Once Yearly Vitamin D for Falls and Fractures: Not A Good Idea

Abstract & Commentary

By David Kiefer, MD. Dr. Kiefer is Clinical Instructor, Family Medicine, University of Washington, Seattle; Clinical Assistant Professor of Medicine, University of Arizona, Tucson; and Adjunct Faculty, Bastyr University, Seattle; he reports no financial relationships relevant to this field of study.

Synopsis: The oral administration of 500,000 IU cholecalciferol in fall or winter resulted in a slightly higher risk of falls and fractures vs placebo in 2,256 community-dwelling women aged 70 and above.

Source: Sanders KM, et al. Annual high-dose oral vitamin D and falls and fractures in older women: A randomized controlled trial. JAMA 2010;303:1815-1822.

The aim of this study was to clarify some of the conflicting results in the literature about whether or not vitamin D supplementation can help to prevent falls or fractures. A variety of demographics and dosing regimens previously have been examined; the researchers behind this study chose a dose of vitamin D known to prevent winter drops in serum 25-hydroxyvitamin D (25(OH)D), and a regimen (once yearly) that would address concerns about adherence that may have affected past studies' results.

In the area of southern Victoria, Australia, recruitment letters were sent to community-dwelling women 70 years old or older who were on a list of registered voters. Women were included in the trial if they were considered "high risk" for fracture, that is, if they had maternal hip fracture, past fracture, or had fallen by self-report. Women were excluded if they were in a high-level care facility, could not provide falls or fracture data, had elevated serum calcium or creatinine levels, were taking vitamin D in a dosage greater than 400 International Units (IU) daily, or were on anti-fracture pharmaceutical treatment. Out of this effort, 2317 women consented to participate, but then 59 withdrew prior to the randomization, leaving 2258 women to be randomized to receive 500,000 IU cholecalciferol (vitamin D3) orally (50,000 IU daily for 10 days) or identical placebo each fall or winter for 3-5 years. These women were contacted regularly by postcard or e-mail about whether or not they had a fall or fracture event; if such an event occurred, follow-up by telephone took place, and all reported fractures had to be corroborated by a radiograph prior to inclusion in the analysis. There were a total of 226 withdrawals (a variety of reasons, similar between treatment and placebo groups) during the study period, but all of these were accounted for in the final analysis by intention-to-treat, increasing the statistical validity of the results.

There were a total of 5404 falls, including 2892 falls in the vitamin D group vs 2512 in the placebo group; the relative risk (RR) was 1.15 (confidence interval [CI], 1.02-1.30; P = 0.03), just barely significant. A statistically significant (RR = 1.31, CI = 1.12-1.54) increased incidence in falls occurred within 3 months of dosing. With respect to total fractures, there were 171 in the vitamin D group, and 135 in the placebo group, with an RR = 1.26 (CI = 1.00-1.59, P = 0.047), again just barely statistically significant. The RR for nonvertebral fractures was not significant between the treatment and placebo groups. There was not a statistically significant temporal effect of dosing on fractures.

A questionnaire about calcium intake found no difference between the treatment and placebo groups; the average intake was 976 mg daily. One hundred thirty-seven people (75 from vitamin D group, 58 from placebo group) consented to blood tests; the median baseline 25(OH)D was 49 nmol/L (no difference between the two groups), and 12 months after dosing the vitamin D group showed marked increases in 25(OH)D (median 55-74 nmol/L, range 25-120 nmol/L), which were significantly higher than the placebo group each year (P < 0.05).

Adverse events, including serious ones, were similar between the treatment and placebo groups, barring injury (which included fracture) that, as mentioned above, occurred more in the vitamin D group (15.2% vs 12.1%, P = 0.03).


Investigations into the use of vitamin D cover many medical conditions, but the primary focus of research has been on bone health and the prevention of falls and fractures. In addition, numerous dosing regimens (dose, frequency, form) have been studied, as reviewed in a past issue of Alternative Medicine Alert.1 When interpreting new vitamin D research results, it is important to keep all of these details in mind, as well as the specific demographic involved, baseline characteristics (diet, serum 25(OH)D, etc.), and any confounding supplement or nutritional variables.

What could explain the surprise results of this approach to vitamin D repletion, which led not to an improvement in bone health, but rather, an increased risk of falls and fractures? Some nuances in the study design may help to answer this question. For example, unique aspects of this study are the use of annual vitamin D dosing, and the fact that the study subjects were community-dwelling women, not those in long-term care facilities as in many past research efforts. In addition, the researchers also did not provide supplemental calcium; most past vitamin D supplementation studies that had a positive effect on fall or fracture risk also included supplemental calcium. The baseline calcium intake for women in both arms of this study was reasonably close to most recommendations, so it is unclear whether or not the lack of calcium supplementation would have made any difference in the results noted.

The authors mention another study with increased fracture risk in women, but not men, receiving 300,000 IU intramuscular ergocalciferol (vitamin D2) once annually.2 Interestingly, fall risk was not increased, though, again, the treatment group did not receive supplemental calcium. There might, indeed, be a risk with high, single-dose treatments without added calcium. The authors note that serum 25(OH)D peaked 1 month after the oral dosing, gradually decreasing (half-life of vitamin D is 60-90 days) to levels 41% higher than the placebo group after 1 year, and they also documented a temporal effect on increased fracture risk at 3 months. Were some patients exceeding a safe level of serum 25(OH)D within the 3 months after dosing? That is possible, although the authors point out that the highest serum 25(OH)D in the subgroup that underwent blood testing reached 208 nmol/L at 1 month, while "toxic" levels are thought to be in the range of 375-500 nmol/L. Nonetheless, the authors postulate a possible connection between temporal high serum 25(OH)D, vitamin D metabolites, or the resulting decrease in levels that "might be causal."

With respect to repletion, the dose in this study was probably sufficient. The 500,000 IU once annually, as referenced by the authors to past research, has been shown to increase serum 25(OH)D sufficiently, and correspond to the thresholds of approximately 700-1000 IU daily oral supplementation that seems to be more effective at fracture reduction than lower-dosed studies using 400 IU daily.

In summary, as much as clinicians might be attracted to once annual vitamin D dosing because of increased patient adherence, the results of this study point out that this may lead to an increased risk of falls and fractures if used in community-dwelling women who are not taking supplemental calcium. Similar improvements in 25(OH)D as would occur with 500,000 IU once annually can be achieved with lower, more frequent dosing, and, as per the medical literature, this appears to not lead to increased falls nor fracture risk. Our recommendations to patients need to be tailored individually based on their demographic, medical history, and baseline nutrition and supplement intake, and, in this respect, the results of this research trial are but one piece of the vitamin D puzzle, highlighting what NOT to do.


1. Kiefer D. Vitamin D supplementation dosage: A road map through the confusion. Alt Med Alert 2010;13: 73-77.

2. Smith H, et al. Effect of annual intramuscular vitamin D on fracture risk in elderly men and women: A population-based, randomized, double-blind, placebo-controlled trial. Rheum 2007;46:1852-1857.