Decoding animals' facial expressions may allow researchers and veterinarians to monitor spontaneous pain over long timescales. This may also aid the discovery of painkillers, because this type of pain is similar to that experienced by humans.
Researchers typically detect pain in mice by eliciting specific reactions. Poking the hind paw, for example, causes a mouse to reflexively withdraw the paw; heating the tail makes it flick. But scientists are not agreed on how to measure unprovoked pain.
To analyse facial expressions in mice, geneticist Jeffrey Mogil at McGill University in Montreal, Canada, and his colleagues have adapted a coding system used to measure pain in infants. The work is published today in Nature Methods1.
Mogil teamed up with Kenneth Craig, a psychologist who studies human pain at the University of British Columbia in Vancouver. Expert expression-spotters from Craig's lab compared video frames of mice filmed for up to 30 minutes before and after receiving a painful injection of acetic acid.
The researchers detected five signs indicative of pain in mice. Three are similar to human responses: the eyes close and the area around them tightens, and the nose and cheeks bulge. Mice also pull back their ears and move their whiskers.
"This is the first study that has examined facial expressions of pain in non-human animals," says Craig.
The mouse grimace scale (MGS) was able to detect pain in experiments typically used by scientists. Grimaces were most pronounced for pain that lasted for a matter of minutes or hours, and for discomfort in joints and internal organs. Superficial harm such as immersing the tail in hot water evoked fewer grimaces. Mice and humans show similar variability in pain response, Mogil says.
Pained expressions differed from those associated with stress and illness, the team found. In addition, more harmful stimuli drew more pronounced grimaces, and pain relievers diminished them.
Mice with a mutation that has been linked to migraines in humans showed pained expressions that subsided when they were given an anti-migraine drug. "I'm very confident that we're measuring pain here," Mogil says.
Lars Arendt-Nielsen, a pain expert at Aalborg University in Denmark, is not convinced that the team has ruled out other emotions, such as fear, but he recognizes the benefits of the approach. "You can probably look into new pain conditions which we have not been capable of looking into with previous methods," he says.
Arendt-Nielsen also questions whether the scale will apply to other animals, because pain reactions differ between species. But Mogil thinks it will. "I'd be incredibly surprised if a version of this scale doesn't work in every mammal," he says.
Next, Mogil would like to compare his method with others used to evaluate painkillers, and test different doses used to treat postoperative discomfort in animals. Currently, it's not known whether prescribed doses really work. "We think in a sense that we can rewrite the veterinarian rule book," he says.