Vitamin D inadequacy is associated with low-energy distal radius fractures: A case–control study

Page 1

Bone 48 (2011) 1140–1145

Contents lists available at ScienceDirect

Bone

j ourna l homepage: www.e lsev ie r.com/ locate /bone

Vitamin D inadequacy is associated with low-energy distal radius fractures: Acase–control study

Jannike Øyen a,⁎, Ellen Margrete Apalset b,c, Clara Gram Gjesdal b,d, Christina Brudvik a,e,Stein Atle Lie a,f, Leiv M. Hove a,g

a Department of Surgical Sciences, University of Bergen, Bergen, Norwayb Department of Rheumatology, Haukeland University Hospital, Bergen, Norwayc Department of Public Health and Primary Health Care, University of Bergen, Bergen, Norwayd Section for Rheumatology, Institute of Medicine, University of Bergen, Bergen, Norwaye Bergen Accident and Emergency Department, Bergen, Norwayf Uni Health, Uni Research, Bergen, Norwayg Department of Orthopaedic Surgery, Haukeland University Hospital, Bergen, Norway

⁎ Corresponding author at: Department of Surgical ScDentistry, University of Bergen, N-5021 Bergen, Norway

E-mail address: [email protected] (J. Øyen).

8756-3282/$ – see front matter © 2011 Elsevier Inc. Aldoi:10.1016/j.bone.2011.01.021

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 5 October 2010Revised 24 January 2011Accepted 25 January 2011Available online 2 February 2011

Edited by: Rene Rizzoli

Keywords:Bone mineral densityCase–controlDistal radius fractureVitamin DOsteoporosis

Introduction: Vitamin D inadequacy is associated with hip fractures, but the relationship has not beenexplored for distal radius fractures.Aims: To compare serum 25-hydroxyvitamin D (s-25(OH)D) status in low-energy distal radius fracturepatients and a group of matched controls, and examine whether observed differences in s-25(OH)D betweenpatients and controls would remain after adjusting for bone mineral density (BMD), body mass index (BMI),and smoking history.Methods: A total of 575 female and 72 male low-energy distal radius fracture patients (50–90 years) and 534female and 52male matched controls were included. The primarymeasure was levels of vitamin D. Secondarymeasures were BMD assessed by dual energy X-ray absorptiometry, BMI and smoking history.Results: Mean s-25(OH)D was 66.5 nmol/L in female patients and 78.7 nmol/L in controls (pb0.001). Thecorresponding figures in men were 64.5 and 77.0 nmol/L (p=0.017). In adjusted conditional logisticregression analyzes, s-25(OH)D b50 nmol/L (OR=2.32, 95% CI: 1.47–3.64, pb0.001), and 50–75 (OR=1.70,95% CI: 1.17–2.47, p=0.005) were associated with distal radius fractures in women. s-25(OH)D b50 nmol/L

(OR=6.27, 95% CI: 1.17–33.66, p=0.032) was associated with distal radius fractures in men.Conclusions: Vitamin D inadequacy is associated with low-energy distal radius fractures in both women andmen. Differences in vitamin D levels are independent of BMD, BMI or smoking history.

© 2011 Elsevier Inc. All rights reserved.

Introduction

Vitamin D and parathyroid hormone (PTH) regulate calcium (Ca)metabolism and influence bone health. Vitamin D deficiency leads torickets in children and osteomalacia in adults [1]. In less severe vitaminD deficiency, gastrointestinal Ca absorption is decreased and theproduction of PTH increases, stimulating Ca release from the skeleton.Thus, people suffering from vitamin D deficiency are susceptible toosteoporosis and fractures [2,3]. Vitamin D deficiency has also beenassociated with muscle weakness, leading to an increased risk of falling[4]. Vitamin D deficiency is mainly caused by insufficient sunlightexposure, diminished capacity of the skin to synthesize vitamin D, orlow dietary intake of vitamin D [2]. The circulating level of serum 25-hydroxyvitamin D (s-25(OH)D) is the most reliable marker for vitamin

iences, Faculty of Medicine and. Fax: +47 55974934.

l rights reserved.

D status [1]. Severe andmoderate vitamin D deficiency are defined as s-25(OH)D lower than 12.5 nmol/L and 12.5–25 nmol/L, respectively [2].The optimal level of vitamin D regarding bone health has not beenestablished, but different thresholds have been proposed; 25(OH)DN30, N50, and N75 nmol/L [2,5–7]. It is difficult to define specificthresholds of circulating 25(OH)D for optimal bone health due to theimprecision of different 25(OH)D assays [8]. Currently vitamin Ddeficiency is defined as a 25(OH)D level b50 nmol/L based on studieson the association between 25(OH)D, bone mineral density (BMD),bone turnover, muscular function and falls [9], and the recentlypublished report from the Institute ofMedicine on dietary requirementsfor Ca and vitamin D [7], which indicate that 50 nmol/L is a suitablethreshold to classify vitaminD inadequacy. In addition, s-25(OH)Dup to75 nmol/L is proposed to be beneficial for several health outcomes;fracture and fall prevention, higher BMD, better extremity strength,dental health, cancer prevention and hypertension prevention [10]. IntheNordic countrieswhere the sunexposure is low, especially inwinter,an increase in mean s-25(OH)D to 105 nmol/L could probably decrease

