Although alcohol is believed to play a role in college sexual assaults, there is no evidence that male students' binge drinking per se boosts their odds of becoming a perpetrator, according to a study.


On the other hand, researchers found, male students who frequently went to bars or college parties were more likely than others to have ever sexually assaulted a woman over the first 5 semesters of college. What's more, an individual student's likelihood of being a perpetrator increased during semesters in which he attended parties or bars more than his usual amount.


The findings, reported in the January 2017 issue of the Journal of Studies on Alcohol and Drugs, suggest that "drinking setting" -- rather than drinking, per se -- might be key.


"People drawn to these settings may be at higher risk," said lead researcher Maria Testa, Ph.D., of the Research Institute on Addictions at the University at Buffalo.


That is, a student who heads to a bar or party might be more interested in sex than one who is content staying in the dorm and watching television, she said.


However, she pointed out, the study did not ask men about the specifics of any sexual assault -- so it's not clear whether assaults stemmed directly from a night out at a bar or party.


The findings are based on surveys of nearly 1,000 U.S. college men. They were first surveyed as freshmen and then again at the end of each of the next five semesters.


Overall, almost 18 percent admitted to sexually assaulting a woman at some point during the study period. That included forced intercourse, "attempted" intercourse, and "unwanted contact."


The students were also asked about binge drinking, defined as having five or more drinks in a row. At first, it looked like men who binge drank more often were more likely to have sexually assaulted a woman.


However, that link disappeared once the researchers accounted for certain personality factors that the survey measured -- such as problems with self-control, antisocial behavior and "impersonal" attitudes toward sex.


But men who frequently went to bars or parties were more likely to be perpetrators, even when personality traits were taken into account.


According to Testa, the findings could have some practical implications. If college parties and bars around campus do provide the context for some sexual assaults, then making those settings "safer" could be helpful.


Testa pointed to "bystander intervention" programs as an example. Many U.S. colleges have already begun using the programs, which train people in how to recognize and intervene in situations where a potential perpetrator may be zeroing in on a potential victim.


Also, most of the students in the study were underage and, technically, should not have gotten into any bars. If laws were enforced, Testa said, that might help address one context related to

sexual assault.