The role of vitamin D in improving physical performance in the elderly


  • Violet Lagari,

    Corresponding author
    1. Department of Medicine, University of Miami Miller School of Medicine, Miami, FL, USA
    2. Endocrinology Section, Miami Department of Veterans Affairs Medical Center, Miami, FL, USA
    • Address correspondence to: Violet Lagari, DO, University of Miami Miller School of Medicine, PO Box 016960 (D-56), Miami, FL 33101, USA. E-mail:

    Search for more papers by this author
  • Orlando Gómez-Marín,

    1. Department of Medicine, University of Miami Miller School of Medicine, Miami, FL, USA
    2. Departments of Epidemiology & Public Health and Pediatrics, University of Miami Miller School of Medicine, Miami Department of Veterans Affairs Medical Center, Miami, FL, USA
    Search for more papers by this author
  • Silvina Levis

    1. Department of Medicine, University of Miami Miller School of Medicine, Miami, FL, USA
    2. Geriatric Research, Education, and Clinical Center and the Bruce W Carter Veterans Affairs Medical Center, Geriatrics Institute, University of Miami Miller School of Medicine, Miami Department of Veterans Affairs Medical Center, Miami, FL, USA
    Search for more papers by this author

  • Presented at the 2012 Topical Meeting on Bone and Skeletal Muscle Interactions of the American Society for Bone and Mineral Research, Kansas City, MO, July 2012, and at the 2012 Annual Meeting of the American Society for Bone and Mineral Research, Minneapolis, MN, October 2012.


There is an ongoing debate over the role of serum 25(OH) vitamin D [25(OH)D] levels in maintaining or improving physical performance and muscle strength. Much of the controversy is because of the variability between studies in participants' characteristics, baseline serum 25(OH)D levels, and baseline physical functioning. The aim of this ancillary study conducted within a randomized controlled clinical trial was to investigate whether supplementation with 400 or 2000 IU vitamin D3 daily for 6 months would improve measures of physical performance and muscle strength in a community-dwelling elderly population aged 65 to 95 years. Those with the slowest gait speed improved their ability to do chair-stand tests after vitamin D supplementation. This finding remained significant after controlling for potential confounding variables. There was also an inverse correlation between serum 25(OH)D levels and fat mass index (FMI) among women, suggesting that higher supplementation with vitamin D is needed as weight increases. The results of this study suggest that supplementation with vitamin D may be most beneficial in older populations who have low baseline physical functioning. © 2013 American Society for Bone and Mineral Research.


The prevalence of low 25(OH) vitamin D [25(OH)D] serum levels is notably high among elderly populations.[1, 2] Hypovitaminosis D likely poses a significant global public health problem as multiple cross-sectional studies have associated low 25(OH)D with negative effects on musculoskeletal health,[3] chronic disease, and increased mortality.[4-6] Supplementation with vitamin D has been shown to decrease the risk of osteoporotic fractures and falls.[7, 8] The question remains if vitamin D supplementation resulting in increased 25(OH)D levels will improve or prevent chronic diseases such as diabetes and cancer, and/or improve or prevent deterioration of physical function in older individuals.

The consistent reduction in falls with vitamin D supplementation in older adults is particularly significant in those with low baseline 25(OH)D levels and is attributed to improvements in balance and strength, measures that are predictive of future mobility limitations.[9-12] In cross-sectional studies of older adults, for the most part, lower 25(OH)D levels have been associated with worse physical performance and decreased muscle strength.[13-20] However, prospective observational studies attempting to establish a relationship between 25(OH)D levels and future physical performance have reported inconsistent findings. In general, low baseline 25(OH)D status seems to be a determinant of progression to disability, suggesting that there may be a benefit from supplementation with vitamin D.[3, 21] Randomized controlled trials testing the effectiveness of vitamin D supplementation on physical performance have also yielded conflicting results.[9-12, 22-24] Among cross-sectional, longitudinal, and randomized trials, studies differ in their baseline serum 25(OH)D levels, frailty and functional status, statistical methods employed, target population (ie, institutionalized versus community dwelling), and the dose of vitamin D used.[13, 15, 25-30] The diverse target population in these studies is especially important in accounting for the variation in results. For example, although various studies may define their elderly target population as ambulatory and community dwelling, there may be differences in their degrees of mobility.[18, 31, 32] Inclusion criteria may allow for certain degrees of mobility disability, making comparisons of studies even more difficult.[13, 15, 25-30] Similarly, studies looking at inpatient or nursing home populations introduce a host of confounding variables that also impede comparisons between studies.[9, 11] It is possible that aggressive vitamin D supplementation may be beneficial if targeted at specific populations, which would require proper identification of groups that would benefit most.

