A new analysis challenges the widely believed notion that moderate alcohol consumption is healthier than abstaining completely. According to researchers, previous studies have flaws that exaggerated the health benefits of drinking.

The Canadian Institute for Substance Use Research’s Tim Stockwell and his team at the University of Victoria, examined 107 studies on drinking habits and longevity, the Guardian reports. They found that many studies compared drinkers with people who didn’t drink or consumed very little, without considering that some had reduced or stopped drinking due to health issues.

This oversight means that the abstainers group included many sick individuals, skewing the results and making moderate drinkers appear healthier by comparison.

“It’s been a propaganda coup for the alcohol industry to propose that moderate use of their product lengthens people’s lives,” Dr. Stockwell told the Guardian. “The idea has impacted national drinking guidelines, estimates of alcohol’s burden of disease worldwide, and has impeded effective policymaking on alcohol and public health.”

Published in the Journal of Studies on Alcohol and Drugs, the research suggests that previous studies showing a J-curve effect — where death rates are lowest among light to moderate drinkers — may be misleading. When the Canadian team re-analyzed the data, they found the supposed benefits disappeared in high-quality studies that excluded former drinkers and occasional drinkers from the abstainers group.

“Estimates of the health benefits from alcohol have been exaggerated while its harms have been underestimated in most previous studies,” Dr. Stockwell said. “Many studies included people who had stopped or cut down their drinking for health reasons, making those well enough to continue drinking appear healthier.”