University Research: Gyms Are Low Risk for COVID Transmission

    Findings from the University of Oregon’s Consulting Group (OCG) demonstrates that health clubs do not pose the same risk as other venues like bars and restaurants. An analysis of Colorado data shows no statistically significant link between fitness club attendance and COVID-19 cases.

    In response to the second surge of COVID-19 cases, multiple states and jurisdictions have moved to reclose businesses to mitigate further spread. Despite existing data attesting to the lower risk profile of health clubs, fitness establishments have been among the businesses mandated to close or operate in a more-reduced capacity.

    IHRSA recently commissioned the University of Oregon’s Consulting Group (OCG) to study the association between health club attendance and COVID-19 cases using Colorado data as a proxy. Colorado’s Department of Public Health & Environment (CDPHE) provides a comprehensive online database of COVID-19 cases and transmission settings.

    “Opposing views from the media and uncertainty in the scientific community create ambiguity for gym-goers, club operators, and legislators about the safety of gyms during the COVID-19 pandemic,” said Megan Evans, senior manager of OCG. “What can we learn from the data?”

    The Results

    In short, the findings are clear: health clubs do not pose the same risk as other venues like bars and restaurants.

    To determine this outcome, University of Oregon researchers examined the relationship between gym attendance and Colorado COVID-19 case data using observational and statistical analytical methods.

    “If we look at Colorado’s record of self-reported outbreak—events in which two or more people have contracted COVID—gyms haven’t made the list up to this point, but bars and restaurants certainly have,” said Callum Kuo, OCG president. “What if we investigate further and compare what we know about gyms in Colorado?”

    By examining the correlation between weekly gym attendance data with the following week of positive COVID-19 rates, researchers found a non-statistically significant correlation between COVID-19 case rates and gym attendance.

    Industry news OCG Research Gym Attendance vs COVID Cases line graph 12 14 column

    Gym attendance vs. COVID-19 cases in Colorado from illness onset (7-day adjusted). SOURCE: OCG

    Not coincidentally, the most recent mandated closures announced by the CDPHE exempted health and fitness centers.

    "We have been talking to stakeholders, and we do not feel this is one of the higher risk settings because people are wearing masks at the gym. Gyms are ensuring social distancing and also cleaning their equipment. So we are comfortable allowing it," said Jill Ryan, executive director of the CDPHE.

    Industry news OCG Research COVID 19 Gym Foot Traffic scatterplot 12 14 column

    Correlation between gym attendance and COVID-19 positivity rates (7-day adjusted). SOURCE: OCG

    Additional analysis by the university group found a negative correlation between health club visits and COVID-19 positivity rates. A negative correlation typically implies an inverse relationship. However, we cannot assert that health club attendance is driving down positivity rates. Instead, a negative correlation should facilitate a discussion on variables that may explain this negative correlation.

    Safety protocols, capacity restrictions, adequate ventilation, and other precautions health clubs have implemented may be impacting a negative correlation between attendance and COVID-19 positivity rates.

    The Key Takeaway

    Based on Colorado data as a proxy, there is no correlation between health club visits and COVID-19 cases. Relative to other public spaces like restaurants and bars, health and fitness facilities are lower risk environments. However, it is possible that safety protocols and measures health and fitness clubs have implemented may be affecting this result.

    “There is no correlation between health club visits and COVID-19 cases.”

    “We need to be reasonable and critical when it comes to evaluating relative risk. And based on what we see in Colorado, gyms do not belong in the same risk category as bars and restaurants,” explained Kuo. “We believe states need to examine their data more closely before hastily closing down gyms and instead making decisions based on what they are seeing in their respective states.”

    The OCG analysis confirms what we already know: health clubs aren’t driving COVID-19 transmission. That latest study adds to the body of knowledge club operators have access to and lawmakers need to consider when mandating business closures. More on the University of Oregon’s findings will be available on OCG’s LinkedIn page.

    The Methodology

    OCG started by examining observational data available on the website of Colorado’s Department of Public Health. The additional statistical methods deployed in the study can be summarized in three steps:

    1. Colorado fitness club attendance data was provided by ABC Fitness Solutions, MindBody, and 24 Hour Fitness. Attendance data amounted to nearly 8.5 million fitness club and studio visits.
    2. Fitness center attendance was examined from March 13 through October 15 along with COVID-19 case data during the same time period.
    3. COVID-19 case data was selected to reflect onset of illness rather than reporting date, then staggered one week (7 days) to account for transmission window based on Centers for Disease Control and Prevention guidance that symptoms appear 2-14 days after contracting the virus.

    “We have been talking to stakeholders, and we do not feel this is one of the higher risk settings because people are wearing masks at the gym. Gyms are ensuring social distancing and also cleaning their equipment. So we are comfortable allowing it.”

    Jill Ryan, Executive Director

    Colorado Department of Public Health & Environment

    Additional Research Needs & Data Caveats

    Like most studies, there are limitations to findings that warrant further research and discussion.

    First, the data analyzed represents a single state. Ideally, if comprehensive COVID-19 case data were as readily available for all states, analysis could be conducted on a nationwide basis.

    Second, the sample size consists of 32 weeks, and with the recent surges in COVID cases, further robust testing is needed to disqualify outlier data.

    Third, positive COVID-19 cases are not available prior to March 12 when health club attendance was at a higher level prior to the pandemic in the U.S.

    Lastly, additional analysis is needed to account for the variables that are omitted from the analysis; the qualitative factors of gyms are unaccounted for, which may, if not likely, contribute to a lower risk environment.

    Author avatar

    Melissa Rodriguez

    Melissa Rodriguez is a Market Research Advisor for IHRSA. When she's not analyzing data and statistics, Melissa enjoys spending time with family, watching superhero series, poring over NBA and NFL box scores, and reading a good book.