In spite of the lack of evidence-based information on the effectiveness of vitamin D supplementation for many outcomes, increased awareness of the potential health benefits of vitamin D among health-conscious adults has resulted in increased self-prescribed supplementation with over-the-counter products. To begin to identify target groups that would benefit from vitamin D supplementation, we investigated the effectiveness of two doses of vitamin D3 in improving measures of physical performance that are known indicators of deteriorating health in an ambulatory, community-dwelling elderly population, regardless of their baseline 25(OH)D levels.[33-36]

Materials and Methods


Men and women enrolled in a vitamin D supplementation study were offered participation in this substudy of physical performance and body composition. The study design, results of screening, randomization, and 6-month follow-up of 25(OH)D levels have been previously reported.[37] Briefly, subjects were recruited in the Miami-Dade County area through community outreach programs, such as posting of flyers and presentations by the investigators in health fairs and to community organizations. Inclusion criteria included being community dwelling, aged 65 to 95 years, ambulation without an assistive device, vitamin D usual intake less than 1000 IU per day, and normocalcemia. Exclusion criteria included history of medical conditions associated with hypercalcemia, poor renal function, and malabsorption. The study was approved by the University of Miami IRB, and all participants signed an informed consent form.

Study design

Subjects were randomized, in a 1:2 ratio, to receive identically looking capsules containing either 400 IU or 2000 IU of vitamin D3 daily for 6 months (Tishcon Corp., Salisbury, MD, USA). This randomization scheme was chosen to maximize the potential benefit of being on the higher dose of 2000 IU vitamin D3.[37] Subjects were instructed to continue with the calcium and vitamin D supplementation they were already taking, if any. Participants were seen at a screening, randomization, and 6-month visit.

Outcome measures

The primary end points were physical performance measures at 6 months. Subjects had serum 25(OH)D and calcium, spot urine calcium and creatinine levels, physical performance, and body composition tested at baseline and at 6 months. Serum calcium, creatinine and albumin, and urinary calcium and creatinine were measured by spectrophotometry and colorimetric assay. Serum 25(OH)D was measured by liquid chromatography, tandem mass spectrometry [LC/MS/MS].

Physical performance measurements

Physical performance tests included 4-meter gait speed, the timed sit-to-stand, the single-leg balance, and the gallon-jug tests. The handgrip test was used as a measure of upper-extremity muscle strength. Body composition was assessed by dual-energy X-ray absorptiometry (DXA).

Four-meter walking speed test

Subjects completed a 4-meter walk test along a corridor at their normal walking speed. Gait speed was calculated by dividing 4 meters by the time in seconds that the subject took to complete the 4-meter walk.

Timed sit-to-stand test

This test is a measure of lower-limb strength.[34] Participants were asked to rise from a standard-height (43-cm) chair without armrests, as fast as possible and with their arms folded. Performance was measured in number of stands in 30 seconds, counting the time from the initial seated position to the final seated position.

Single-leg balance test

Subjects were instructed to stand on one leg without the support of the upper extremities or bracing of the unweighted leg against the standing leg. Subjects were asked to hold this position for as long as possible up to 60 seconds with their eyes open. The test was terminated if the foot touched the support leg, hopping occurred, the foot touched the floor, or arms touched something for support.

