Reply to Vitamin D Supplements: Is Bone Loss by HR-pQCT Really Negative?

The author questions whether the dose-response effect of vitamin D supplementation we showed using HR-pQCT should be of concern to clinicians, particularly because those findings were not reflected in our DXA-based area BMD (aBMD) measures, and that estimated strength changes by HR-pQCT and finite element (FE) analysis were not statistically significant. The correspondent paraphrases the conclusion of our study as "higher levels could be dangerous for use to prevent bone loss," although this exaggerates what is reported in our manuscript, which is that "our findings do not support a benefit of high-dose vitamin D supplementation for bone health, and raise the possibility of harm for females." We purposely did not use words such as "dangerous" to avoid overstating our findings. Although clinicians almost never prescribe high doses of vitamin D for the prevention of osteoporosis, there is growing evidence of individuals taking ≥4000 IU daily.(1) This number may increase, given the observation of associations between low serum 25-hydroxyvitamin D and coronavirus disease 2019 (COVID-19) and suggestions that this warrants increased supplement use.(2) Our study was completed in 2017, prior to the COVID-19 pandemic, and was not designed to assess the effects of vitamin D on nonskeletal health outcomes. Weighing the benefits of high-dose vitamin D for prevention and/or treatment of reduced BMD is an open question. Although some sources have recommended that the vitamin D requirements for optimal immune system function would need higher-dose supplementation than the standard National Academy of Medicine recommendations for prevention of the bone consequences of D deficiency, that remains to be demonstrated in clinical trials.(3) By design, our study does not include an explicit placebo control group; however, the 400 IU daily group serves as a basis for comparison, receiving the National Academy of Medicine–recommended intake for adults younger than age 70 years (accounting for dietary input). It should also be noted that the 400 IU daily group's serum 25-OH Vitamin D levels did not change from baseline throughout the 3 years of study, suggesting that arm of the study is effectively a placebo group of healthy individuals. Furthermore, in our previous population-based study(4) the dietary intake vitamin D in adults in the same age range was similar to our 400 IU daily group. In reference to our previous population-based data,(4) the 3-year age-related decline in HR-pQCT–measured BMD in the 400-IU and 4000-IU study groups is comparable to what was observed in those participants. It is not appropriate, as stated by the correspondent, to comment whether the 400-IU group was protected from bone loss, just as it would have been inappropriate for us to comment that the 4000-IU group was worse. We re-emphasize our conclusion in the article for this exact reason—our findings do not support a benefit of high-dose vitamin D, and raise the possibility of harm for females. Our data clearly showed a dose-response effect in females, with increased bone loss associated with higher doses. We are not suggesting that all future trials of high-dose vitamin D require assessment of skeletal outcomes using HR-pQCT, but it will be important for potential study participants to be aware of our findings when making an informed decision of whether to participate in such studies. The use of three-dimensional (3D) image registration with HR-pQCT data(5) leads to excellent reproducibility compared to traditional technologies such as DXA. The fact that the loss of BMD by HR-pQCT, particularly at 10,000 IU daily, was not mirrored by femoral DXA aBMD may simply be due to the lesser sensitivity of DXA hip measurements. Or, perhaps less likely, there are skeletal site-specific differences. It is unusual that our BMD findings were not paralleled in our strength estimates, but most likely this is a methodological problem rather than a biological phenomenon. We suspect the reason our FE strength estimates were not significant for pooled male and female (radius: p = .058; tibia: p = .123)(6) is because we could not employ 3D image registration for FE modeling (there was no available method at the time of analysis). The situation is worse for the sex-stratified data (radius: p = .362; tibia: p = .440), likely because of the confounding effects of being underpowered in our secondary analysis. The problem is that serial measures of a 1-cm slab of bone by HR-pQCT strongly warrant the use of alignment strategies such as registration, otherwise reproducibility is greatly affected. We have recently developed methods that can be applied to FE(7) and look forward to examining our data with improved precision. We are skeptical of the suggestion that there was fundamental adaptation to the bone structure resulting in an undetected increase in cross-sectional area resulting from vitamin D supplementation, but our re-analysis of estimated bone strength will address that question. Overall, our data clearly show that as the supplemented dose increases, so does the decrease in BMD at the radius and tibia in females. Whether there is "harm" at supplement levels of 400 and 4000 IU daily is doubtful, but there is also likely no benefit. However, at 10,000 IU daily, it raises the possibility of harm. The authors have no conflicts to declare related to this study or response letter. Funding for the Calgary Vitamin D Study was provided by Pure North S'Energy Foundation in response to an investigator-initiated research grant proposal. Authors' roles: All authors participated in the writing of this response letter.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.