I wouldn’t call that evidence thin biff. The role of vitamin D in tuberculosis was actually pretty well nailed by studies of immigrants in the UK which demonstrate that latent TB infections will reactivate in the sunless climate. The demonstration of the role of Vit D in making defensins seems the likely physiological explanation.
I am wary though of some of the more hyperbolic claims I’ve seen about vitamin D lately. It’s fascinating stuff, sure. But until replicated in RCT rather than epidemiologically I think it’s too early to recommend universal supplementation with a vitamin that does have a toxic syndrome. Many vitamins have appeared similarly astounding in such studies and the benefits failed to pan out. This is likely due to certain epidemiologic biases that tend to show people who actively take care of themselves (take supplements, exercise, etc.) are generally wealthier, and healthier
First, in regard to toxicity, this site (NIH affiliated) states that 2,000 IU is the upper bound for Vitamin D daily intake. It recommends 200 IU for the young adult, and up to 600 IU for the elderly. I have brown skin, and it turns out that that means I’m about 1/6 as productive in producing Vitamin D via sun exposure as a white skinned person, so I doubt that the 200 IU recommendation really applies to me. To clear up my Vitamin D deficiency I consumed 3,000 IU per day for a month (3 1,000 IU tablets) and now am going through a regime of 2,000 IU in winter and 1,000 IU in summer. Note, even with the toxicity caution this is below that threshold.
The second issue is a serious one. People do tend to overreact to medical hypotheses and neglect confounds. That being said, I think this is a case where evolution has something to say; multiple evolutionary genomics papers have suggested that skin color is a trait that has been under very powerful selection of late in the northern hemisphere. One locus, OCA2, is the third longest haplotype in the European genome, and has a noticeable affect on skin color. Another locus, SLC24A5 has recently fixed in Europeans and seems to have been subject to selection on the order of 10% per generation across many populations. It accounts for 1/3 of the skin color difference between Europeans and Africans. Skin color has also been driven by selection in East Asia.
The most parsimonious explanation for these selective events, some of the most powerful sweeps detectable in the human genome, is that there were Vitamin D deficiency issues across Eurasia. I have posited that the rise of agriculture resulted in a concomitant decline in intake of Vitamin D through diet and an increase in the frequency and virulence of epidemics because due to increased population density. In any case, the point is that there is a strong likelihood that Vitamin D had selective implications in the past, and I think that should weigh in as a prior when considering its health implications. The possibility that people with skin color in the range of my own were simply out-bred by their lighter skinned relatives 10,000 years ago across vast swaths of Eurasia suggests a major fitness penalty, and also makes me weigh my priors in assessing the validity of current medical research. That being said, just because the Vitamin D thesis is the best we have right now does not mean it is a definite answer (I have started to suspect that not all the light skin producing alleles were selected because of the light skin, and ergo Vitamin D production), and that uncertainty reduces the weight that one can place upon this avenue of insight.