Gallon-jug shelf-transfer test

This test combines both motor-skill performance and physical working capacity.[38] The equipment for this test includes a bookshelf with adjustable shelves and five water-filled 1-gallon milk jugs. The lower shelf is aligned with the participant's patella and the upper shelf at the level of the top of the participant's shoulder. Each participant was asked to sequentially transfer the five jugs as quickly as possible from the lower shelf to the higher shelf, one jug at a time and without alternating hands. Performance was measured as the time to complete each trial was measured with a handheld stopwatch.

Handgrip test

Isometric handgrip force was used as an indicator of upper-extremity muscle strength and was measured using a JAMAR digital dynamometer, (Sammons Preston, Bolingbrook, IL, USA). The maximum strength (kg) of three attempts of the dominant hand was used.

Body composition measurements

Body composition (BC) was assessed by DXA using a Lunar iDXA scanner (Lunar Inc., Madison, WI, USA). Changes in body composition were determined by examining the variables total lean mass, lean mass of arms, lean mass of legs, and total fat mass. Total skeletal muscle mass (SM) was estimated in kilograms as described by Heymsfield and colleagues by determining appendicular lean soft-tissue mass of the arm and leg regions of the whole-body scan.[39] The appendicular skeletal muscle index (ASMI) was calculated by dividing appendicular muscle mass (kg) by height in meters squared. Similarly, fat mass was estimated by the fat mass index (FMI), which was calculated by total fat mass (kg) divided by height in meters squared, as described by Kelly and colleagues.[40]

Statistical analyses

All data from the 86 contributing patients were examined by intention-to-treat analysis. Relative changes in total lean mass, total fat mass, ASMI, FMI, grip strength, all physical performance measurements, and serum levels of 25(OH) vitamin D between the baseline and follow-up examination at 6 months were divided by their baseline respective values and multiplied by 100. The data were stratified according to gender. Data are presented for both genderes, but because of the small sample of male participants, results are mostly applicable to women. Baseline comparisons between randomization groups were performed to determine statistically significant differences among baseline variables that include age, BMI, race, ethnicity, smoking status, alcohol use, weight, height, 25(OH)D, calcium and vitamin D supplements (other than what the subject was randomized to), grip strength, and total lean and fat mass. Independent t tests were performed on continuous variables, and chi-square analyses were performed on dichotomous variables. Separate multivariate linear regression analyses were performed to determine correlations between relative changes in total lean mass, total fat mass, ASMI, FMI, physical performance measurements, and relative grip strength change (dependent variables) and relative change in vitamin D (independent variable), after adjusting for potential confounding variables.[41] Given that men and women differ with respect to strength and body composition, separate multivariate linear regression analyses were performed stratified by gender. Preliminary multivariate analyses revealed a significant gender and treatment interaction. Thus, within gender, separate correlation and regression analyses were performed to assess different interrelationships and the effect of treatment on relative changes in 25(OH)D, total lean mass, total fat mass, FMI, ASMI, single-leg stand, gallon-jug test, 4-meter walk, timed chair-stand test, and grip strength. Analyses were performed using SPSS for Windows (Version 17.0, SPSS Inc., Chicago, IL, USA).


All analyses were performed adjusting for potential confounding variables, which include age, race, gender, BMI, randomization group, calcium and vitamin D supplementation, smoking, alcohol consumption, and waist and hip circumference.


Baseline characteristics

Among the 105 subjects who were randomized in the vitamin D3 supplementation study, 86 (15 men, or 17%, and 71 women, or 83%) agreed to participate in this substudy of physical performance and body composition. Because a large proportion of the participants were females, these results are mostly applicable to women. The baseline demographic, clinical, physical performance, and body composition characteristics of the study participants are listed in Tables 1 and 2. Gender-specific comparisons yielded no significant differences between treatment groups with respect to baseline characteristics, with the exception of greater height and a longer single-leg balance test among men and smaller ASMI among women.

