Vitamin D3 for Life

Live Enlightened

Article

Vitamin D, Physical Performance and Falls: Full Evidence Review

Previous related pageVitamin D, Muscle Respiration and Mitochondrial Signalling: Study SummaryNext related pageVitamin D, Physical Performance and Falls: Study Summary
Study at a glance

How was vitamin D tested for movement and falls in older adults?

Two trials compared vitamin D with a placebo, a treatment without vitamin D, in older adults. One trial, called D-Vitaal, tested daily supplements for a year in 155 people with depressive symptoms, difficulties with everyday activities and relatively low vitamin D levels. The other, called D-Health, tested monthly supplements for up to five years in a much larger group whose vitamin D levels were generally adequate.

Did supplements improve movement or reduce falls overall?

D-Vitaal raised blood vitamin D but did not improve overall walking, chair-standing, balance, grip strength or reported limits on everyday activities compared with placebo. D-Health did not show an overall reduction in falls. These findings separate a rise in blood vitamin D from a demonstrated improvement in movement or fall prevention.

Were there benefits or possible harms in smaller groups?

A small D-Vitaal subgroup reported fewer limits on everyday activities, but physical performance did not clearly improve and the finding could be due to chance. In D-Health, participants with a body mass index below 25 had more falls with vitamin D in both survey and diary analyses. This possible harm signal matters, but subgroup results need caution because many comparisons were made.

Do these findings settle how severe deficiency should be treated?

D-Vitaal included few people with very low vitamin D levels. D-Health did not measure everyone’s starting vitamin D and mainly studied people with adequate levels. Its annual surveys covered falls in the preceding month, rather than every fall over five years. Neither trial settles treatment of severe deficiency or identifies a personal dose. The results do not support assuming that more vitamin D improves physical performance or prevents falls.

Read the full article below ↓

Two placebo-controlled trials separate a rise in blood vitamin D from a measurable improvement in movement or fewer falls. D-Vitaal found no overall physical-performance benefit in older adults selected for symptoms, functional difficulty and relatively low vitamin D status. D-Health found no overall fall-prevention benefit in a much larger, generally vitamin D-replete older population. For the companion appraisal, see Vitamin D, Physical Performance and Falls: Study Summary.

The question behind low vitamin D and poor performance

A low vitamin D concentration and impaired movement can occur together without proving which caused the other. Less outdoor activity, illness and other characteristics may affect both. A randomised supplementation trial tests whether changing intake improves an outcome under its particular conditions. It does not experimentally create deficiency, and a null treatment result does not establish that untreated severe deficiency is harmless.

The physiology described in Vitamin D and Skeletal Muscle Function makes a muscle effect plausible. Clinical consequences still require direct measurement. Strength, walking, balance and falling are related but distinct outcomes: a fall also depends on vision, medication, cognition and environmental hazards. Reading them as interchangeable would obscure what either trial actually established.

D-Vitaal selected participants with symptoms and relatively low status

The Dutch D-Vitaal trial randomised 155 community-dwelling adults aged sixty to eighty years. Participants had clinically relevant depressive symptoms, at least one reported functional limitation, and screening 25-hydroxyvitamin D between 15 and 50 nmol/L in winter or between 15 and 70 nmol/L in summer. The seasonal upper limits matter: not every participant was below 50 nmol/L at entry. Baseline medians were approximately 46 and 44 nmol/L in the intervention and placebo groups.

Participants received 1,200 IU/day vitamin D3 or identical placebo for twelve months. Both groups could continue up to 400 IU/day from other supplements. Calcium intake was supported through dietary advice, with calcium tablets for participants reporting low dairy intake. The study therefore tests an assigned regimen within that background care, rather than isolated vitamin D exposure without any accompanying nutritional advice. These quantities describe research methods, not an individual prescription.

Performance improved neither consistently nor selectively in low-status participants

The primary physical outcomes were reported functional limitations and a modified Short Physical Performance Battery combining walking, repeated chair stands and balance. Grip strength and timed up-and-go were secondary measurements. At six months, mean 25-hydroxyvitamin D was 85 nmol/L with supplementation and 43 nmol/L with placebo. The clear biochemical separation did not produce an overall advantage in physical performance, functional limitations, grip strength or timed up-and-go.