Page 2

1141J. Øyen et al. / Bone 48 (2011) 1140–1145

the mortality rate [11]. About one in three people aged 65 years or olderexperiences at least one fall yearly [12], and about 6% of all falls result in afracture [13]. Vitamin D has a direct effect onmuscle strength modulatedby specific vitamin D receptors present in human muscle tissue [14].Randomized controlled trials show that high doses of vitamin D (700–1000 IU a day) reduce both the risk of falling [15] and the risk of hip andother non-vertebral fractures [16]. In a recently published study high oraldose of vitamin D supplement (500 000 IU once yearly) resulted in anincreased risk of falls and fractures [17]. However, the evidence for anassociation of s-25(OH)D levels with fractures is not consistent [18].

The incidence of distal radius fractures in Norway is high [19–21],and many of these patients have osteoporosis [22,23]. However, distalradius fracture patients are often not evaluated and treated forosteoporosis [24–27]. Vitamin D deficiency is found to be associatedwith hip fractures [3,28]. A large proportion of distal radius fracturepatients have also demonstrated vitamin D inadequacy [29,30], but sofar, vitamin D status among patients with distal radius fractures has notbeen compared to subjects without such fractures. The main aim of thisstudy was to determine s-25(OH)D status in female and male low-energy distal radius fracture patients compared to matched controls,and to ascertain whether observed differences in s-25(OH)D betweenpatients and controls where associated with BMD, body mass index(BMI) or smoking history.

Materials and methods

Population

All patients aged 50 years or older with a low-energy distal radiusfracturewere invited to participate in this case–control study. Theywererecruited from Bergen Accident and Emergency Department (AED) andHaukeland University Hospital (HUH), located in the city of Bergen,Western Norway. Control subjects were randomly selected from thegeneral population in the same area. The fracture patients and controlswere offered referral to the osteoporosis clinic at HUH for BMDmeasurements, blood tests and clinical assessment of osteoporosis.During the inclusion period (from October 2003 until October 2007)1252 female and 185 male patients 50 years or older with low-energydistal radius fracturewere treated at Bergen AED andHUH. Of these, 121declined participation because of earlier evaluation for osteoporosis, and291declined for unknown reasons. A further 268were excludedbecausetheywere over 90years old, tourists, or incapacitateddue to confusionorserious illness. The group of patients completing BMD measurementsand registration of risk factors for osteoporosis comprised 757. Eightpatients who had their examination at the osteoporosis clinicmore thansixmonths after the distal radius fracture were excluded. Blood sampleswere missing in 102 patients. The final study sample comprised 575female and 72 male patients. The average time between fracture andexamination at the osteoporosis clinic was 61 days (range 7–169) infemale and 57 days (range 18–152) in male patients.

Controls were randomly selected by Statistics Norway using theNorwegianPopulationRegistry. Theywerematchedonplace of residence,age (±2 years) and gender, and controlswere examined the samemonthas theirmatched cases. A total of 1352 female and 172male controlswereinvited bymail. Of these, 667 completed BMDmeasurements and clinicalassessment of osteoporosis. Fifty-five were excluded because of previouslow-energy distal radius fracture after the age of 50. In four femalecontrols neither hip nor spine scans could be used because of bilateral hipfracture or surgery and degenerative changes of the spine. Furthermore,22 blood samplesweremissing. A total of 534 female and 52male controlsubjects were included in the study from April 2008 until June 2009.

Demographic and clinical data

A low-energy fracture was defined as a fracture following a minortrauma, equivalent to falling from standing height or less [31]. All

distal radius fractures were radiologically confirmed. Weight andheight were measured in light clothing at the osteoporosis clinic, andbody mass index (BMI in kg/m2) was calculated. A self-administeredquestionnaire provided information on chronic illness (data notshown), use of glucocorticoids, bisphosphonates, vitamin D and Casupplements, history of hip fracture in a parent, and previousfractures. Previous fracture was defined as a fracture of upper arm,rib, spine, hip, distal femur or lower leg from low-energy trauma afterthe age of 50. Smoking was recorded as previous, current or never.Age at menopause, use of selective estrogen receptor modulators(SERM) and postmenopausal estrogen therapy were recorded forwomen. Early menopause was defined as loss of regular menstruationbefore the age of 45 (Table 1).

Measurements of bone mineral density

BMD was measured by dual energy X-ray absorptiometry (DXA)(Lunar Prodigy, GE Madison, WI, USA) at the femoral neck, total hipand lumbar spine (L2–L4). The results were based on measurementsof femoral neck BMD as this is recommended as a reference standard[32]. The densitometer was calibrated every day, and was stableduring the whole measurement period. The in vitro long-termprecision expressed as the coefficient of variation was 0.86%, and thein vivo short-term precision for femoral neck was 1.47%. BMD valuesfor the left hip were used unless there was a history of previousfracture or surgery. Scans from the right hip were used in 31 patientsand in nine controls. BMD scans of the hip were missing in ninefemale patients and eight female controls because of previousbilateral hip fracture or prosthesis. BMD was categorized accordingto different levels of T-score. The T-score calculations were derivedfrom a combined European/US reference population supplied by theDXA manufacturer Lunar [33,34]. T-scores in men were determinedfrom the database of young healthy Caucasian men. Osteoporosiswas defined as T-score ≤−2.5 SD, and osteopenia as T-score b−1.0SD and N −2.5 SD. Normal BMD was defined as T-score ≥ −1.0 SD[35].

