2.7833 0.0054 0.016 0.0919 23.53 0.0027 0.0016 66.0008 2.9412

3 0.0586 0.024 2.437 0.0148 0.0115 0.1057 22.1508 0.0046 0.0022 63.8839 2.7688

4 0.0623 0.0165 3.7839 0.0002 0.0301 0.0946 13.2307 0.1042 0.0007 39.5346 1.6538

5 0.0619 0.0235 2.6365 0.0084 0.0159 0.108 23.7195 0.0026 0.0021 66.2725 2.9649

6 0.0601 0.0226 2.6605 0.0078 0.0158 0.1044 23.7212 0.0026 0.0019 66.2749 2.9651

7 0.0486 0.0184 2.6442 0.0082 0.0126 0.0847 21.0271 0.0071 0.0014 61.9539 2.6284

8 0.0438 0.0166 2.6455 0.0082 0.0114 0.0763 17.505 0.0253 0.001 54.2989 2.1881

9 0.0484 0.0182 2.6532 0.008 0.0126 0.0841 20.7307 0.0079 0.0013 61.4098 2.5913

10 0.061 0.0241 2.5344 0.0113 0.0138 0.1081 23.4956 0.0028 0.0022 65.951 2.9369

Fail-safe N Calculation Using the Rosenthal Approach

Observed Significance Level: <.0001

Target Significance Level: 0.05

Fail-safe N: 121

Trim-and-fill analysis Estimated number of missing studies on the left side: 3 (SE = 2.3645)

Random-Effects Model (k = 15; tau^2 estimator: DL)

tau^2 (estimated amount of total heterogeneity): 0.0017 (SE = 0.0015)

tau (square root of estimated tau^2 value): 0.0409

I^2 (total heterogeneity / total variability): 59.23%

H^2 (total variability / sampling variability): 2.45

Test for Heterogeneity:

Q(df = 14) = 34.3394, p-val = 0.0018

Model Results:

estimate se zval pval ci.lb ci.ub

0.0606 0.0188 3.2296 0.0012 0.0238 0.0974 **

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Dementia continuous (CES-D/HAM)

Random-Effects Model (k = 8; tau^2 estimator: DL) logLik deviance AIC BIC AICc

11.3939 17.6476 -18.7878 -18.6290 -16.3878

tau^2 (estimated amount of total heterogeneity): 0.0004 (SE = 0.0005)

tau (square root of estimated tau^2 value): 0.0211

I^2 (total heterogeneity / total variability): 54.82%

H^2 (total variability / sampling variability): 2.21

Test for Heterogeneity:

Q(df = 7) = 15.4937, p-val = 0.0302

Model Results:

estimate se zval pval ci.lb ci.ub

0.0457 0.0111 4.1261 <.0001 0.0240 0.0674 ***

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Leave one out sensitivity analysis estimate se zval pval ci.lb ci.ub Q Qp tau2 I2 H2

1 0.0492 0.0189 2.5979 0.0094 0.0121 0.0863 21.8313 0.0052 0.0015 63.3553 2.7289

2 0.054 0.0194 2.7833 0.0054 0.016 0.0919 23.53 0.0027 0.0016 66.0008 2.9412

3 0.0586 0.024 2.437 0.0148 0.0115 0.1057 22.1508 0.0046 0.0022 63.8839 2.7688

4 0.0623 0.0165 3.7839 0.0002 0.0301 0.0946 13.2307 0.1042 0.0007 39.5346 1.6538

5 0.0619 0.0235 2.6365 0.0084 0.0159 0.108 23.7195 0.0026 0.0021 66.2725 2.9649

6 0.0601 0.0226 2.6605 0.0078 0.0158 0.1044 23.7212 0.0026 0.0019 66.2749 2.9651

7 0.0486 0.0184 2.6442 0.0082 0.0126 0.0847 21.0271 0.0071 0.0014 61.9539 2.6284

8 0.0438 0.0166 2.6455 0.0082 0.0114 0.0763 17.505 0.0253 0.001 54.2989 2.1881

9 0.0484 0.0182 2.6532 0.008 0.0126 0.0841 20.7307 0.0079 0.0013 61.4098 2.5913

10 0.061 0.0241 2.5344 0.0113 0.0138 0.1081 23.4956 0.0028 0.0022 65.951 2.9369

Fail-safe N Calculation Using the Rosenthal Approach

Observed Significance Level: <.0001

Target Significance Level: 0.05

Fail-safe N: 131

Trim-and-fill analysis Estimated number of missing studies on the left side: 3 (SE = 1.8667)