Table 1. Baseline Demographic and Clinical Characteristics of Study Participants According to Treatment Arm and Gender
Randomization groupAll (N = 86)Men (n = 15)Women (n = 71)
400 IU D3 (n = 6)2000 IU D3 (n = 9)p Value400 IU D3 (n = 25)2000 IU D3 (n = 46)p Value
  1. BMI = body mass index; 25(OH)D = 25-hydroxyvitamin D; PTH = parathyroid hormone.
Age (years ± SD)73.4 ± 6.474.8 ± 7.973.3 ± 6.30.68972.0 ± 5.474.0 ± 6.70.215
Race, n (%)   0.205  0.660
White81 (94)5 (83)9 (100) 24 (96)43 (91) 
Black5 (6)1 (7)0 1(4)3 (9) 
Ethnicity, n (%)   0.833  0.343
Not Hispanic27 (33)3 (50)5 (56) 5 (20)14 (30) 
Hispanic59 (67)3 (50)4 (44) 20 (80)32 (70) 
Smoking status, n (%)   0.114  0.676
Former and/or current28 (33)03 (33) 8 (32)17 (37) 
Never smoker58 (67)6 (100)6 (67) 17 (68)29 (63) 
Current drinker, n (%)   0.069  0.777
Yes34 (40)1 (17)2 (22) 10 (40)20 (43) 
No52 (60)5 (83)7 (78) 15 (60)26 (57) 
Calcium supplement intake (mg/d)428 ± 456366 ± 476122 ± 2440.210444 ± 487487 ± 4560.718
Vitamin D supplement intake (IU)239 ± 31767 ± 103178 ± 3380.454288 ± 342246 ± 3150.606
Weight (kg)65.3 ± 11.975.7 ± 12.679.1 ± 11.90.60065.4 ± 10.862.4 ± 11.30.172
Height (cm)158.6 ± 8.7165.2 ± 6.0174.5 ± 7.70.026156.7 ± 7.2155.6 ± 5.70.485
BMI (kg/m2 ± SD)25.9 ± 4.027.8 ± 4.926.0 ± 3.70.42326.4 ± 4.325.7 ± 4.40.277
Hip circumference (cm)101.3 ± 7.6104.8 ± 7.9101.6 ± 7.00.612102.4 ± 7.9100.3 ± 7.60.270
Waist circumference (cm)87.1 ± 11.3100.5 ± 12.895.4 ± 9.50.43665.0 ± 10.861.4 ± 9.70.198
25(OH)D (ng/mL ± SD)33.0 ± 10.031.3 ± 9.132.6 ± 15.50.86533.5 ± 8.433.0 ± 9.90.832
25(OH)D <20 ng/mL, n (%)7 (8)02 (13)0.2152 (3)3 (4)0.816
25(OH)D <30 ng/mL, n (%)30 (35)2 (13)4 (16)0.6678 (11)16 (23)0.813
Table 2. Baseline Physical Performance and Body Composition Characteristics of Study Participants According to Treatment Arm and Gender
Randomization groupAll (N = 86)Men (n = 15)Women (n = 71)
400 IU D3 (n = 6)2000 IU D3 (n = 9)p Value400 IU D3 (n = 25)2000 IU D3 (n = 46)p Value
Physical performance measurements
Gait speed (m/s)0.94 ± 0.21.0 ± 0.221.1 ± 0.160.6000.93 ± 0.10.91 ± 0.20.628
Single-leg balance (s)18.8 ± 19.89.0 ± 8.936.5 ± 25.60.03019.3 ± 19.916.8 ± 18.50.633
Chair stands (n/30 s)15.3 ± 4.515.3 ± 6.120.1 ± 4.00.08313.7 ± 4.015.1 ± 3.80.208
Grip strength (kg)23.8 ± 6.730.0 ± 9.930.6 ± 6.60.83622.4 ± 4.222.3 ± 6.30.950
Gallon-jug test (s)10.7 ± 2.49.5 ± 1.58.4 ± 1.60.20711.0 ± 2.011.1 ± 2.50.815
Body composition
Appendicular skeletal muscle mass index (ASMI) (kg/m2)6.4 ± 1.18.3 ± 0.667.9 ± 0.730.2456.3 ± 0.85.9 ± 0.720.029
Fat mass index (FMI) (kg/m2)10.0 ± 3.19.3 ± 3.77.3 ± 2.70.23510.7 ± 3.110.3 ± 2.90.528
Skeletal muscle mass (%)58.7 ± 7.163.6 ± 7.262.7 ± 9.00.84857.4 ± 5.358.0 ± 7.20.692
Fat mass (%)38.1 ± 7.532.9 ± 7.633.9 ± 9.40.99139.6 ± 5.638.9 ± 7.60.688
Fat mass (kg)24.8 ± 7.221.5 ± 4.924.4 ± 10.60.54426.0 ± 6.624.6 ± 7.10.423
Skeletal muscle mass (kg)37.9 ± 7.742.2 ± 9.144.6 ± 12.10.69231.1 ± 5.236.4 ± 7.00.665