Biochemical analyzes

Serum samples were kept frozen at −70 °C and all samples wereanalyzed at the Hormone Laboratory, HUH and the Laboratory forClinical Biochemistry, HUH. s-25(OH)D was measured using aradioimmunoassay from Immunodiagnostic System, Boldon, UK. Theinterassay coefficients of variation were 8.2%, 8.1% and 7.3% forconcentrations of 19.6 nmol/L, 56.7 nmol/L and 136 nmol/L, respec-tively. The LC-MS/MS-method has been validated against an LC-MS/MS-method developed at the Hormone Laboratory, HaukelandUniversity Hospital, Bergen, Norway. The LC-MS/MS-method showson the average 94% accuracy measured against an SI certifiedreference standard from Chromsystems, at two levels (88 nM and182 nM). The correlation coefficient between the RIA-IDS-assay usedin this study and the LC-MS/MS-method was 0.94. Below 50 nmol/Lthe RIA method measured on the average 5% higher values than theLC-MS/MS method, above 50 nmol/L the difference was less than 3%.Levels of Ca in serum (s-Ca) and albumin in serum (s-albumin) weremeasured by colorimetric assays using a Roche/Hitachi Modularanalyzer (Roche Diagnostics GmbH, Germany). The interassay varia-tions were 2.0% for s-Ca at a concentration of 2.44 mmol/L and 2.0% fors-albumin at a concentration of 44.0 g/L . s-Ca mmol/L was correctedfor s-albumin (s-Ca mmol/L+(40−s-albumin)×0.02).

Statistical analyzes

Categorical variables were expressed as numbers and percentages,and continuous variables as means with variance expressed as SDand/or range. We used independent sample t-tests for continuous

Page 3

Table 1Demographic variables, clinical characteristics and BMD measurements in distal radius fracture patients and controls. Mean (SD) for continuous variables and numbers (%) forcategorical variables.

Women Men

Patients(n=575)

Controls(n=534)

p-value Patients(n=72)

Controls(n=52)

p-value

Age (years) 66.2 (10.0) 65.1 (7.5) 0.028 65.1 (9.9) 67.1 (8.6) 0.247BMI (kg/m2) 25.4 (4.2) 26.4 (4.8) b0.001 25.5 (3.9) 27.1 (3.8) 0.019Smoking 0.728a 0.016a

Never 325 (56.5) 301 (56.4) 31 (43.1) 17 (32.7)Previous 159 (27.7) 140 (26.2) 21 (29.2) 28 (53.8)Current 91 (15.8) 93 (17.4) 20 (27.8) 7 (13.5)s-25(OH)D(nmol/L)

66.5 (29.2) 78.8 (33.5) b0.001 65.4 (27.5) 77.0 (28.5) 0.024

s-25(OH)D(nmol/L)

b0.001a 0.017a

b25 7 (1.2) 5 (0.9) 1 (1.4) 1 (1.9)≥25, b50 162 (28.2) 81 (15.2) 22 (31.0) 7 (13.5)≥50, b75 214 (37.3) 186 (35.0) 28 (39.4) 16 (30.8)≥75 191 (33.3) 260 (48.9) 20 (28.2) 28 (53.8)s-Ca (mmol/L)b 2.31 (0.11) 2.33 (0.08) 0.012 2.28 (0.12) 2.30 (0.06) 0.473

Glucocorticoids 0.159a

Never 545 (94.8) 505 (94.6) 72 (100) 52 (100)Previous 26 (4.5) 19 (3.6) – – –

Current 4 (0.7) 10 (1.9) – – –

Estrogen 19 (3.3) 3 (0.6) 0.001 NA NABisphosphonates 21 (3.7) 22 (4.1) 0.756 0 1 (1.9) –

SERM 30 (5.2) 1 (0.2) b0.001 – – –

Calciumsupplement

119 (20.7) 99 (18.5) 0.406 4 (5.6) 1 (1.9) 0.398

Vitamin Dsupplement

119 (20.7) 92 (17.2) 0.146 7 (9.7) 5 (9.6) 1.000

Menopauseb45 years

87 (15.1) 55 (10.3) 0.019 NA NA

Previous fracture 77 (13.4) 74 (13.9) 0.861 0 7 (13.5) –

Hip fracture in aparent

51 (8.9) 69 (13.0) 0.033 3 (4.2) 6 (11.5) 0.164

BMD femoralneck (g/cm2)

0.78 (0.12) 0.84 (0.13) b0.001 0.85 (0.09) 0.91 (0.13) 0.011

T-score femoralneck

b0.001a 0.075a

Normal BMD 77 (13.6) 211 (40.1) 13 (18.1) 19 (36.5)Osteopenia 283 (50.0) 263 (50.0) 47 (65.3) 27 (51.9)Osteoporosis 206 (36.4) 52 (9.9) 12 (16.7) 6 (11.5)

BMI: body mass index. BMD: bone mineral density. s-25(OH)D: serum 25-hydroxyvitamin D. s-Ca: serum calcium. SERM: selective estrogen receptor modulators. Normal BMD: T-score ≥ −1.0 SD. Osteopenia: T-score b −1.0, N −2.5 SD. Osteoporosis: T-score ≤ −2.5 SD. NA: not applicable.

a Overall p-value for the categorical variable.b s-Ca is corrected for serum albumin. Total numbers may vary between different variables according to different numbers of missing data.

