Faced with a crippling depression, Olympic swimmer Allison Schmitt would wake every morning wanting nothing more than to go back to bed. “I was failing every time I dove into the pool,” she told Today in August. A years-long battle was exacerbated by a family suicide in 2015, and it took a concentrated effort by Schmitt and the people close to get her out of the hole. In Rio, she won gold and silver medals in the 4x200m and 4x100m freestyle relays, respectively.

Athletes, like everyone else, suffer from mental-health issues—ailments generally far more difficult to assess than a pulled muscle or broken bone.  Unlike everyone else, however, athletes perform in controlled, quantified environments. A person who isn’t in training doesn’t always have crystal-clear markers for how an anxiety disorder impacts their life, but an athlete faces cold numbers every time they step on the field: distances run, assists made, goals scored, games won.

Because of this, sports may offer unique ways to measure mental health. As brain-imaging technologies and advances in neuroscience help link the state of the mind to the state of the body, the ability to objectively trace mental health’s influence on athletic performance could provide tools for investigations into how mental health more broadly affects everyone.

Daniel Eisenberg, an associate professor at the University of Michigan's School of Public Health, is taking the first steps toward this end. He heads up a team of researchers at Athletes Connected, a multi-departmental initiative that helps athletes deal with mental-health problems and raises awareness about mental-health issues on campus. By tracking objective measurements of athletic and academic performance as they compare to athletes’ mental states, he believes that the connections between mental health and performance may be able to be more concretely studied.

Eisenberg recently began modeling this approach by collecting self-reported surveys and statistics from Michigan athletes. During the university’s previous semester, Eisenberg and his team asked 43 Wolverines to complete weekly mental-health surveys, which focused on the athletes’ levels of anxiety and depression and their assessments of their academic and athletic performances. Questions about injuries, sleep, and stressors helped to control for variables surrounding mental health, because these external factors are known to inhibit performance. The athletes provided roughly 530 data points, which Eisenberg and his team now are analyzing.

But self-reported surveys have their limits. “People tend to over-report their own abilities … and their own performance,” Eisenberg says. “There's systematic biases, and there's also probably random biases, or noise, around subjective measures.” What he’s really after is hard data: the squat maxes, 100m dash times, assist/turnover ratios, and putt lengths that demonstrate an athlete’s fitness and focus. Combined with survey information, Eisenberg envisions the data providing numbers to go with self-reports, allowing the connections between the two to be more easily understood.

Eisenberg and his team are still in the process of gathering the performance points to marry to their survey data. They have acquired academic information from the registrar, and they’re working on tracking down sports stats from the university’s immense and highly competitive athletic department.

As one example of what the team has in mind, consider the bounty of data available for a stat-centric sport like baseball. There are the obvious statistics, like a player’s batting average or the number of runs a pitcher allows, but also Sabermetric formulas complex enough to demonstrate how valuable a player is over a replacement (VORP) and even how well that player performs in “late-inning pressure situations” (LIPS)—a measurement of clutch.

The ultimate benefit of these more objective measurements may not only be a more scientifically precise understanding of mental health’s impacts on people, but also the ways societies handle mental-health issues. For instance, Eisenberg's study could have implications for how mental wellness is assessed in insurance structures that reward doctors for keeping patients healthy and avoiding expensive medical care, like Blue Cross of Massachusetts's Alternative Quality Contract (AQC) system.

Moreover, objective measures could change how depression, anxiety, and other mental-health issues are perceived. “I think sports and celebrity are two places where we can begin to lift the mental-health stigma, by showing that real people who perform, and who are well valued by society through their athletic contributions, do also suffer from symptoms of ill mental health,” says Chris Gibbons, a post-doctoral fellow and the director of health assessment and innovation at the University of Cambridge's Psychometrics Centre.

If the evidence is strong, however, it could cut both ways. “I suppose one of the great risks of [Eisenberg’s] study is that it actually increases mental-health stigma by showing that people who have mental-health issues perform much worse in sporting situations,” Gibbons notes. “So I think that the results will have to be handled very carefully, if that is the finding. There are already people who are very worried about revealing their mental-health status to their employers because they assume that will stymie their career and make it difficult for them to progress.”

Eisenberg’s work, at least, seems to have had a net-positive impact on people’s attitudes toward mental health in Ann Arbor. Adam Kern, a research assistant to Eisenberg, relayed the results of an after-study focus group, in which athletes found the surveys useful in their own performances, though felt that coaches needed better instruction on how to provide support. Most hearteningly, they reported that the atmosphere around mental health at UM, and especially among athletes, was open and de-stigmatized, something that Kern, a former Wolverine himself, says wasn’t the case as little as four years ago.