Does COVID-19 care what you weigh?

CW: discussion of body weight and fat shaming in news and medical writing.

Even in the midst of a global pandemic, some folks manage to carve time out of their busy schedules for fat-shaming, patient-blaming and promoting all-purpose weight hysteria. On April 16, the NY Times wrote this story:

Okay, but what are those “studies”, and what do they “show”?

Here’s the lede, which is rather unpromising:

The research is preliminary, and not peer reviewed, but it buttresses anecdotal reports from doctors who say they have been struck by how many seriously ill younger patients of theirs with obesity are otherwise healthy.

No one knows why obesity makes Covid-19 worse, but hypotheses abound.

It’s worth noting a few key points here: There are no studies. There are preliminary reports, based on gathering some information about some patients in some places. These reports sound similar to anecdotes from medical workers about particular patients who were 1) young; 2) severely ill with COVID-10; and 3) had BMIs>30.

I looked at the report the NYT was going on about. It’s here, and is accompanied by a serious disclaimer:

Good to know. But, just out of curiosity, what’s in there?

The upshot is this: the biggest relative risk (by a factor of 10) for being hospitalized for COVID-19 is age (75 and over for huge relative risk, 65-74 for smaller risk). Smaller relative risk factors were BMI>40 and heart failure. BUT: they didn’t control for race, socioeconomic status or quality of care (according to this article in Wired, and also according to me from looking at the original article).

We do know that it’s not always true that people with higher body weights automatically have higher risks of complications or death when they’re hospitalized for respiratory diseases. Here’s what one study on pneumonia found:

The cohort [of veterans who were pneumonia patients] comprised of 18,746 subjects. Three percent [had BMI <18.5], 30% [had BMI 18.5–25], 35% [had BMI 25–30], 26% [had BMI 30–40], and 4% [had BMI>40]. In the regression models, after adjusting for potential confounders, [BMI>40] was not associated with mortality (odds ratio 0.96, 95% confidence interval 0.72-1.28), but BMI 30-40 was associated with decreased mortality (0.86, 95% 0.74-0.99). Neither [BMI 30-40] nor [BMI>40] were associated with ICU admission, use of mechanical ventilation or vasopressor utilization. BMI <18.5 patients had increased 90-day mortality (1.40, 1.14-1.73).

The only group with increased mortality risk was the BMI<18.5 group. The other BMI groups either had a lowered risk or a non-increased risk.

Of course, this is only one study, but there are a lot of studies that fail to show a connection between higher body weights and risks of complications and death during hospitalization for some respiratory illness.

Other investigations are at ongoing and at various stages of revision and peer-review. This is important, as one thing we know for sure is that trying to tease out the influence of one feature of patients on particular health outcomes is very very hard. Small sample size, lack of representativeness, potential confounders and methodological flaws all get in the way of reliable results. Wired gives a good and detailed analysis of ways that some claims about the relationship between BMI and COVID-19-related health outcomes are unwarranted. And they offer a possible explanation:

The fact that researchers have been pointing to body size as a risk factor for weeks now, even in the absence of much evidence, is a clear example of how weight stigma gets enacted in science.

I don’t work in medicine, but I do know that there is a humongous evidence gap between what’s happening clinically in a particular hospital and its patients (each with their own complex medical and other histories), and what is true about everyone with higher BMIs in the US (not to mention other countries) with respect to risks related to COVID-19. Right now we can’t say much of anything. So maybe we shouldn’t. Which means the answer to my blog title question is, “we don’t have evidence right now to answer this question”. It doesn’t make for exciting news copy, but it’s the closest thing to the truth right now.