Table 2Mean s-25(OH)D nmol/L in female and male patients and controls with normal BMD,osteopenia, and osteoporosis. Data are given as mean (SD).

Women Men

1142 J. Øyen et al. / Bone 48 (2011) 1140–1145

variables and chi-square tests for categorical variables in comparisonsbetween fracture patients and controls. Odds ratios (ORs) for distalradius fracture were estimated in unadjusted and adjusted condi-tional logistical regression analyzes for s-25(OH)D. In the adjustedanalyzes, adjustments were made for femoral neck BMD (g/cm2),body mass index (kg/m2), and smoking. A generalized additivelogistic regression model (GAM) was used for the dose–responserelation between vitamin D (s-25(OH)-D nmol/L), and OR for distalradius fractures. The models were adjusted for femoral neck BMD (g/cm2), body mass index (kg/m2), age and smoking. Two-tailed p-valuesb0.05 were considered statistically significant. The analyzeswere performed using SPSS software for Windows, version 15.0 (SPSSInc., Chicago, Illinois). For the GAM models in R we used the GAMlibrary maintained by Hastie and Tibshirani [36].

Patients(n=566)

Controls(n=526)

p-value

Patients(n=72)

Controls(n=52)

p-value

Normal BMD 70.0 (29.0) 76.4 (31.4) 0.111 65.6 (21.2) 73.7 (15.4) 0.220Osteopenia 66.7 (28.0) 80.9 (32.0) b0.001 62.9 (27.9) 78.2 (33.4) 0.038Osteoporosis 64.8 (31.0) 79.0 (47.7) 0.045 75.0 (32.0) 82.5 (39.1) 0.668

s-25(OH)D: serum 25-hydroxyvitamin D. Normal BMD: T-score ≥ −1.0 SD.Osteopenia: T-score b −1.0, N −2.5 SD. Osteoporosis: T-score ≤ −2.5 SD.

Ethics

The study was approved by the National Data Inspectorate and theRegional Committee for Medical Research Ethics. Each participantsigned an informed consent form.

Results

Characteristics of patients and controls

As a group, the female patients were older and had lower BMI, s-25(OH)D, and s-Ca than the female controls (Table 1). The use ofpostmenopausal estrogen therapy and SERM was higher amongpatients than controls, and more patients than controls reported early

Page 4

Table 3Odds ratio (OR) for distal radius fractures according to levels of s-25(OH)D in womenby unadjusted and adjusted conditional logistic regression.

Patients, n=575Controls, n=534

Unadjusted Adjusteda

OR 95% CI p-value OR 95% CI p-value

s-25(OH)D (nmol/L)≥75 1 1b75, ≥50 1.54 1.09–2.19 0.015 1.70 1.17–2.47 0.005b50 2.28 1.50–3.46 b0.001 2.32 1.47–3.64 b0.001

s-25(OH)D: serum 25-hydroxyvitamin D.a Adjusted for femoral neck bone mineral density (g/cm2), body mass index (kg/m2),

and smoking.

Table 4Odds ratio (OR) for distal radius fractures according to levels of s-25(OH)D in men byunadjusted and adjusted conditional logistic regression.

Patients, n=72Controls, n=52

Unadjusted Adjusteda

OR 95% CI p-value OR 95% CI p-value

s-25(OH)D (nmol/L)≥75 1 1b75, ≥50 2.56 0.76–8.62 0.128 2.62 0.69–9.93 0.156b50 6.14 1.43–26.44 0.015 6.27 1.17–33.66 0.032

s-25(OH)D: serum 25-hydroxyvitamin D.a Adjusted for femoral neck bone mineral density (g/cm2), body mass index (kg/m2),

and smoking.

1143J. Øyen et al. / Bone 48 (2011) 1140–1145

menopause. Furthermore, a higher proportion of female patientsreported previous fractures than female controls. However, thepercentage of hip fracture in a parent was higher among femalecontrols than patients. Finally, the female patients had significantlylower BMD and higher prevalence of osteoporosis than the femalecontrols. The male patients had significantly lower BMI, s-25(OH)D,and BMD than the male controls (Table 1).

Bone mineral density and serum 25-hydroxyvitamin D

Mean s-25(OH)D was significantly lower in female patients withosteopenia and osteoporosis compared to osteopenic and osteoporoticcontrols. In male patients the corresponding figures were significant forosteopenia (Table 2).

