On election night last year, testosterone levels dropped rapidly among male voters of losing parties.

After the outcome of the U.S. presidential election was declared, neuroscientists at Duke University found that although male voters for Barack Obama, the winner, had stable levels of testosterone, the hormone's levels rapidly dropped in males who cast ballots for John McCain or Robert Barr, the losers. In a questionnaire, the McCain and Barr voters reported feeling significantly more controlled, submissive, unhappy and unpleasant after the loss than the Obama backers.

The researchers monitored testosterone levels from the saliva of 163 college-age volunteers in North Carolina and Michigan by asking them to chew sugar-free gum and then spit before and after the results were announced. The male participants would normally have shown a slight nighttime drop in testosterone levels anyway, because the body doesn't need it during sleep, but on election night, they departed dramatically from this routine: Obama voters' levels did not fall as they should have, whereas those of McCain and Barr backers dropped more than would have been expected.

No significant effects were seen in the 106 female volunteers. Women have testosterone, but in much lesser amounts, making them less likely to experience rapid testosterone changes following victory or defeat.

Past research had shown that personally winning and losing in sports matches and other competitions raised and lowered testosterone levels in men. These new findings, appearing online October 21 in PLoS ONE, reveal that politics can influence testosterone in men "just as if they directly engaged head-to-head in a contest for dominance," says researcher Kevin LaBar of the Duke University Center for Cognitive Neuroscience. "I wouldn't be surprised if there was an Obama baby bump nine months after the election."

Anthropologist Coren Apicella at Harvard University, who did not participate in this study, noted she and her colleagues discovered similar results with a smaller group, findings that will appear next year in a book. "It's an exciting time for people that study political behavior, where biological factors have largely been ignored," she notes. "Political scientists are starting to recognize the role of biology and more and more research is showing there may be some reciprocal interactions between how elections make one feel, and how feelings can affect political behavior."

Testosterone is linked to aggression, risk-taking and responses to threats. Bumps and drops in testosterone levels in response to competition in a variety of species can help both winners and losers, explains researcher Steven Stanton of the Duke Center for Cognitive Neuroscience—victors may get motivated to pursue further gains, whereas also-rans are encouraged to back down so as not to press onward and potentially get injured. Apicella noted she and her colleagues found when voters were given $5 that they could donate in toto or in part to political parties or keep for themselves, after the 2008 election, voters for losing candidates whose testosterone levels dropped donated less, "showing this withdrawal behavior."

LaBar and his colleagues conjecture that because the shift in the hierarchy of dominance in the nation following a presidential election is stable for at least four years, the stress of having one's political party lose executive control of policy decisions could plausibly lead to continued testosterone suppression in males. However, "testosterone levels fluctuate every day, so with so many other factors that might influence it, it's hard to know how long-lasting these effects might be," LaBar says.

An open question is whether changes on the level of business or international politics can also drive changes in physiology—for instance, winning or losing wars, or economic booms and busts. Another mystery is whether a similar effect would be seen with local competitions, such as mayoral races, or in balloting where voters are not as emotionally invested, such as in off-year elections.

"We're now going to try and explore this in spectator sports, by looking at Duke and [University of North Carolina] basketball fans," LaBar says. "Sports competitions are not like the political process, where you can have a direct influence on the outcome, but obviously avid sports fans are highly invested personally in the outcome of a game."

Neuroscientist and psychologist Robert Josephs at the University of Texas at Austin, who did not participate in this study, adds, "It would be very useful and productive to assess individual differences in voters, such as need for power. Also, I would love to scan these folks [with fMRI]. Specifically, what brain regions are activated during these changes in testosterone levels? Could we link success/failure to specific neural substrates?"