25-hydroxyvitamin D levels after supplementation

Results of supplementation with either 400 or 2000 IU of vitamin D3 for 6 months have been published elsewhere.[37] In brief, the majority of participants had normal baseline 25(OH)D levels. The study found that either dose was more effective in raising and maintaining levels of 25(OH)D when baseline levels were lower than 30 ng/mL. However, results suggest that 400 IU of vitamin D3 daily may not be sufficient as a maintenance dose. Among those receiving 400 IU daily, there was a drop in 25(OH)D levels in the subjects with baseline values ≥30 ng/mL; in those with baseline levels <30 ng/dL, there was an increase in 25(OH)D levels, but this increase was not statistically significant.

Physical performance outcomes

Overall, 6 months of supplementation with vitamin D3, either 400 IU or 2000 IU daily, did not result in statistically significant changes in physical performance or strength. Absolute changes in physical performance outcomes are listed in Table 3. However, separate multivariate regression analyses showed a statistically significant improvement in the timed sit-to-stand test in the subjects whose gait speed was in the lowest tertile (≤0.8576 m/s) (Fig. 1). Among the 22 participants who fell in the subcategory of having baseline gait speed ≤0.8576 m/s, there was a statistically significant correlation between relative change in chair stands (mean 5.1 ± 21.7%) and relative change in 25(OH)D level (mean 14.2 ± 43.4%), p = 0.033. This remained statistically significant after controlling for potential confounding variables. Of these 22 participants, 10 were receiving 400 IU and 12 were receiving 2000 IU of vitamin D3 daily.

Table 3. Absolute Changes in Physical Performance and in Body Composition According to Treatment Arm and Gender
Randomization groupAll (N = 86)Men (n = 15)Women (n = 71)
400 IU D3 (n = 6)2000 IU D3 (n = 9)p Value400 IU D3 (n = 25)2000 IU D3 (n = 46)p Value
Absolute change in:
Gait speed (m/s)0.02 ± 0.11–0.10 ± 0.18–0.01 ± 0.100.2910.05 ± 0.100.03 ± 0.090.438
Single-leg balance, (s)–1.7 ± 11.93.8 ± 8.2–8.6 ± 13.30.0810.62 ± 1.14–2.7 ± 12.40.326
Chair stands (n/30 s)–0.17 ± 11.9–5.0 ± 1.50.67 ± 3.60.4770.26 ± 3.1–0.50 ± 3.70.453
Grip strength (kg)–0.9 ± 2.9–1.0 ± 4.51.3 ± 2.40.252–1.3 ± 2.5–1.1 ± 2.80.776
Gallon-jug test (s)–0.34 ± 1.30.35 ± 1.1–0.52 ± 0.740.128–0.62 ± 1.1–0.28 ± 1.40.395
Appendicular skeletal muscle mass (ASMI) (kg/m2)0.00 ± 0.31–0.08 ± 0.23–0.12 ± 0.500.8770.09 ± 0.28–0.02 ± 0.30.162
Fat mass index (FMI) (kg/m2)0.09 ± 0.610.36 ± 1.10.14 ± 0.680.6440.06 ± 0.670.06 ± 0.520.956
Lean mass (kg)–0.19 ± 2.2–3.57 ± 6.42–0.62 ± 2.450.2290.18 ± 0.990.13 ± 1.00.826
Fat mass (kg)–0.16 ± 2.0–1.58 ± 6.210.47 ± 1.580.3550.09 ± 1.38–0.23 ± 1.30.334
25(OH)D (ng/mL)2.4 ± 12.0–1.2 ± 5.96.1 ± 12.60.208–3.4 ± 10.85.3 ± 12.30.004
Figure 1.

