Deep breath; Let's talk about Ethnicity, Vitamin D and COVID19 hospitalisation. TL;DR we don't have enough data/good enough studies to provide conclusive information of whether vitamin D is part of the causal pathway for COVID19 severity
Some baseline. In the UK and US, people from minority ethnicity (labelled "BAME" in UK, and "African-American" or "non-White" in US) are substantially more likely to be in hospital with COVID19. There will be undoubtedly be many reasons for this.
(A side note: would be great to have a French perspective on this but in France recording of ethnicity for health is far more complex and I have not seen studies with this broken out - does anyone know of one?)
Clearly the urbanisation of BAME / African-American is the biggest part of this effect (London and NYC) but it does look like there is more than this.
A potential genetic pathway is the vitamin D pathway; vitamin D production happens via sunlight, and people with lighter skin make more vitamin D in weak sunlight (think - UK winters) and usually overall.
As long time followers know, I am *very* leery of rushing to a "genetic" explanation of "ethnic group" differences - for starters human genetics is way way messier and complex than the simple buckets we are asked to sort ourselves into.
At a deeper level the "bucket" view of human ethnicities is profoundly wrong - its best to think of human genetics as a complex family tree with many joins and splits - this way "hispanics" and "afro-caribbeans" and "black british in liverpool" - mixtures of recent people that>>
<<don't fit the "bucket" view of ethnicity due to recent mixing (and sometimes therefore trigger new buckets to be made in the labelling process) fit absolutely fine in humans as "its a big complicated tree" view. As it happens "Europeans" (the bucket view) are a complex mixture
...but... if there is *one* phenotype where the phrase "white" vs "non white" has an alignment between "what box do you tick on the survey" and "what is your phenotype" it is going to be skin pigmentation.
(NOTE; I go ballistic when people wave their hands around other phenotypes - from height to intelligence - being somehow well captured at the genetic level via this box ticking process. It's so wrong. For other threads).
So - it is pretty well known that vitamin D levels are lower in BAME in northern climates. And studies like this one medrxiv.org/content/10.110… points to some (weak) association of Vitamin D levels and COVID status
*BUT* this is association, and frustratingly for the people with the vitamin D hypothesis the fact that skin-pigmentation <=> vitamin D levels are correlated opens up vitamin D levels <=> COVID assocaitions via other routes (very simply exposure due to cities).
Untangling this is all is a real mess. And because this is one of the few areas where genetics is tangled up with ethnicity labels which itself is tangled up with all sorts of societal processes (straightforwardly - poverty - but lots of other subtle things)
Usually we can use genetics as a sort of sword through the Gordian knot of correlations (I'm a big fan) but even here genetics just because one more complicated rope tangled up with all the others. Irritating.
So, the most responsible thing to say here - which is super annoying - is that it *possibly* might be part of the explanation, given the complexity of things unlikely to be the only thing and we will need to be very sophisticated to untangle it.
