Male Scientists' Threat to the Integrity of Research

A new study shows that the mere scent of men can skew the results of animal tests.

Robert F. Bukaty/AP

In the field of pain studies, urban legend had it that the presence of experimenters decreased pain behavior in mice and rats, and that more pain would be recorded if experimenters left the room.

"It was sort of lab lore. A thing that people whispered about sometimes in meetings," says Dr. Jeffrey Mogil, chair of the Pain Genetics Lab at McGill University. "After hearing this story on and off for 15 years I finally said, 'Listen, if you think this is true, let's do it properly and see if it is.'"

Stress is known to have an analgesic, or pain-inhibiting, effect. So for the first experiment, Mogil and his team injected lab mice with an inflammatory agent and recorded the pain responses with an experimenter present in the room, and the mice's responses when the experimenter was absent. In subsequent experiments, the experimenter was replaced with t-shirts worn by human males and females, as well as bedding material from the mice's own cages and from different animals including other mice, guinea pigs, rats, cats, and dogs—both castrated and "gonadally intact."

The results were a "huge shock," says Mogil. The researchers' suspicions were confirmed, but only halfway: Experimenters did decrease pain in mice—but only male experimenters.

As noted in their findings, which were published in Nature Methods this week, male experimenters and olfactory stimuli produced no effect and neutralized the effect of male stimuli when presented together. On average, the pain response in the mice decreased a significant 36 percent after the introduction of stimuli from male experimenters and un-castrated male animals, and the effect proved to be "exactly as stressful" as two comparison conditions—a three-minute swim in cold water and 15 minutes of restraint in a tube.

"We weren't surprised that those [comparison conditions] would produce stress because we knew they would, but we were surprised that the simple presence of male olfactory stimuli could produce stresses of an equivalent level," Mogil says. "It's quite impressive. This is a very robust amount of stress to not know about until 2014."

He speculates that since bedding from stranger mice produces the effect, what causes the stress is the thought of imminent territorial aggression from males in general, rather than the threat of predation from human males specifically. "What they're afraid of is strange male mice," he explains. "It's just that other male mammals, including us, smell like male mice."

Mogil's theory might also explain why the presence of female researchers and olfactory stimuli defuse the effect of stress-induced analgesia: If the mice smell a mixed-sexed group, the "strange" male mouse is likely to be within a group or with its family, and much less likely to be aggressive or defending territory. The scientist has received a grant to observe whether the same effect is true for people. He believes that it will be, however, since the stress disappears once the animal convinces itself there's no actual danger, "humans would be able to do that fairly quickly"—and any observable stress response would likely be much smaller and shorter-lived.

"A lot of people may interpret this study as bad news that animal testing doesn't work or can fatally confound [research], but I don't look at it that way," says Mogil. "I think a stronger case can be made that, actually, this is good news for animal research."

Growing concerns about the ability to replicate the results of animal testing and the occurrence of false positives may be assuaged by the premise of this new explanation: That it isn't about one lab being right and another wrong, but that both labs are right given their particular environmental context.

"The biological phenomena under study are interacting with these laboratory environmental factors that we don't know about. That means that the results are true, but are just more complicated than we currently understand," says Mogil. "I've seen how much inertia there is in science, but if we started to pay attention to this sort of thing, then we would understand what factors actually matter, and that would make research going forward stronger."

In this case, while the sex effect on mice does go away after extended exposure to male olfactory stimuli, Mogil understands that it is unreasonable to mandate that a female researcher always be present in the room, or even to ask male researchers to sit in a room with mice or rats for 30 to 45 minutes before testing. Instead, he suggests the simple solution of including information about the researchers' sex in their methods section.

"Now that people know this, they can go back and look over their old data and hopefully be able to explain why, for example, they had an effect when with one student, but when a new student took over the project, the effect went away; or why they weren't able to replicate the lab in their field in another city," Mogil says. "I'm hoping that this will solve some existing mysteries."