> Although results of an observational study, such as this one, need to be interpreted with caution, as done by the authors [1], due to the potential of residual confounding or reverse causality (i.e., vitamin D insufficiency resulting from poor health status at baseline rather than vice versa),
You need a control group with vitamin D deficiency but no other health issues to prove the hypothesis. They are just stating there's a correlation worth looking into.
Almost half of Americans are vitamin D deficient, a lot more if you are Black or Hispanic: https://www.healthline.com/nutrition/vitamin-d-deficiency-sy...
Vitamin D is very cheap, and it takes a lot to overdose: https://www.mayoclinic.org/healthy-lifestyle/nutrition-and-h...
Up until a very short time ago, humans spent most of their time outside, and sunscreen did not exist. Now we spend nearly all our time inside, and many of us wear sunscreen whenever it's really sunny.
And there's a serious correlation between low vitamin D levels and COVID-19 fatalities.
We haven't proved causation yet, and we might never, but given the previous three points, there's every good reason to take vitamin D.
I see this in infosec all the time. People think, for example, "I'm running antivirus software; I'm safe," which they'll use to justify all sorts of risky behavior that could (and sometimes does) lead to a security breach. They don't understand that antivirus software only reduces the risk of a security breach, never eliminating it, and it only works on certain kinds of attack vectors.
The same goes for vitamin D. Sure, lots of people are deficient, and the current evidence-based medical consensus is that people who are deficient should supplement. And maybe there's a weak correlation with COVID-19 infection/mortality risk, but people who aren't experts might misinterpret that and think they should hang out with their buddies in the sunshine or something equally risky while we're in the middle of a pandemic, instead of following the other bits of the current evidence-based medical consensus, which tells us to stay away from one another and to wear masks over our mouths and noses, which we know will greatly reduce virus transmission rates.
Spending time with people outdoors, especially in sunshine, is not risky, and for regions with serious outbreaks is an ideal way of finding compromise between keeping people isolated when indoors and allowing people some much needed social interaction.
I'm from Melbourne Australia which is just coming out of one of the longest/strictest lockdowns in the world, and all the evidence has been that letting people have some social interaction in the outdoors is beneficial to people's mental health and not a contributor to viral spread.
The outbreak was painstakingly contact-traced, transmission patterns modeled in great detail.
The transmissions were indoors; initially in workplaces and homes, then it tore through aged care homes.
Transmission outdoors just wasn't a factor. The government still imposed limits on how much outdoor interaction could happen, and mandated masks after things got bad, but the number of outdoor transmissions didn't change; it was always insignificant. Indoor contact was clearly the driver.
To be fair, the facts here probably don’t contradict your point much; severe restrictions were imposed, and adhered to by our largely compliant population, including significant limits on outdoor contact (exercise with only one person outside one's household for up to one hour per day) and requirements for distancing and mask-wearing.
Incidentally, Melbourne is the only large city in Australia that has cool/dark winters and a high incidence of Vitamin D deficiency, and was the only city to have a major covid outbreak. The other (warmer/sunnier/higher Vit D) cities didn’t limit outdoor contact at all, and had no trouble preventing outbreaks.
As TFA says, it "may" account for it. Or may not. You need a control group with D (but no other) deficiency to upgrade from "may" to "does".