Relative change in 25(OH)D (%) versus relative change in chair-stand test scores (%) among men and women with baseline gait speed scores in the lowest tertile (p = 0.033).

Gender-specific comparisons of relative changes in physical performance measures, handgrip strength, body composition measurements, and 25(OH)D levels only yielded significant differences between treatment groups for relative change in 25(OH)D levels among women, with a greater increase among women randomized to the 2000 IU vitamin D3 group versus 400 IU group (24.5 ± 51.1 versus –6.3 ± 28.1, p = 0.007) (Table 3). Also among women, increases in FMI correlated with declining levels of 25(OH)D, suggesting that supplementation with higher doses is required with increases in weight (Fig. 2). This relationship remained significant after adjusting simultaneously for treatment, age, smoking, alcohol consumption, calcium and nonstudy vitamin D supplementation, race, ethnicity, waist and hip circumference, and BMI.

Figure 2.

Relative change in 25(OH)D (%) versus relative change in total fat mass (%) among women (p = 0.027).

When analyzing subgroups, there were no statistically significant correlations observed in relative change in 25(OH)D and relative change in physical performance or handgrip strength measurements among Hispanics versus non-Hispanics or among those participants with baseline 25(OH)D levels below or above 20 ng/mL.


Compliance among the 86 subjects in this substudy was excellent, with overall compliance of 90.5%. Among participants receiving 400 IU of vitamin D3, mean compliance with study medication was 95.9%, and in the group randomized to 2000 IU, compliance was 95.1%.


In this study of community-dwelling older persons, increases in 25(OH)D levels improved lower-extremity strength in those with the poorest baseline physical functioning. Supplementation with vitamin D improved chair-stand test scores among those who had the slowest gait speed scores. This finding remained significant after controlling for potential confounding variables and is mostly applicable to women because few men participated in the trial.

It is challenging to explain the improvement in the chair-stand test but the lack of response in the other tests of physical performance. The chair-stand test demands mostly proximal lower-extremity muscle strength. One can speculate that our finding is comparable to the clinical observations in cases of severe vitamin D deficiency, which, in addition to osteomalacia, often presents with marked proximal muscle weakness.[42] In this respect, our results correlate with those of two other studies. Bischoff and colleagues demonstrated an improvement after vitamin D supplementation in the Timed Up & Go test, which includes standing from a chair before the gait component of the test.[43] In addition, a recent cross-sectional study of postmenopausal osteoporotic women demonstrated a weak but positive correlation between 25(OH)D levels and knee muscle strength and postural balance during the act of standing.[44]

Our results have two important implications. First, vitamin D supplementation may improve physical functioning only in the frailest individuals, as shown in other randomized clinical trials conducted on nursing home subjects, where physical function was noted to improve with vitamin D supplementation.[9-12] Bunout and colleagues demonstrated that in elderly nursing home residents with low baseline 25(OH)D levels, subjects who were supplemented with 400 IU vitamin D3 daily for 9 months had significant improvements in gait speed and other tests of muscle strength and physical performance.[12] Other intervention studies that supplemented vitamin D in community-dwelling populations yielded conflicting results regarding effects on gait speed, muscle strength, and other tests of physical performance.[24, 45, 46] In addition to using a community-dwelling and presumably less frail population, the differences noted in these studies also stem from different baseline 25(OH)D levels, different vitamin D metabolites analyzed, and different assessment methods for evaluation of muscle strength and physical performance.