For the adjusted performance analysis, the estimated between-group difference was −0.20 points, with a 95% confidence interval from −0.79 to 0.38. Grip strength likewise showed no established benefit. These estimates are more informative than saying only that a significance threshold was missed: they show the direction and uncertainty of the observed contrast. Vitamin D Status vs Vitamin D Effect provides the relevant distinction between changing a blood marker and changing function.

An exploratory subgroup with starting concentrations at least 50 nmol/L reported fewer functional limitations with supplementation. The difference was small, did not establish a corresponding performance benefit and was absent in the lower-status subgroup. The authors regarded it as likely a chance finding. The trial was not powered for those subgroup comparisons. The result must remain visible, but it cannot justify selecting people for treatment by that threshold.

What D-Vitaal leaves uncertain

The physical-performance analysis included 151 of the 155 randomised participants because four lacked adequate follow-up. Dropout was relatively low and adherence was high, supporting the interpretation that the main null finding was not simply a failure to change exposure. Nevertheless, only about sixteen percent entered below 30 nmol/L. Small subgroup numbers limit conclusions about more marked deficiency, and the eligibility criteria excluded concentrations below 15 nmol/L.

Depressive symptoms and functional difficulty defined this population. Results cannot automatically be extended to every older adult, to institutional care or to a patient with a diagnosed deficiency-related myopathy. Adverse-event totals did not differ significantly between groups; one participant in the intervention group died. The few reported falls in the adverse-event table were not a sufficiently powered falls-prevention experiment. The separate D-Health trial addresses falls directly.

D-Health tested a different population and dosing schedule

D-Health randomised 21,315 Australians aged sixty to eighty-four to monthly 60,000 IU vitamin D3 or placebo for up to five years. Five withdrew consent and their data were removed. Falls were a prespecified tertiary outcome of a trial primarily designed around mortality. To preserve blinding for other analyses, the investigators used randomly selected subsets rather than presenting every participant as part of this falls analysis.

The annual-survey analysis included 15,416 participants with at least one response from a sample of 16,000. A separate three-month diary analysis included 2,200 respondents, with 2,093 in the fully adjusted comparison. The main falls endpoint asked whether a participant had fallen in the month before each annual survey. This provides repeated snapshots, not continuous observation of every fall throughout five years. The diaries offered more detailed prospective recording over a much shorter interval.

No overall reduction in falling

The odds ratio for a reported fall in the preceding month was 1.02, with a 95% confidence interval of 0.95 to 1.10. The fully adjusted diary comparison was also compatible with no effect: odds ratio 1.07, confidence interval 0.84 to 1.36. Diary fall incidence showed no clear reduction either. These are different analyses with different precision, but neither demonstrated the proposed overall prevention benefit.

An odds ratio is not the percentage of participants who fell, and it is not automatically an identical relative risk. The confidence interval also does not prove exact equivalence. Here the large survey analysis constrains the size of plausible overall effects more tightly than the smaller diary analysis. The trial concerns Vitamin D in Older Adults, but it did not directly measure a microscopic contraction mechanism or explain every contributor to an individual fall.

A safety signal in the lower-BMI subgroup

In the prespecified subgroup with body mass index below 25 kg/m², supplementation was associated with more falls in both survey and diary analyses. The survey odds ratio was 1.25, with a 95% confidence interval of 1.09 to 1.43; the diary estimate was 1.72, with an interval of 1.09 to 2.73. The corresponding interaction tests supported a difference between BMI strata. Participants with higher BMI did not show the same pattern.

This signal deserves attention alongside the overall null result. Multiple comparisons were not adjusted, and subgroup interpretation remains less secure than the principal analysis. Exploratory exclusions of underweight participants did not remove the pattern, but they do not establish its mechanism. Hypercalcaemia and kidney-stone incidence did not differ materially between groups. Vitamin D and Calcium therefore supplies physiological context without allowing the falls signal to be relabelled proven calcium toxicity.