Random-Effects Model (k = 11; tau^2 estimator: DL)

tau^2 (estimated amount of total heterogeneity): 0.0009 (SE = 0.0007)

tau (square root of estimated tau^2 value): 0.0294

I^2 (total heterogeneity / total variability): 65.12%

H^2 (total variability / sampling variability): 2.87

Test for Heterogeneity:

Q(df = 10) = 28.6690, p-val = 0.0014

Model Results:

estimate se zval pval ci.lb ci.ub

0.0377 0.0129 2.9318 0.0034 0.0125 0.0629 **

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Dementia categorical (All compatible)

Random-Effects Model (k = 9; tau^2 estimator: DL) logLik deviance AIC BIC AICc

-6.6395 14.0693 17.2790 18.0748 18.7790

tau^2 (estimated amount of total heterogeneity): 0.0844 (SE = 0.0998)

tau (square root of estimated tau^2 value): 0.2905

I^2 (total heterogeneity / total variability): 38.34%

H^2 (total variability / sampling variability): 1.62

Test for Heterogeneity:

Q(df = 10) = 16.2182, p-val = 0.0936

Model Results:

estimate se zval pval ci.lb ci.ub

0.6855 0.1441 4.7573 <.0001 0.4031 0.9679 ***

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Leave one out sensitivity analysis estimate se zval pval ci.lb ci.ub Q Qp tau2 I2 H2

1 0.6725 0.157 4.2824 0 0.3647 0.9803 15.9913 0.0671 0.1034 43.7193 1.7768

2 0.704 0.155 4.5409 0 0.4001 1.0078 16.0193 0.0665 0.1014 43.8178 1.7799

3 0.6222 0.1476 4.2152 0 0.3329 0.9116 13.46 0.1429 0.0698 33.135 1.4956

4 0.8265 0.118 7.0026 0 0.5952 1.0579 2.8003 0.9717 0 0 1

5 0.666 0.1524 4.3692 0 0.3672 0.9648 15.7327 0.0727 0.0962 42.7944 1.7481

6 0.6865 0.1643 4.1774 0 0.3644 1.0087 16.1967 0.0629 0.1155 44.4332 1.7996

7 0.6745 0.165 4.0876 0 0.3511 0.9979 15.975 0.0674 0.115 43.662 1.775

8 0.6842 0.1607 4.2566 0 0.3691 0.9992 16.1867 0.0631 0.1098 44.3989 1.7985

9 0.6659 0.1531 4.3498 0 0.3658 0.9659 15.7466 0.0724 0.0969 42.8446 1.7496

10 0.7004 0.1565 4.4761 0 0.3937 1.0071 16.1187 0.0644 0.1039 44.1642 1.791

11 0.6627 0.1551 4.2715 0 0.3586 0.9668 15.6741 0.074 0.0984 42.5805 1.7416

Fail-safe N Calculation Using the Rosenthal Approach Observed Significance Level: <.0001

Target Significance Level: 0.05

Fail-safe N: 142

Trim-and-fill analysis Estimated number of missing studies on the left side: 4 (SE = 2.1937)

Random-Effects Model (k = 15; tau^2 estimator: DL)

tau^2 (estimated amount of total heterogeneity): 0.1145 (SE = 0.0999)

tau (square root of estimated tau^2 value): 0.3384

I^2 (total heterogeneity / total variability): 44.24%

H^2 (total variability / sampling variability): 1.79

Test for Heterogeneity:

Q(df = 14) = 25.1089, p-val = 0.0335

Model Results:

estimate se zval pval ci.lb ci.ub

0.5206 0.1344 3.8736 0.0001 0.2572 0.7840 ***

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Dementia categorical (CES-D >20)