Fig. 1. Dose–response curves for the relation between vitamin D (s-25(OH)D nmol/L), and ouse of a generalized additive model (GAM). The curves are adjusted for femoral neck bonindicates the estimated dose–response curve and the dotted lines the 95% confidence inter

Serum 25-hydroxyvitamin D and risk of distal radius fracture

In both unadjusted and adjusted conditional logistic regressionmodels, low s-25(OH)D levels were associated with distal radius fracturein women (Table 3) and men (Table 4).

Exclusion of subjects who received medication for osteoporosis(e.g. bisphosphonates, SERM) did notmaterially alter the results (datanot shown).

Fig. 1 shows the adjusted dose–response relation (change in logodds) for s-25(OH)D for the distal radius fractures by generalizedadditive logistic regression (GAM model). The negative relationbetween s-25(OH)D and the change in log odds for distal radiusfracture was close to linear within the s-25(OH)D range both forwomen and for men, but the association tended to reach a thresholdlevel at approximately 100 nmol/L. Above this threshold, there was nostatistical decreasing risk for distal radius fractures for increasing s-25(OH)D levels. These results were also supported in conditional logisticregression analyzes with s-25(OH)D as a categorized variable (datanot shown).

Discussion

In this matched case–control study we found a lower mean s-25(OH)D in distal radius fracture patients compared to controls in bothwomen and men. This difference between patients and controls wasalso observed in subgroups of different BMD levels (osteopenia/osteoporosis). There was an inverse relationship between vitamin Dstatus and risk of fracture that was independent of BMD, BMI andsmoking.

To our knowledge, this case–control study is the first to comparevitamin D levels in distal radius fracture patients and matchedcontrols. Our control subjects werematched onmonth of examinationto adjust for varying sun exposure according to season. The study alsoincludes men with distal radius fractures. Male patients have notpreviously been studied so extensively. Selection bias may haveaffected our results as the participation rate was 45% in patients and39% in controls. Many fracture patients declining participation hadpreviously been evaluated for osteoporosis. If a large proportion ofthese have osteoporosis, the differences in BMD between cases andcontrols may have been underestimated. We lack information aboutcontrols refusing to participate, but the reason may be similar [37]. Ifso, this should not materially affect the reported differences invitamin D between fracture patients and controls. The patients were

dds ratio (OR) for distal radius fractures among women (left) and men (right) with thee mineral density (g/cm2), body mass index (kg/m2), age and smoking. The solid linevals.

Page 5

1144 J. Øyen et al. / Bone 48 (2011) 1140–1145

somewhat older and had lower BMI than the controls. However, thedifferences were small and probably without clinical significance andthe age matching was done with a possible difference of ±2 yearsbetween the patients and their matched controls. In any case,adjustment for these differences was done in the regression analyzes.The control subjects were included six months to four years after thepatients were assessed. This is clearly a weakness of the study andmight have introduced a bias as increased awareness of vitamin Dmay have changed the patterns of vitamin D supplementation. Bothpatients and controls were, however, examined at the sameosteoporosis clinic, by the same personnel, in the same month ofthe year, and with the same DXA device. The laboratory used the sameassay for s-25(OH)D during the whole period of the study [38].Vitamin D results could have been influenced by extraordinary sunexposure if the participants had been to sunnier countries or had useda solarium previous to the blood tests. Unfortunately, we did notcollect this information, but probably this influenced controls andpatients equally.

A dose–response relationship between vitamin D and distal radiusfracture was observed for vitamin D levels up to about 100 nmol/L forboth women and men. Above this level, there was no statisticaldecreasing risk for distal radius fractures for increasing vitamin Dlevels. In a report from the Institute ofMedicine [7] and in the study bySanders and co-workers [17] it was indicated that high levels ofvitamin D might be adverse for bone health outcomes. However,relatively few of the patients in our study had 25(OH)D levels above100 nmol/L as shown by the wide 95% CI (Fig. 1).

Vitamin D levels predicted fractures independently of differencesin BMD. Vitamin D may affect other bone strength factors than BMD.BMD predicts fractures but its sensitivity and specificity is quite low asgenerally many fractures occur in people without osteoporosis andmany people with osteoporosis do not sustain a fracture [39]. Theaccidental fall itself is the obvious cause of the fracture. In a previousincidence study from Bergen, about half of the distal radius fractureshappened during outdoor walking, and the number of fractures was3.6 times higher on days with snow on the ground compared to dayswithout snow [20]. Patients are presumably not more unlucky onslippery surfaces than their controls, and a higher fall tendencybecause of lower muscle strength caused by low vitamin D levels isprobably not the case among these patients as their vitamin D levelsare rather high [4,14,40]. Thus, the explanation might be the higherprevalence of low BMD and osteoporosis in the fracture patients. Thisis revealed in a previous study from the same population [23].

