The data for this cohort study was extracted from the larger randomized and double-blinded Vitamin D and Omega-3 Trial (VITAL), which evaluated the effect of supplementation with 2000 IU cholecalciferol (vitamin D3) on a variety of health outcomes. Although meta-analyses of VITAL and other randomized trials did not initially report a benefit in terms of cardiovascular disease or cancer risk reduction, secondary analyses did find evidence of lower cancer incidence/mortality and autoimmune disease when the data was stratified by BMI. For example, a detailed secondary analysis of VITAL found that supplementation at 2000 IU per day reduced the risk of developing advanced (metastatic or fatal) among adults without a cancer diagnosis at baseline by 38%, but only among participants with a BMI < 25. Among adults with a BMI > 25, no benefit was observed. Similarly, a meta-analysis of randomized trials found that vitamin D supplementation given to participants with prediabetes reduced the risk for progression to type 2 diabetes, and also increased the reversion rate to normoglycemia, but only among non-obese participants.
This post hoc cohort analysis in JAMA Network was done to better understand why supplementation appeared to have benefits for participants with a BMI <25, but not those with a BMI greater than 25. Data from over 16,000 VITAL participants were available for analysis, with an approximately equal number of men and women. In addition to measuring total 25-hydroxyvitamin D (25-OHD) levels, multiple other vitamin D-related biomarkers were examined in a subset of these participants, including 25-OHD3, free vitamin D (FVD), bioavailable vitamin D (BioD), vitamin D–binding protein (VDBP), albumin, parathyroid hormone (PTH), and calcium.
As previously observed, 25-OHD levels were lower among those with a BMI > 25, both at baseline and after receiving the same dosage of vitamin D. Additionally, although supplementation significantly increased total 25-OHD, 25-OHD3, FVD, and BioD levels at 2 years compared to placebo, these increases were blunted as BMI increased. No differences in VDBP, albumin, PTH, or calcium levels were observed between placebo and supplemental groups, with no associations found in all BMI categories.
Thus, markers of vitamin D availability and bioactivity appear to be diminished in individuals with overweight and obesity. These trends were observed at baseline as well as after supplementation. BMI modified the relationship between disease incidence and 25-OHD levels as well. For example, even though most individuals with a BMI < 25 had 25-OHD levels greater than 20ng/mL (i.e., they were not deficient), supplementation was still associated with a reduction in cancer incidence in this group. Yet despite having lower 25-OHD levels at baseline, no benefit was observed among individuals with a BMI > 25. The typical lack of any analysis of bioactivity may offer some insight. 25-OHD alone does not necessarily reflect the FVD level and may bias estimates of deficiency.
Although this publication does not clarify mechanisms, the authors do propose two plausible hypotheses. The first is that among people with a BMI > 25, more vitamin D is sequestered into adipose cells, and is removed from circulation. This concept was supported by their findings, as well as previous weight loss trials suggesting that greater loss is associated with an increase in 25-OHD levels, an effect also observed in weight loss due to gastric bypass surgery, without any vitamin D supplementation.
The second hypothesis proposes that adiposity is associated with hepatic dysfunction. The activation of vitamin D requires the enzyme cytochrome P450 (CYP) 2R1, the main vitamin D 25-hydroxylase. Animal and human studies suggest that this enzyme is both downregulated with obesity, and upregulated with weight loss.
There are several important takeaways from this analysis. First, among people with a BMI <25, there is consistent evidence for significant chronic disease risk reduction with vitamin D supplementation at a dose of 2000 IU per day. This includes cancer, autoimmune disease, and diabetes/metabolic risk reduction. This is an important point to make, as initial reports suggesting that supplementation did not have significant benefit may have discouraged practitioners from recommending this effective intervention, yet supplementation has shown a clear benefit in this subset of people. Second, although 25-OHD is generally relied upon as the gold-standard biomarker for vitamin D sufficiency, this study suggests that there may be more to the story, and more in-depth analysis of vitamin D bioactivity may be necessary, particularly among patients with a BMI > 25.
Finally, this study does not provide a clear message of optimal dosing for those patients with an elevated BMI, and more work here is needed. However, we do have evidence that weight-based dosing is sensible here, as suggested by a prospective controlled trial which found that rather than a daily dose of 2000 IU per day, 25-OHD was more effectively increased among participants with obesity when dosed at 125 IU/kg/m2 of vitamin D3. Given that many medications are based on dose, it doesn’t seem that far-fetched to think vitamin D should be as well, and this study helps make that case.