Measured and predicted vitamin D status are different evidence

D-Health did not measure baseline vitamin D in every participant. It predicted lower starting status using a model developed from placebo-group samples. The model had only moderate discrimination and a low positive predictive value for concentrations below 50 nmol/L. Survey subgroup estimates suggested a more favourable effect at predicted lower status, but the diary analyses did not confirm that pattern. Prediction error and inconsistent outcomes prevent treating it as a demonstrated benefit in deficient people.

During follow-up, sampled mean concentrations were approximately 115 nmol/L in the vitamin D group and 78 nmol/L in placebo recipients. Only thirteen percent of placebo samples were below 50 nmol/L. The trial consequently offers strong information about the tested monthly regimen in a largely replete population, while leaving falls in populations with prevalent severe deficiency unresolved. It should not be described as a direct replication of D-Vitaal’s daily regimen and symptom-selected participants.

Measurement and follow-up limit the inference

D-Health added an explicit definition of falling from the third annual survey. Reports increased in both groups at that point, showing why a change in the questionnaire should not be mistaken for a sudden treatment effect. Excluding participants without any outcome data also makes the analysis a modified intention-to-treat comparison. The diary group was selected later in follow-up among people who remained enrolled; high adherence in that subset does not establish identical experience among everyone originally randomised.

The D-Vitaal measurements illustrate a different distinction. Reported difficulty with daily activities is meaningful to participants, but it does not measure exactly the same thing as timed walking, chair stands, balance or dynamometer grip strength. A favourable questionnaire subgroup without a corresponding performance effect cannot establish stronger muscle. Conversely, a small physical test battery cannot exclude every possible physiological response. The correct inference remains tied to the measured outcomes and their precision, rather than an unrestricted claim of either benefit or biological inactivity.

How the trials fit muscle physiology

The results do not remove the importance of Vitamin D Receptors in experimental muscle biology. Receptor deletion, low circulating precursor and additional oral supplementation are different perturbations. An intact pathway may contribute to normal tissue function without a higher intake producing better performance. Vitamin D Activation adds another distinction: an administered dose, the circulating precursor and local active-hormone signalling are not the same measurement.

For Vitamin D and Muscle Function, D-Vitaal provides a direct performance test in a relatively low-status symptomatic group, while D-Health provides a much larger falls test in a different population. Neither study traces a causal sequence from vitamin D insufficiency through a particular muscle defect to a fall. Their combined contribution is to constrain treatment claims and expose the remaining population-specific uncertainty, rather than to turn an association into a universal cause or cure.

The D-Health supplementary year-by-year analysis included one favourable second-year survey estimate, with an odds ratio of 0.84 and a 95% confidence interval of 0.71 to 0.99. Other yearly estimates and the combined analysis did not establish an overall reduction. This isolated time-point result should remain visible without replacing the overall interpretation.

Conclusion

These trials raised blood vitamin D without demonstrating overall improvements in the physical outcomes tested. D-Vitaal’s small favourable subgroup result and D-Health’s possible harm signal should be reported with their limitations. The findings argue against assuming that higher vitamin D exposure improves movement or prevents falls. They do not settle treatment of profound deficiency, identify a universal safe dose or supply a personal treatment schedule.

Sources and publication details

Vitamin D supplementation for the prevention of depression and poor physical function in older persons: the D-Vitaal study, a randomized clinical trial. de Koning EJ and colleagues. American Journal of Clinical Nutrition. 2019;110(5):1119 to 1130. DOI: 10.1093/ajcn/nqz141. PMID: 31340012.

Vitamin D supplementation and risk of falling: outcomes from the randomized, placebo-controlled D-Health Trial. Waterhouse M and colleagues. Journal of Cachexia, Sarcopenia and Muscle. 2021;12(6):1428 to 1439. DOI: 10.1002/jcsm.12759. PMID: 34337905.

Related reading

Original sources

Source

Vitamin D supplementation for the prevention of depression and poor physical function in older persons: the D-Vitaal study, a randomized clinical trial

Published in de Koning EJ and colleagues. American Journal of Clinical Nutrition (2019).

Publication details

Vitamin D supplementation for the prevention of depression and poor physical function in older persons: the D-Vitaal study, a randomized clinical trial. de Koning EJ and colleagues. American Journal of Clinical Nutrition. 2019;110(5):1119 to 1130. DOI: 10.1093/ajcn/nqz141. PMID: 31340012.

Read more →