Did Adult Obesity Spike in the Pandemic? Meh

An appealing narrative is hard to resist. When it’s grounded in facts, it can be a powerful way to inform people. But it can just as easily be a tool for misinformation. All too often, a storyline forms around an anecdote or mere speculation. Such is the case with the one about a spike in obesity prevalence for adults in the pandemic. It generates lots of headlines and a few papers. But hard evidence for it is scarce.

Nonetheless, a new paper this week is spawning fresh headlines about how the “pandemic packed on the pounds,” as Newsday puts it. Yet once again, close look at the evidence for this leaves us yawning.

A Huge Set of Data

Brandon Restrepo used data from the BRFSS to support his contention that adult obesity rates rose with the COVID-19 pandemic in the U.S. He looked closely at data on self-reported height and weight, as well as four obesity related behavioral risk factors. Those were smoking, alcohol, sleep, and exercise. To support his conclusions, he constructed linear regression models in which he claimed to have 3,555,865 data points for self-reported BMI. He used log-transformed values in his model for BMI and all of the risk factors.

Thus he found that “obesity prevalence rates were higher by 1.1 percentage points or 3%” when he compared 2019 data to data from 2020 starting with March 12. He also found that people slept a bit more, drank more alcohol, exercised more, and smoked less.

So he concludes that the first year of the pandemic “worsened the pre-existing epidemic of adult obesity in the U.S.”

Opaque Methods, Modest Results

Interpreting these results is a challenge, because the paper does not provide a very detailed description of the methods or the findings. Diana Thomas, a professor of mathematics known for her research on obesity, tells us:

“Based on the paper’s limited description of study methods, It’s hard to know what to make of these findings. The linear regression model uses log-transformed values for BMI to make inferences about the trend in BMI. This raises serious questions for me about the reliability of the study’s conclusions.”

It’s worth noting here that self-reported BMI is subject to bias. People report that they are taller than they really are and that they weigh less. The magnitude of misreporting can vary with other confounding factors. In the midst of a pandemic lockdown, people might be thinking about their bodies a bit differently.

One other thing to note is that a one percentage point rise in obesity prevalence is not very impressive. Even before the pandemic was a thing, this sort of rise was pretty typical. Between 2016 and 2018, the adult obesity rate went from 39.8 to 42.4 percent. That’s 2.6 percentage points in just two years – comparable to 1.1 percentage point rise that Restrepo reports.

Nothing to See Here, Folks

The bottom line is that the appealing narrative about a spike in adult obesity due to the pandemic still lacks objective evidence to support it. In fact, data from measured height and weight in electronic health records says there was no spike.

While good data do suggest that child obesity went up because of the pandemic, this does not appear to be the case for adults. With respect to adults, all we have evidence to say is that the well-established upward trend likely continued.

Click here for the Restrepo study and here for further perspective.

Self-Portrait,Yawning; painting by Joseph Ducreux / WikiArt

Subscribe by email to follow the accumulating evidence and observations that shape our view of health, obesity, and policy.

April 7, 2022