Vitamin D inadequacy has previously been studied in distal radiusfracture patients without a matched control group. In a Dutch study of100 distal radius fracture patients (74 women, 26 men) with a meanage of 67 years, the prevalence of osteoporosis (T-score ≤ −2.5)measured at total hip, spine or distal radius was 58%, and the vitaminD inadequacy (b 50 nmol/L) was 69% in the patients with low BMD (T-score ≤ −2.0) [30]. Of our female and male patients (mean age66 years), 51% had osteoporosis measured at femoral neck, and only34% of these had s-25(OH)D lower than 50 nmol/L. In a study of maledistal radius fracture patients in Northern Ireland (mean age 54 years)[29], the prevalence of osteoporosis was lower (11%) than in our malepatients (mean age 65 years) (18%), however, the prevalence ofvitamin D inadequacy was 49% compared to 33% in the male patientsin our study. Different measurement methods could affect the results,but generally, the vitamin D status has been measured to be higheramong Scandinavians than in people from other parts of Europe [41].The international Multiple Outcomes of Raloxifene Evaluation(MORE) study [42] showed high levels of mean s-25(OH)D inScandinavian patients with osteoporosis. This can partly be explainedby a traditionally high intake of cod liver oil in Norway [41,43], whichuntil a few years ago had a vitamin A content above recommendedlevels [44]. High intakes of vitamin A have been suggested to have anegative effect on BMD and risk of fractures [45,46]. Although the

amount of vitamin A in commercial cod liver oil has now beenreduced, a negative association between childhood cod liver oilintakes and BMD in a Norwegian population has been described [44].However, our distal radius fracture patients had lower levels of s-25(OH)D than the controls. Hence, a relative vitamin D inadequacyseems to be a risk factor for fractures. Genetic differences betweenethnic Scandinavians and other populations might be an explanationof why our population seems to need a higher vitamin D level thanother populations in order to prevent fractures [40].

Distal radius fractures occur on average 15 years earlier in life thanhip fractures [47,48], and may predict a risk of both subsequentvertebral and hip fractures [49]. To reduce the risk of future hip andvertebral fractures, it may be clinically appropriate to put more effortinto identifying risk factors in patients after their first low-energy distalradius fracture. BMD measurement with DXA is effective in clinicalpractice to diagnose osteoporosis in fracture patients [50]. Today, only3–20% of patients with distal radius fractures are examined for possibleosteoporosis, and 8–30% are treated for osteoporosis by medication[24,26,51,52]. s-25(OH)Dabove75 nmol/L is thought to be beneficial forhealth outcomes, like fracture risk [16] and risk of falling [15],while s-25(OH)D level of at least 50 nmol/L is stated as adequate for bone healthoutcomes in a North American population [7]. There is no consensus onthe optimal vitaminD level for bone health [9,10]. However, the currentstudy seems to confirm this assumption regarding distal radius fracturesin our population. With an increased focus on identifying these riskfactors, fracture preventive treatment for both osteoporosis [49,53] andvitamin D inadequacy [3] could be offered.

In conclusion, our study indicates that vitamin D inadequacy isassociated with low-energy distal radius fractures in both women andmen. Differences in vitamin D levels between patients and controlswere independent of differences in BMD, BMI or smoking history.

Disclosure statement

This work has been supported by research grants from The ResearchCouncil of Norway, The University of Bergen and The Western NorwayRegional Health Authority.

This work has not been presented or published elsewhere.

Acknowledgments

The authors are grateful to the medical staff at Bergen Accident andEmergency Department, the Department of Orthopaedic Surgery atHaukeland University Hospital and the technicians at the osteoporosisclinic at the Department of Rheumatology at Haukeland UniversityHospital. We would also like to thank PhD Bjørg Almås and PhD fellowMonika H.E. Christensen for implementation and interpretation ofvitamin D analyses.

References

[1] Holick MF. Vitamin D: a millenium perspective. J Cell Biochem 2003;88:296–307.[2] Lips P. Vitamin D deficiency and secondary hyperparathyroidism in the elderly:

consequences for bone loss and fractures and therapeutic implications. Endocr Rev2001;22:477–501.

[3] Bakhtiyarova S, Lesnyak O, Kyznesova N, Blankenstein MA, Lips P. Vitamin D statusamong patients with hip fracture and elderly control subjects in Yekaterinburg,Russia. Osteoporos Int 2006;17:441–6.

[4] Bischoff-Ferrari HA, Dietrich T, Orav EJ, Hu FB, Zhang Y, Karlson EW, et al. Higher25-hydroxyvitamin D concentrations are associated with better lower-extremityfunction in both active and inactive persons aged N or = 60 y. Am J Clin Nutr2004;80:752–8.

[5] Dawson-Hughes B, Heaney RP, Holick MF, Lips P, Meunier PJ, Vieth R. Estimates ofoptimal vitamin D status. Osteoporos Int 2005;16:713–6.

[6] Malabanan A, Veronikis IE, Holick MF. Redefining vitamin D insufficiency. Lancet1998;351:805–6.

[7] Ross AC, Manson JE, Abrams SA, Aloia JF, Brannon PM, Clinton SK, et al. The 2011report on dietary reference intakes for calcium and vitamin d from the institute ofmedicine: what clinicians need to know. J Clin Endocrinol Metab 2011;96:53–8.

Page 6

1145J. Øyen et al. / Bone 48 (2011) 1140–1145

[8] Cranney A, Weiler HA, O'Donnell S, Puil L. Summary of evidence-based review onvitamin D efficacy and safety in relation to bone health. Am J Clin Nutr 2008;88:513S–9S.

