Future Tense

That One Study Does Not Mean It’s Safe to Go to the Gym

A woman lifts weights in a gym.
Maybe gyms should stay empty a little longer. Danielle Cerullo/Unsplash

This article is part of Viral Studies, a Slate series in which we break down recent viral articles and—most importantly—their caveats.

Even as Florida and Texas are taking steps toward reclosing, other states are continuing to open back up, and people are itching to return to their pre-COVID-19 routines. Many states have recently reopened gyms, but the question remains: Are they safe?

According to a recent preprint paper (that is, a paper that has not yet gone through peer review) discussed in the New York Times , it’s safe in Norway, where the study was conducted —and its authors say the results are generalizable to other parts of the world, with the caveat that “there may be places where there is a lot of COVID, or where people are less inclined to follow restrictions.” But that’s a big caveat—especially if you’re trying to generalize these results to the U.S.—and in looking at the study’s results, it’s unclear whether they reflect the transmission risk at gyms, or just the general, relatively low transmission risk in Oslo.

In the preprint study, which was commissioned by the Norwegian government, more than 3,700 Norwegians at five gyms in Oslo were randomly assigned to either work out at the gym or stay at home. After two weeks, researchers tested the participants again but lost around 20 percent of the original participants in the process. Of the roughly 3,000 people tested after two weeks, the researchers found no difference between the gym-going group and the staying-home group. But it’s noteworthy that only one person in the entire study tested positive for COVID-19. They happened to be in the gym-going group, but the authors note that “the individual had been present at the workplace where two other individuals had tested positive for SARS-CoV-2 shortly before the participant tested positive in the trial. The transmission was most likely unlikely related to the trial intervention.”

That one positive test result is not enough to draw conclusions, says Sandra Tilmon, an epidemiologist at the University of Chicago. She points to the overall low prevalence of COVID-19 in Norway; the preprint authors write that over the two-week period of their study, there were just 105 confirmed COVID cases in Oslo. Meanwhile, in the U.S., Florida has just announced its highest-ever single day total of nearly 9,000 cases. The background prevalence of COVID-19 is just generally low in Norway, which means these results could just be showing that it’s generally safe to be out in public in Norway. “If you don’t have community transmission, you’re not going to have transmission in the gym,” says Tilmon. “There’s no way you can leap from one transmission to ‘gyms are safe.’ If we got to low background transmission, then everything would be safe.”

Nicole Carnegie, a statistician at Montana State University, also says she doesn’t believe the researchers’ conclusions are supported by their results. “There isn’t enough information to make a call about the difference in risk between the two arms in the study,” she says.
“Really what this says is ‘we don’t know’; the rate was so low inside and outside the gym that we can’t measure [a difference].”

These different rates of community transmission mean that these results are not generalizable to other cities, says Tilman. To truly understand whether this study actually has anything to do with gym behavior, this study would need to be replicated in cities with varying COVID-19 prevalence—but that could be unethical, since gymgoers in areas with higher COVID-19 rates would be more likely to get sick.

While this study concludes that the hand hygiene and 3 to 6 feet of distance the researchers mandated will keep people safe, that bears further investigation, too. Tilmon points out that being in an experiment can change people’s behavior, often for the better—and future studies could look at how people in the real world behave in gyms when they don’t think anyone’s watching (or testing them later). Carnegie points out that there may just be lower overall compliance with distancing rules in U.S. gyms. And again, with an overall low COVID-19 rate, hygiene may not have made a huge difference at all, and other factors—like air flow at these facilities—could play a role.

But there is one lesson the U.S. could take from the study. For the research, these scientists tested a sizable group of people who seemed otherwise healthy and reported no symptoms. That’s similar what many U.S. epidemiologists and public health specialists advise we do more of. Currently, much of the info about the U.S.’s COVID-19 rate comes from people who are actively seeking out tests, which means those numbers represent primarily people who are potentially sick, or have had an exposure that makes them concerned enough to get checked. Random community sampling, on the other hand, can give us a better glimpse into the transmission rates in the community in general; it will include cases in folks who may be asymptomatic or mistook mild symptoms for a cold or allergies.

The researchers write that they had expected that 1 percent of their participants in each group to come down with COVID, based on Norwegian data on COVID-19 rates. That would’ve been roughly 30 people, but only one tested positive; that could be a sign that community spread is low. More evidence from community testing would be necessary to draw any conclusions. The community sample taken could also account for the low rate. “People who want to go back to their gym or at least have a gym membership are probably a bit healthier or lower risk than the general population,” says Carnegie.

But with massive spikes in cases and hospitalizations in many states, that type of good news does not extend to much of the U.S., where many experts believe community transmission might be significantly higher than revealed through our still-limited testing. “I see people grasping onto … any messaging that says we can go back to normal,” says Tilmon. “But we’re nowhere near normal.”

Future Tense is a partnership of Slate, New America, and Arizona State University that examines emerging technologies, public policy, and society.