Looking to go back into the office anytime soon? Evidence is starting to mount against it, according to a new report from the Centers for Disease Control and Prevention (CDC).

The data released Friday revealed that, in July, of 248 people, those who worked from home for two weeks before developing symptoms of any sort were less likely to be infected with the coronavirus.

In other words, coronavirus-infected people were more likely to report going into school or offices in the two weeks leading up to illness.

This conclusion was upheld even after accounting for those who work in critical sectors, like health care and education, for example.

“Businesses and employers should promote alternative work site options, such as teleworking, where possible, to reduce exposures to SARS-CoV-2,” reads the CDC report. “Where telework options are not feasible, worker safety measures should continue to be scaled up to reduce possible worksite exposures.”

The report drew from a multi-state case study of 314 symptomatic adults in total; 153 tested positive for the virus while 161 tested negative. Participants answered questions about employment status, telework status, and community exposures, among other questions.

(Not all participants answered the question about telework status, which explains the discrepancy in totals.)

Further, of coronavirus-infected participants, only 35% reported being able to telework full-time or part-time, while 53% of people who tested negative reported teleworking. Meanwhile, 65% of coronavirus-infected people were going into the office or school regularly.

“This investigation provides evidence of the potential health benefits of teleworking associated with the COVID-19 pandemic,” the authors added.

However the study also further confirmed socioeconomic disadvantages related to the virus. Teleworkers were more likely to be White, have a college degree, earn over $75,000, and have health insurance, the agency wrote.

Finally, the percentage of people who reported teleworking in the study is higher than the national estimate, which was 26% during July, per the agency. The authors also noted several limitations, including that the findings “might not be representative of the U.S. population,” due to people who declined to partake.