[9] Lips P, Bouillon R, van Schoor NM, Vanderschueren D, Verschueren S, Kuchuk N,et al. Reducing fracture risk with calcium and vitamin D. Clin Endocrinol Oxf2010;73:277–85.

[10] Bischoff-Ferrari HA. Optimal serum 25-hydroxyvitamin D levels for multiplehealth outcomes. Adv Exp Med Biol 2008;624:55–71.

[11] Grant WB, Juzeniene A, Moan JE. Health benefit of increased serum 25(OH)Dlevels from oral intake and ultraviolet-B irradiance in the Nordic countries. Scand JPublic Health 2011;39:70–8.

[12] Graafmans WC, Ooms ME, Hofstee HM, Bezemer PD, Bouter LM, Lips P. Falls in theelderly: a prospective study of risk factors and risk profiles. Am J Epidemiol1996;143:1129–36.

[13] Bischoff-Ferrari HA, Stahelin HB, Dick W, Akos R, Knecht M, Salis C. Effects ofvitamin D and calcium supplementation on falls: a randomized controlled trial. JBone Miner Res 2003;18:343–51.

[14] Glerup H, Mikkelsen K, Poulsen L, Hass E, Overbeck S, Andersen H, et al.Hypovitaminosis D myopathy without biochemical signs of osteomalacic boneinvolvement. Calcif Tissue Int 2000;66:419–24.

[15] Bischoff-Ferrari HA, Dawson-Hughes B, Staehelin HB, Orav JE, Stuck AE, Theiler R,et al. Fall prevention with supplemental and active forms of vitamin D: a meta-analysis of randomised controlled trials. BMJ 2009;339:b3692.

[16] Bischoff-Ferrari HA, Willett WC, Wong JB, Giovannucci E, Dietrich T, Dawson-Hughes B. Fracture prevention with vitamin D supplementation: a meta-analysisof randomized controlled trials. JAMA 2005;293:2257–64.

[17] Sanders KM, Stuart AL, Williamson EJ, Simpson JA, Kotowicz MA, Young D, et al.Annual high-dose oral vitamin D and falls and fractures in older women: arandomized controlled trial. JAMA 2010;303:1815–22.

[18] Chung M, Balk EM, Brendel M, Ip S, Lau J, Lee J, et al. Vitamin D and calcium: asystematic reviewofhealthoutcomes. EvidRepTechnolAssess Full Rep2009;1:1–420.

[19] Falch JA. Epidemiology of fractures of the distal forearm in Oslo, Norway. ActaOrthop Scand 1983;54:291–5.

[20] Hove LM, Fjeldsgaard K, Reitan R, Skjeie R, Sorensen FK. Fractures of the distalradius in a Norwegian city. Scand J Plast Reconstr Surg Hand Surg 1995;29:263–7.

[21] Lofthus CM, Frihagen F, Meyer HE, Nordsletten L, Melhuus K, Falch JA.Epidemiology of distal forearm fractures in Oslo, Norway. Osteoporos Int2008;19:781–6.

[22] Oyen J, Gjesdal CG, Brudvik C, Hove LM, Apalset EM, Gulseth HC, et al. Low-energydistal radius fractures in middle-aged and elderly men and women — the burdenof osteoporosis and fracture risk: a study of 1794 consecutive patients. OsteoporosInt 2010;21:1257–67.

[23] Oyen J, Brudvic C,Gjesdal CG, Tell GS, Lie SA, Hove LM.Osteoporosis as a risk factor fordistal radius fractures. A case–control study. J Bone Joint Surg Am 2011;93:1–9.

[24] Freedman BA, Potter BK, Nesti LJ, Cho T, Kuklo TR. Missed opportunities in patientswith osteoporosis and distal radius fractures. Clin Orthop Relat Res 2007;454:202–6.

[25] Khan SA, de Geus C, Holroyd B, Russell AS. Osteoporosis follow-up after wristfractures following minor trauma. Arch Intern Med 2001;161:1309–12.

[26] Gong HS, Oh WS, Chung MS, Oh JH, Lee YH, Baek GH. Patients with wrist fracturesare less likely to be evaluated and managed for osteoporosis. J Bone Joint Surg Am2009;91:2376–80.

[27] Vestergaard P, Rejnmark L, Mosekilde L. Osteoporosis is markedly under-diagnosed: a nationwide study from Denmark. Osteoporos Int 2005;16:134–41.

[28] LeBoff MS, Kohlmeier L, Hurwitz S, Franklin J, Wright J, Glowacki J. Occult vitaminD deficiency in postmenopausal US women with acute hip fracture. JAMA1999;281:1505–11.

[29] Wright S, Beringer T, Taggart H, Keegan D, Kelly J, Whithead E, et al. A study ofmale patients with forearm fracture in Northern Ireland. Clin Rheumatol 2007;26:191–5.

[30] Hegeman JH, Willemsen G, van Nieuwpoort J, Kreeftenberg HG, van der Veer E,Slaets JP, et al. Effective tracing of osteoporosis at a fracture and osteoporosis clinicin Groningen; an analysis of the first 100 patients. Ned Tijdschr Geneeskd2004;148:2180–5.

