In the endless news cycle that is COVID-19 there’s an almost ceaseless cry that is repeated over and over again. While the subject changes, the basic idea has remained remarkably similar over the last 14 months: “[x] will protect you against coronavirus!”.
Sadly, it’s almost never true.
The newest case of this hopeful thinking has to do with eyeglasses. Headlines across the globe have, for the last week, been proclaiming that people who wear glasses are less likely to catch COVID-19. The New York Post even said that people who wear glasses are “2–3 times less likely to get COVID-19” than people who don’t which, if true, is a truly remarkable fact. Maybe instead of social distancing and masks we could’ve just all learned from the myopic masterminds in our lives and gotten some frames?
Unfortunately, the truth is a bit less fantastic than the headlines suggest. The science is, at best, incredibly shaky, and probably means almost nothing to your life at all.
People who wear glasses probably aren’t immune to COVID-19.
The basic idea behind these headlines is actually not that outrageous. One of the ways that COVID-19 can spread is by getting into your eyes through droplets or smaller particles, and having a barrier there — even a fairly flimsy one like spectacles — might prevent some of this from happening. As a glasses-wearer myself, the idea is pretty appealing!
So, what does the science actually show?
The study that has been in the news so much is a preprint on a server called Medrxiv, written by a single author. It’s a very simple piece of research — the author conducted a survey of about 300 people who had COVID-19 in a hospital in India, and asked them about their glasses-wearing behaviour. They then compared this group of coronavirus patients to previous estimates of the number of people who wear glasses in India, and found that there were fewer glasses-wearers in the COVID-19 group than they would’ve expected. Thus, they argued, glasses were protective against COVID-19.
There are a few massive issues here that immediately spring out. Firstly, this was a small, selected sample of inpatients at a single hospital in northern India. What we can infer from these people is limited enormously, because they may not be representative of most places in the world.
Worse than that, however, is the comparison group. The author took a 2019 study that estimated that about 40% of people in the entire country of India (population 1.3 billion) had vision problems that required correction to compare to the group of inpatients who wore glasses most or all of the time. Now, firstly, this 2019 study is a systematic review of research covering the entirety of India, and so it’s nonsensical to compare it to a small sample of COVID-19 patients in one hospital in the country. Moreover, the systematic review didn’t even estimate what proportion of people wore glasses all or most of the time.
What this study estimated was the proportion of people who had vision issues (50%), and the percentage of people who had vision issues which weren’t corrected (10%). Subtract the one from the other, and you get ~40% of people who wear glasses at any time. Not only is this a terrible way of estimating the number, it means that the two groups are inherently different and the numbers make no sense at all.
Somehow, the problems don’t stop there. The author also used some odd statistical comparisons*, completely ignored any potential for confounding (for example, the age and gender of the survey population might be an issue!), and because this is a preprint instead of giving vital information left blank spaces saying [table 2 — should be here].
It’s rare that I can say this, but I’m not sure we can take any meaning from this study at all. Maybe fewer people diagnosed with COVID-19 wore glasses in one hospital in India than we would’ve expected based on the population, but we don’t actually know if that’s true and even if it was it tells us almost nothing about glasses and your risk of the disease.
The take-home here is pretty simple — we have no idea whether wearing eyeglasses significantly lowers your risk of contracting COVID-19. It’s plausible, the hypothesis is not crazily bad, and it’s entirely possible that there will eventually be some research showing compellingly that glasses can stop coronavirus infections.
But as of now there is not one shred of good evidence that this is the case.
I’ve covered this issue before on my podcast. An almost identical study came out 6 months ago with almost identically useless findings, and it went viral in a similar way. It’s perhaps not surprising — we all wish that there was an easy solution to the pandemic — but it’s still a bit depressing to see this play out over and over in the same way.
One thing we do know is that barriers are not a bad idea against COVID-19. Protecting your eyes, whether that’s with cool spectacles like mine, or a face shield and goggles like a true warrior, probably isn’t going to be a terrible idea.
But the headlines are so misleading that they are just wrong.
Glasses won’t cut your risk of COVID-19 by 2–3x. They might do nothing at all.
*For the epi nerds — this was technically a case-control study, but the author computed relative risk instead of odds ratios and also reversed the calculations making the end numbers mostly meaningless.