Identifying sick people while sick yourself: a study of identification of facial cues and walking patterns of sick individuals during experimental endotoxemia.

Sick humans and other animals often withdraw from social interactions. It has been suggested that social withdrawal might enable avoidance of contagious individuals, but experimental evidence is lacking on how the state of sickness may affect perception of sick others. Here, we investigated if individuals were more likely to rate others as sick, while being sick themselves, compared to when healthy. Furthermore, we assessed whether the intensity of the fever response and sickness behavior would predict changes in sickness detection. Thirty-four participants were experimentally made sick using an intravenous injection of the bacterial endotoxin lipopolysaccharide (LPS condition; dose of 1.0 ng/kg body weight) and completed a sickness detection task during the peak of the inflammatory and sickness response. Participants performed the same task when they were healthy (control condition, n = 32), in a randomized order before or after the main study day. In the sickness detection task, participants watched photos of individuals' faces as well as video recordings of the same individuals walking, and rated the individual on each stimulus as sick or healthy. The photos and video recordings were obtained from twenty-two individuals who had participated in a previous study, and had been made sick with an intravenous injection of lipopolysaccharide (2.0 ng/kg body weight) on one occasion, and remained healthy after an intravenous injection of a placebo (0.9 % NaCl) on another occasion. Participants could detect sick individuals based on photos and walking patterns above chance level during both the LPS and the control condition. There was no significant difference in how often participants identified sick faces and sick walkers in the LPS condition - when they were sick themselves - compared to in the control condition. However, healthy walkers (but not healthy faces) were more often rated as sick by participants in the LPS condition compared to the control condition. Neither the fever response nor the intensity of sickness behavior predicted changes in sickness detection. The results do not indicate more accurate sickness detection in others during own sickness. Nevertheless, the data from walking patterns indicate that sick individuals may be more prone to categorize healthy individuals as sick. If replicated, this could in speculation be related to a need to reduce the risk of becoming infected while already fighting a pathogen.