[31] Cummings SR, Nevitt MC. A hypothesis: the causes of hip fractures. J Gerontol1989;44:M107–11.

[32] Kanis JA, McCloskey EV, Johansson H, Oden A, Melton 3rd LJ, Khaltaev N. Areference standard for the description of osteoporosis. Bone 2008;42:467–75.

[33] Gjesdal CG, Aanderud SJ, Haga HJ, Brun JG, Tell GS. Femoral and whole-body bonemineral density in middle-aged and older Norwegian men and women: suitabilityof the reference values. Osteoporos Int 2004;15:525–34.

[34] Mazess RB, Barden H. Bone density of the spine and femur in adult white females.Calcif Tissue Int 1999;65:91–9.

[35] Assessment of fracture risk and its application to screening for postmenopausalosteoporosis. Report of a WHO Study Group, 843. World Health Organ Tech RepSer; 1994. p. 1–129.

[36] Hastie TJ, Tibshirani RJ. Generalized additive models. London: Chapman & Hall;1990.

[37] Buist DS, LaCroix AZ, Brenneman SK, Abbott 3rd T. A population-basedosteoporosis screening program: who does not participate, and what are theconsequences? J Am Geriatr Soc 2004;52:1130–7.

[38] ChristensenMH, Lien EA, Hustad S, Almas B. Seasonal and age-related differences inserum25-hydroxyvitaminD, 1, 25-dihydroxyvitaminD and parathyroid hormone inpatients fromWestern Norway. Scand J Clin Lab Invest 2010;70:281–6.

[39] Bouxsein ML, Seeman E. Quantifying the material and structural determinants ofbone strength. Best Pract Res Clin Rheumatol 2009;23:741–53.

[40] Cranney A, Horsley T, O'Donnell S, Weiler H, Puil L, Ooi D, et al. Effectiveness andsafety of vitamin D in relation to bone health. Evid Rep Technol Assess Full Rep2007;1:1–235.

[41] Lips P. Worldwide status of vitamin D nutrition. J Steroid Biochem Mol Biol2010;121:297–300.

[42] Lips P, Duong T, Oleksik A, Black D, Cummings S, Cox D, et al. A global study ofvitamin D status and parathyroid function in postmenopausal women withosteoporosis: baseline data from the multiple outcomes of raloxifene evaluationclinical trial. J Clin Endocrinol Metab 2001;86:1212–21.

[43] Brustad M, Sandanger T, Aksnes L, Lund E. Vitamin D status in a rural population ofnorthern Norway with high fish liver consumption. Public Health Nutr 2004;7:783–9.

[44] Forsmo S, Fjeldbo SK, Langhammer A. Childhood cod liver oil consumption andbone mineral density in a population-based cohort of peri- and postmenopausalwomen: the Nord-Trondelag Health Study. Am J Epidemiol 2008;167:406–11.

[45] Feskanich D, Singh V, Willett WC, Colditz GA. Vitamin A intake and hip fracturesamong postmenopausal women. JAMA 2002;287:47–54.

[46] Melhus H, Michaelsson K, Kindmark A, Bergstrom R, Holmberg L, Mallmin H, et al.Excessive dietary intake of vitamin A is associated with reduced bone mineraldensity and increased risk for hip fracture. Ann Intern Med 1998;129:770–8.

[47] Owen RA, Melton 3rd LJ, Ilstrup DM, Johnson KA, Riggs BL. Colles' fracture andsubsequent hip fracture risk. Clin Orthop Relat Res 1982:37–43.

[48] Bu FX, Armas L, Lappe J, Zhou Y, Gao G, Wang HW, et al. Comprehensiveassociation analysis of nine candidate genes with serum 25-hydroxy vitamin Dlevels among healthy Caucasian subjects. Hum Genet 2010;128:549–56.

[49] Mallmin H, Ljunghall S, Persson I, Naessen T, Krusemo UB, Bergstrom R. Fracture ofthe distal forearm as a forecaster of subsequent hip fracture: a population-basedcohort study with 24 years of follow-up. Calcif Tissue Int 1993;52:269–72.

[50] Leslie WD, Tsang JF, Caetano PA, Lix LM. Effectiveness of bone densitymeasurement for predicting osteoporotic fractures in clinical practice. J ClinEndocrinol Metab 2007;92:77–81.

[51] Freedman KB, Kaplan FS, Bilker WB, Strom BL, Lowe RA. Treatment ofosteoporosis: are physicians missing an opportunity? J Bone Joint Surg Am2000;82-A:1063–70.

[52] Talbot JC, Elener C, Praveen P, Shaw DL. Secondary prevention of osteoporosis:calcium, vitamin D and bisphosphonate prescribing following distal radialfracture. Injury 2007;38:1236–40.

[53] Earnshaw SA, Cawte SA, Worley A, Hosking DJ. Colles' fracture of the wrist as anindicator of underlying osteoporosis in postmenopausal women: a prospectivestudy of bone mineral density and bone turnover rate. Osteoporos Int 1998;8:53–60.