A recently published Lancet article about vitamin D and bone mineral density is making waves in the medical world and on the web. It calls into question whether vitamin D should be recommended to all community-dwelling adults for osteoporosis prevention.
This meta-analysis sought to investigate the interactions between vitamin D supplementation and bone mineral density (BMD). BMD is a proxy measurement for fracture risk that, while not always accurate, is usually a good indicator of fracture risk. Changing your lifestyle can have an impact on your BMD within the year, whereas to measure risk directly would require much larger sampler sizes and numerous years of tracking.
While bone mineral density (BMD) is not a direct measurement of fracture risk, it is a solid and measurable indicator.
The authors argued that previous meta-analyses on vitamin D and fractures were of low quality due to the available data being of limited statistical power, inappropriate doses, or not in deficient populations.
A new meta-analysis questions whether blanket vitamin D recommendations are suitable for osteoporosis prevention.
Recap
The meta-analysis assessed 19 studies on Vitamin D supplementation which analyzes bone mineral density. The studies invested were studies using vitamin D supplementation of either D2 or D3 form in adults. Age was not limited to older people, and if calcium was taken with Vitamin D then the placebo group must have also received calcium. Thus, some of the studies are just vitamin D, while others are vitamin D + calcium vs placebo + calcium.
The inclusion criteria were a bit vague, and this may have been to broaden the search (and thus include more data). That being said, while permissive inclusion criteria can include more data, they can also add so much data that benefits to one group of people (e.g. the elderly, women, etc.) can be lost because people in other groups are included. For these complications, we look further in the data to when factors are controlled for.
Inclusion criteria are the characteristics of the subjects in the study. Having both young and elderly in this meta-analysis could have influenced the results.
At the outset of the study, the meta-analysis found that at the four areas of the body where bone mineral density was measured, only one (femoral neck) saw a small benefit with supplementation. The other four (hip/trochanter, lumbar, forearm, and total) did not; the confidence intervals (CI values) all crossed the zero point, indicating no statistical benefit.
The authors concluded the study prior to doing any analysis of the cohorts in their data, and as such we cannot use this data to conclude anything about the age groups.
Lets take a closer look
At the outset, a major complication with this study would be how it had a very large and permissive inclusion criteria set to allow many studies to be analyzed yet it did not do further analysis into whether there were commonalities in the studies that showed benefit versus those that don't. The question "did including younger adults cause an apparent benefit in elderly people to be missed" cannot be answered without further statistical analysis. It should be noted that six of the included studies (comprising 871 subjects, about 21% of the total subjects analyzed) had a mean age under 50. This could have influenced the results if we assume the elderly would benefit more from Vitamin D supplementation than youth (which, according to the previous literature, is a very sound hypothesis).
Furthermore, vitamin D levels in serum were quite variable, as at baseline in these studies there seems to be quite a mix of studies with baseline Vitamin D below 30nM (five studies), between 30-50nM (three studies), 50-75nM (eleven studies) and one above 75nM. When looking at Vitamin D concentrations in serum after supplementation, it remains a variable 53-92nM (the range of the averages across the studies).
Penultimately, while perhaps of lesser importance due to standardization in the past suggesting intramuscular and oral supplementation of the following doses are somewhat bioequivalent, there was still a large variability in studies giving daily (thirteen) or weekly/monthly (four) oral supplementation and injections once a year (two studies).
Finally, while the studies in the subject were 92% women, which may simply be something that should be mentioned in the reporting of this study as the actions of vitamin D on bone health are somewhat comparable between men and women (although perhaps a bit more robust in women due to usually more advanced and common osteoporotic symptoms than men).
What should I know?
Overall, there are very limited practical conclusions we can draw from this study.
To start, remember that we need to keep an eye on the prize. While bone mineral density (BMD) is a useful measurement for bone fractures, the ultimate issue we are concerned with is injuries and death from falling, especially in the elderly. Vitamin D supplementation has been shown studies to decrease falls and fractures in the elderly. That's important!
Various studies show that in the elderly, vitamin D supplementation decreases falls and fractures.
The broad inclusion criteria means that in general, we can say that vitamin D does not help improve BMD. As vitamin D is not generally recommended to alleviate osteoporosis in the general population, it would have been beneficial for the analysis to include a breakdown for a variety of age groups.
The bottom line is that we care most about preventing fractures and subsequent death. The people most at-risk for those are the elderly, and several studies show that vitamin D helps prevent such falls. The Lancet study focused on the entire adult population.
When you factor in the prevalence of vitamin D deficiency, the myriad of other health benefits of vitamin D, its low cost, and how safe it is, it is our opinion that vitamin D supplementation is a safe and intelligent choice.