Random-Effects Model (k = 5; tau^2 estimator: DL) logLik deviance AIC BIC AICc

-4.6477 9.2448 13.2954 12.5143 19.2954

tau^2 (estimated amount of total heterogeneity): 0.3891 (SE = 0.3983)

tau (square root of estimated tau^2 value): 0.6238

I^2 (total heterogeneity / total variability): 71.85%

H^2 (total variability / sampling variability): 3.55

Test for Heterogeneity:

Q(df = 4) = 14.2079, p-val = 0.0067

Model Results:

estimate se zval pval ci.lb ci.ub

0.6033 0.3337 1.8080 0.0706 -0.0507 1.2574 .

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Leave one out sensitivity analysis estimate se zval pval ci.lb ci.ub Q Qp tau2 I2 H2

1 0.5467 0.4162 1.3135 0.189 -

0.2691

1.3625 13.4257 0.0038 0.5237 77.6548 4.4752

2 0.6427 0.4113 1.5627 0.1181 -

0.1634

1.4487 14.1859 0.0027 0.5217 78.8523 4.7286

3 0.4305 0.3577 1.2035 0.2288 -

0.2706

1.1316 8.5327 0.0362 0.3257 64.8412 2.8442

4 0.9413 0.2066 4.5557 0 0.5363 1.3462 1.5507 0.6706 0 0 1

5 0.5201 0.3882 1.3399 0.1803 -

0.2407

1.2808 13.2095 0.0042 0.4585 77.289 4.4032

Fail-safe N Calculation Using the Rosenthal Approach Observed Significance Level: 0.0005

Target Significance Level: 0.05

Fail-safe N: 16

Trim-and-fill analysis Estimated number of missing studies on the left side: 1 (SE = 1.7009)

Random-Effects Model (k = 6; tau^2 estimator: DL)

tau^2 (estimated amount of total heterogeneity): 0.4046 (SE = 0.3582)

tau (square root of estimated tau^2 value): 0.6361

I^2 (total heterogeneity / total variability): 74.33%

H^2 (total variability / sampling variability): 3.90

Test for Heterogeneity:

Q(df = 5) = 19.4798, p-val = 0.0016

Model Results:

estimate se zval pval ci.lb ci.ub

0.4367 0.3056 1.4292 0.1529 -0.1622 1.0356

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Dementia categorical (CES-D >16)

Random-Effects Model (k = 9; tau^2 estimator: DL)

logLik deviance AIC BIC AICc

1.7584 2.7745 0.4832 0.8776 2.4832

tau^2 (estimated amount of total heterogeneity): 0 (SE = 0.0322)

tau (square root of estimated tau^2 value): 0

I^2 (total heterogeneity / total variability): 0.00%

H^2 (total variability / sampling variability): 1.00

Test for Heterogeneity:

Q(df = 8) = 2.7745, p-val = 0.9477

Model Results:

estimate se zval pval ci.lb ci.ub

0.5268 0.0771 6.8328 <.0001 0.3757 0.6779 ***

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Leave one out sensitivity analysis estimate se zval pval ci.lb ci.ub Q Qp tau2 I2 H2

1 0.5131 0.0823 6.2378 0 0.3519 0.6743 2.5459 0.9236 0 0 1

2 0.5157 0.0784 6.5817 0 0.3622 0.6693 2.1484 0.9512 0 0 1

3 0.5335 0.0782 6.8198 0 0.3801 0.6868 2.5213 0.9255 0 0 1

4 0.5131 0.08 6.4179 0 0.3564 0.6698 2.3577 0.9374 0 0 1

5 0.5299 0.0792 6.6914 0 0.3747 0.6852 2.7445 0.9076 0 0 1

6 0.5504 0.0805 6.8402 0 0.3927 0.7081 1.725 0.9735 0 0 1

7 0.5183 0.1058 4.8982 0 0.3109 0.7256 2.7606 0.9062 0 0 1

8 0.5253 0.0808 6.4972 0 0.3668 0.6837 2.7705 0.9054 0 0 1

9 0.5377 0.0793 6.7826 0 0.3824 0.6931 2.4244 0.9327 0 0 1

Fail-safe N Calculation Using the Rosenthal Approach Observed Significance Level: <.0001

