Women are more accurate than men at detecting signs of sickness in faces

New research suggests that women may possess a heightened ability to detect subtle signs of illness in the faces of others. The study found that female participants were more accurate than male participants at distinguishing between photos of healthy individuals and photos of the same individuals when they were naturally sick. These findings were published in the journal Evolution and Human Behavior.

Humans have a fundamental need to avoid infectious diseases. To help with this, people often rely on nonverbal signals to identify potential health threats in those around them. When a person is ill, their face often changes in specific, observable ways.

This collection of facial signs is known as “lassitude.” The expression typically includes features such as drooping eyelids or red, sleepy eyes. It may also involve pale lips that are slightly parted and drooping corners of the mouth.

Recognizing these cues helps observers avoid contagion. It also signals that the sick individual may need assistance or rest. While humans generally possess the ability to spot these signs, it is not clear if everyone is equally skilled at it.

Scientific models propose that certain groups may have evolved to be more sensitive to these cues. Some individuals might be generally more avoidant of pathogens. The authors of the current study hypothesized that biological sex could be a key factor in this variation.

Previous research has shown that females often display a small but consistent advantage in recognizing emotional expressions. This includes emotions like fear, disgust, and sadness. There is also evidence that women are faster at categorizing faces as sick or healthy. However, prior studies often relied on manipulated images. Some used software to morph faces. Others induced temporary inflammation in participants using injections.

“Previous research has shown sex differences in experimentally manipulated (through endotoxin injection or photo editing) sick face photos. Building on this work, our study asked whether these sex differences also appear when people view naturally occurring signs of illness in faces,” said study author Tiffany Leung, a graduate student in the Social Cognition Lab at the University of Miami.

“We also aimed to strengthen how sick face perception is measured by combining multiple ratings (i.e., safety, healthiness, approachability, alertness, social interest, and positivity) into a single, more reliable assessment, rather than relying on any one judgment alone.”

For their study, the researchers recruited 280 adult participants. The sample was drawn from a university in the United States. It included an even split of 140 males and 140 females.

The participants viewed a series of 24 facial photographs. These images came from 12 different donors. Each donor provided one photo taken when they were healthy and one taken when they were sick.

The sick photos captured the donors when they were suffering from a naturally occurring contagious illness. Examples of these illnesses included influenza and COVID-19. The donors maintained a neutral facial expression in both the sick and healthy states.

Participants did not just categorize the faces as sick or healthy. Instead, they rated the images on six different dimensions using a 9-point scale. This allowed the researchers to capture subtle perceptions that a simple categorization might miss.

The first dimension was safety. Participants rated how safe they would feel around the person in the photo. The second dimension was healthiness, asking how sick or healthy the person appeared.

The third dimension assessed approachability. Participants rated their urge to approach or avoid the individual. The fourth dimension focused on alertness, measuring how tired or awake the person looked.

The fifth dimension was positivity. Participants gauged whether the person seemed to be feeling positive or negative. The final dimension was social interest. This asked how interested the person seemed in interacting with others.

The researchers calculated a difference score for each participant. This score represented the gap between the ratings for the healthy faces and the ratings for the sick faces. A larger difference indicated greater accuracy in distinguishing the two states.

The analysis showed that all six dimensions were related. They all contributed to a single underlying ability which the researchers termed “latent lassitude perception.” This suggests that judgments about safety, alertness, and healthiness all tap into a unified perception of sickness.

The study found that biological sex was a significant predictor of this ability. Female participants demonstrated higher accuracy than male participants. This female advantage was consistent with the researchers’ hypothesis.

“Sensitivity to facial signs of sickness varies across individuals, with women generally showing greater sensitivity to lassitude expressions than men,” Leung told PsyPost.

The results align with the idea that women are generally more attuned to nonverbal emotional cues. The findings specifically support the notion of an advantage in detecting natural lassitude. This effect persists even when the signs of sickness are subtle and unforced.

The effect size observed in the study was small to moderate. This means that while women are generally better at this task on average, the difference is not overwhelming.

“In other words, our effects do not necessarily suggest that all women are highly sensitive to sick faces or that all men are insensitive,” Leung said. “In everyday life, these effects would be unlikely to stand out in a single interaction, but over many social encounters they could meaningfully influence how people notice and respond to signs of illness in others.”

Small differences can have meaningful consequences over time. In everyday life, a slightly better ability to spot sickness could influence many social interactions. It could lead to more effective disease avoidance or faster caregiving responses.

But what might be the evolutionary roots of the differences observed between men and women? One theory is the Primary Caretaker Hypothesis. It suggests that women have historically been the primary caregivers for infants.

This role would create evolutionary pressure to recognize nonverbal signals. Mothers needed to identify signs of distress or illness in preverbal infants to ensure their survival. This heightened sensitivity might extend to recognizing sickness in adults as well.

Another possibility is the Contaminant Avoidance Hypothesis. Females often experience higher levels of disgust and are more cautious regarding contaminants. This may be due to the need to protect themselves and offspring during periods of immune suppression, such as pregnancy.

There are some limitations to this research. The study relied on static photographs. In the real world, sickness is also conveyed through movement and sound.

For instance, a person’s posture or the sound of their voice might provide additional clues. It is possible that sex differences also exist in the perception of these dynamic signals. Future research could investigate whether women are better at detecting sickness from vocal cues or body language.

Another limitation is that the study focused on adult participants. It is not yet known when these sex differences emerge during development. Future studies could examine children to see if girls display this advantage early in life.

The researchers also noted that the these effects are likely influenced by individual experiences and contexts. Cultural factors could also play a role in how sickness is perceived and evaluated.

“We have begun developing approaches to support adults’ ability to accurately recognize and avoid signs of illness in faces,” Leung said. “Our next step is to extend this work to children (stage 1 registered report here).”

The study, “Individual differences in sick face sensitivity: females are more sensitive to lassitude facial expressions than males,” was authored by Tiffany S. Leung and Elizabeth A. Simpson.