The Latest on COVID-19 and Weight

Every day, the media keeps pumping out content about the purported link between poor COVID-19 outcomes to higher body-mass index (BMI), and yet to date I haven’t seen convincing evidence that this link definitively exists. Different studies have come to very different conclusions, including some showing no link at all—or even a protective effect of high BMI—in certain populations, as we’ll discuss below.

Moreover, even if a COVID-BMI link does genuinely exist, I’ve seen no good evidence to date that higher weight somehow causes poor COVID-19 outcomes. Right now, all we have is (inconsistent) evidence of correlation, and the golden rule in statistics is that correlation is not causation. Researchers have merely been speculating on the ways in which body fat itself might be to blame for these outcomes—and that rush to implicate fatness is a form of weight stigma, which ironically is one of the real contributors to poor health for higher-weight people.

Take, for example, this August 12 study (content warning: weight-stigmatizing language), which followed 6,916 patients in a Southern California healthcare system. Media coverage of this study has reliably included speculation about why fatness supposedly led to higher COVID-19 mortality among higher-weight participants—but a close reading of the science tells a very different story (CW: specific BMI numbers):

First, this study found NO increased mortality risk for people of any age or sex with BMIs between 25 and 40—the “overweight” category and much of the “obese” category. And when the researchers drilled down into the risks for people with BMIs of 40 and above, they found that mortality risk was *only* increased for men—NOT for women with those same BMIs. Women at the very highest end of the weight spectrum weren’t found to be at greater risk of dying from COVID-19 at all. So despite headlines making blanket statements about a link between higher BMI and mortality risk, this study found that risk was only elevated for a very specific subset of larger-bodied people.

Even if these results were airtight (which I’d argue they’re not; more on that below), men with BMIs of 40 and above shouldn’t freak out, or really to do anything different to keep themselves safe from COVID-19 than the rest of the population: Everyone, regardless of body size, simply needs to continue to practice physical distancing and wear masks in public, wash your hands frequently, avoid touching your face, avoid contact with people who are known or suspected to be sick, work from home if possible, etc. 

The reality is that even if weight were a COVID risk factor (and again, the jury is very much still out), we just don’t have a safe, sustainable way for people to lose weight. The overwhelming majority of weight-loss attempts end in weight cycling, which is a known health-risk factor, independent of BMI. 

Weight stigma is another independent health-risk factor, but the August 12 study doesn’t control for the medical weight bias that we know harms people in larger bodies. Having a BMI at the higher end of the spectrum can mean that a person is exposed to more weight stigma, so I think it’s likely that much of the increased risk found in this study for men with BMIs of 40 and above would disappear if the researchers had been able to control for this form of discrimination. 

We see a clear example of how weight stigma plays out in infectious disease in the case of the H1N1 flu virus: During that pandemic, study after study came out blaming BMI, and even today studies (including this one) are still using the example of H1N1 to supposedly show that people in the “obese” BMI category are at higher risk of severe illness and mortality. But as I reported in my April 2020 WIRED piece, a systematic review and meta-analysis of studies on H1N1 and BMI found that there was actually NO increased risk of mortality for people with BMIs of 25 and above, and any associations between high BMI and severe, non-fatal complications disappeared after the researchers adjusted for the fact that smaller-bodied H1N1 patients were more likely to get better-quality treatment. In other words, high BMI had no bearing on mortality risk, and medical weight bias—not weight itself—was responsible for the increased risk of severe but non-fatal H1N1 outcomes among people with BMIs in the “obese” category. 

I think it’s likely that we’ll see the same results for COVID-19, once meta-analyses come along that actually control for quality of care as they should. 

It’s also odd that this California study didn’t find any differences in outcomes among different races/ethnicities, given that we know this disease is disproportionately impacting Black, Latinx, and Indigenous communities (not because of anything to do with their biology, but because of systemic racism resulting in higher rates of COVID-19 exposure and less access to healthcare). The researchers themselves acknowledge that this may be because the healthcare system where the study was conducted (Kaiser) has "more equalized health care access” and therefore doesn’t reflect the racial health disparities we usually see in the U.S. That’s a major limitation of this research when it comes to controlling for the impact of race on COVID-19 mortality (although it’s certainly a good thing for the well-being of the study participants). 

Finally, this study was funded by the pharmaceutical company Roche-Genentech, which likely has an interest in developing weight-loss drugs—Roche was the originator of the weight-loss drug Xenical (orlistat). I recommend taking a critical eye toward studies with this type of funding that find some adverse effect for people with high BMIs, because of course that’s beneficial for the funder.

If fatness itself really were the COVID-19 risk factor it’s made out to be, we’d see that across the board in every study, in every high-BMI category, and in all different populations and age groups. Instead, the research is inconsistent at best.

Another recent study, for example, suggested that high BMI may actually have a significantly protective effect. That study (CW: weight-stigmatizing language), published in BMJ Open on August 11, found that Black and Hispanic ethnicity are associated with significantly higher odds of getting COVID-19, even when controlling for “obesity”—and in fact, the study shows (in Table 4) that Black and Hispanic people in the “obese” BMI category are actually 17 to 18 percent LESS likely to contract the virus.

The researchers don’t speculate as to why that is, but it just goes to show that the scientific data is still very much in flux, and we can’t draw any firm conclusions about larger body size being linked to increased COVID risk. In my view, the push to jump to those conclusions is a product of diet culture and the industries that profit from it, including the diet industry, the bariatric-surgery industry, and the pharmaceutical industry—all of which seem to be behind a lot of the research showing a purported link between COVID-19 and high BMI.

If we really want to help improve COVID-19 outcomes, we need to stop reflexively blaming body fat, and start looking at the real culprits: systemic injustice and discrimination against marginalized communities.