Target Significance Level: 0.05

Fail-safe N: 114

Trim-and-fill analysis Estimated number of missing studies on the right side: 0 (SE = 2.0456)

Random-Effects Model (k = 9; tau^2 estimator: DL)

tau^2 (estimated amount of total heterogeneity): 0 (SE = 0.0322)

tau (square root of estimated tau^2 value): 0

I^2 (total heterogeneity / total variability): 0.00%

H^2 (total variability / sampling variability): 1.00

Test for Heterogeneity:

Q(df = 8) = 2.7745, p-val = 0.9477

Model Results:

estimate se zval pval ci.lb ci.ub

0.5268 0.0771 6.8328 <.0001 0.3757 0.6779 ***

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

VaD continuous (All)

Random-Effects Model (k = 3; tau^2 estimator: DL) logLik deviance AIC BIC AICc

1.3925 3.1371 1.2149 -0.5879 13.2149

tau^2 (estimated amount of total heterogeneity): 0.0027 (SE = 0.0124)

tau (square root of estimated tau^2 value): 0.0521

I^2 (total heterogeneity / total variability): 22.36%

H^2 (total variability / sampling variability): 1.29

Test for Heterogeneity:

Q(df = 2) = 2.5760, p-val = 0.2758

Model Results:

estimate se zval pval ci.lb ci.ub

0.0631 0.0544 1.1599 0.2461 -0.0435 0.1696

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Leave one out sensitivity analysis estimate se zval pval ci.lb ci.ub Q Qp tau2 I2 H2

1 0.5017 0.4457 1.1258 0.2602 -

0.3717

1.3752 12.9869 0.0015 0.5039 84.5999 6.4934

2 0.4038 0.4305 0.938 0.3482 -0.44 1.2476 8.5973 0.0136 0.415 76.737 4.2987

3 0.9324 0.1994 4.6768 0 0.5417 1.3232 1.1598 0.56 0 0 1

4 0.5834 0.5536 1.0538 0.292 -

0.5017

1.6685 13.4257 0.0012 0.7673 85.1032 6.7129

Fail-safe N Calculation Using the Rosenthal Approach Fail-safe N Calculation Using the Rosenthal Approach

Observed Significance Level: 0.0003

Target Significance Level: 0.05

Fail-safe N: 14

Trim-and-fill analysis Estimated number of missing studies on the left side: 1 (SE = 1.6103)

Random-Effects Model (k = 5; tau^2 estimator: DL)

tau^2 (estimated amount of total heterogeneity): 0.4494 (SE = 0.4100)

tau (square root of estimated tau^2 value): 0.6704

I^2 (total heterogeneity / total variability): 80.15%

H^2 (total variability / sampling variability): 5.04

Test for Heterogeneity:

Q(df = 4) = 20.1561, p-val = 0.0005

Model Results:

estimate se zval pval ci.lb ci.ub

0.4056 0.3388 1.1969 0.2313 -0.2585 1.0697

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

VaD categorical (All)

Random-Effects Model (k = 2; tau^2 estimator: DL) logLik deviance AIC BIC AICc

-1.6261 3.7765 7.2522 4.6385 19.2522

tau^2 (estimated amount of total heterogeneity): 0.3646 (SE = 0.6958)

tau (square root of estimated tau^2 value): 0.6038

I^2 (total heterogeneity / total variability): 74.10%

H^2 (total variability / sampling variability): 3.86

Test for Heterogeneity:

Q(df = 1) = 3.8603, p-val = 0.0494

Model Results:

estimate se zval pval ci.lb ci.ub

1.0370 0.4947 2.0961 0.0361 0.0673 2.0066 *

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Fail-safe N Calculation Using the Rosenthal Approach Fail-safe N Calculation Using the Rosenthal Approach

Observed Significance Level: <.0001

Target Significance Level: 0.05

Fail-safe N: 11