Although the majority of observational longitudinal studies show a positive correlation between serum 25(OH)D and tests of muscle strength and physical performance,[8, 10, 21, 22, 47-49] the critical question as to who will benefit most from vitamin D supplementation remains to be determined. Our results suggest that it may be important to target high-risk populations that are most vulnerable to frailty and falls. The lack of response in nonfrail individuals is supported by the findings of a recent prospective observational study of highly functional elderly men with normal mean 25(OH)D levels, 77.9 nmol/L (31 ng/mL). The majority (94%) of participants had 25(OH)D levels greater than 50 nmol/L (20 mg/mL), and there was no correlation between serum 25(OH)D levels and baseline or 4-year change in physical performance measures.[18] Thus, there may be no role for additional vitamin D supplementation in those individuals who are vitamin D replete and/or have high baseline physical functioning. Although our study population did not have low average baseline serum 25(OH)D levels, the finding that improvements in physical performance were only achieved among those with slowest baseline gait speed provides support for vitamin D's benefits in frail populations.

Second, our study confirmed that in women, higher fat mass was associated with declining levels of vitamin D, supporting the general belief that obese individuals require supplementation with higher vitamin D doses. This has already been demonstrated in previous studies and, as vitamin D is liposoluble, it has been explained as the result of vitamin D sequestration and storage in fat.[50] Although this is also likely the case in men, because of the small number of males enrolled in this study, we are unable to confirm these findings for men.

There are several limitations of this study. The study population was relatively small, there were few men enrolled, and the subset of patients who had low baseline physical functioning was also small. The average baseline 25(OH)D levels were not low in this community-dwelling population, and it is possible that further benefit could have been observed if baseline levels had been lower. In addition, the overall design of the study did not include a placebo control group, which could make it seem as though vitamin D supplementation may not have had the desired therapeutic effect, rather than the result of a possible “ceiling effect.” Despite these limitations and even after controlling for potentially confounding variables, there was a statistically significant improvement in chair-stand test scores among those with slow baseline gait speed.

Although the results of this small substudy cannot be applied to the general population, findings of this randomized clinical trial assessing the effects of vitamin D supplementation on tests of muscle strength and physical performance suggest a benefit in older women with poor physical performance, regardless of baseline 25(OH)D levels. The significant inverse correlation between relative changes in total fat and in 25(OH)D supports previous evidence that higher supplemental doses are needed in persons who are overweight or obese. Because of the small sample size, this study cannot determine if there is a need for rigorous or long-term monitoring when providing vitamin D3 2000 IU daily to older persons.


This study was supported by an investigator-initiated grant from Merck & Co. and by the Marjorie Cowan Family Foundation. The sponsors had no input regarding study design, conduct, or data analysis. All authors state that they have no conflicts of interest.


We thank Luis Rochel, RN, and Rafael Franjul for their assistance in the conduction of this study; Joseph Signorile, PhD, Sara Mow, and Les Bradley for their assistance with the physical performance assessments; and Sean Rinehart, Eric Ponce, and Stephen DeGennaro for their help with data management.

Authors' roles: Study design: SL. Study conduct: SL. Data collection: SL. Data analysis: VSL and OG. Drafting manuscript: VL and SL. Revising manuscript content: VSL, SL, and OG. Approving final version of manuscript: VSL, SL, and OG. VSL takes responsibility for the integrity of the data analysis.