How can we improve identification of contagious individuals? Factors influencing sickness detection

The comment by Kurvers and Wolf [1] on our recent publication [2] highlights some urgent questions that emerge not only in the wake of our study, but also in relation to the developing field of how behavioural mechanisms have evolved to defend against microbial infections. Their additional analyses of our data show that, even though there were small individual differences in the ability to discriminate between healthy and sick individuals (although some expert raters were identified), there were substantial differences in response bias, that is, how individuals balance the trade-off between sensitivity and specificity. Kurvers and Wolf also focus on how social-learning strategies could influence the detection of disease cues, and thereby disease dynamics, in the population in relation to the social network position of the ‘detector’. Thus, how do individuals differ in their ability to discriminate between sick and healthy peers, and in what way can social information—learning from others—influence one's responses to potentially sick people? There are good reasons to believe that exploring both these questions has the potential to cast light on a wide range of phenomena, from psychopathological mechanisms related to dysregulation of disease detection [3] to more societal matters such as attitudes, specifically prejudicial intergroup attitudes [4].

Kurvers and Wolf ask if ‘differences in cautiousness reflect an adaptive response to differences in the costs and benefits associated with behavioural immunity?’ This is likely true, given that variation is seen both across and within individuals in central disease-avoidance processes. For example, disease primes, such as watching images of viruses or bacteria, and hearing stories of how easily they are transferred, make people less positive to immigrants [5]. A number of similar studies show that disease primes affect attitudes, disgust, and avoidance behaviours. Watching images of sick people can also cause a more reactive immune system [6]. Together, these data indicate that humans are sensitive to sickness cues in the environment, and that such contextual factors shape expectations, immune responses, and how one behaves towards other people. A recent model similarly proposes that people continuously evaluate pathogen risks when deciding what to eat, what to touch, and with whom to have sex [7]. Since outgroup members pose a larger risk for pathogen exposure, this has also been suggested to contribute to favouring ingroup relationships and a xenophobia of outgroup members [8].

With respect to social-learning strategies, a recent review suggests that learning pathogen detection and proper avoidance behaviour depends on the informational content of the observed behaviour, as well as how this information is used by the observer [9]. The informational content is believed to include typical facial expressions related to disgust and withdrawal responses. While it would clearly be most efficient to learn from expert individuals, it is still unclear how one learns from others and how effective that learning is. Interesting aspects in this regard include the degree to which children learn these behaviours from their parents, from encounters with sick individuals and subsequent sickness, as well as whether people in healthcare settings should be considered expert individuals.

Disgust is a central emotion in avoidance of contamination and poisoning, but disgust sensitivity—the tendency to overestimate the potential negative consequences of disgust—varies greatly between individuals [10]. This suggests that the cost–benefit analysis of avoidance versus more active behavioural choices differs from person to person. While it is clear that women are more disgust sensitive than men [11], some research also suggests that individuals scoring high on disgust sensitivity suffer more from anxiety, have a tendency to avoid new foods, and are more constrained in their sexual behaviours [12]. A recent study found that individuals with health anxiety are more disgusted, worried, anxious, and believe they are more likely to be infected if interacting with other people [3]. While this may protect against pathogenic threats, it can also lead to reduced chances for obtaining sufficient food and social opportunities.

To conclude, there is ample support for the notion that the trade-off between sensitivity and specificity is modulated by a number of factors, including those related to context, traits, and learning. It seems likely that people change their cost–benefit analysis when sitting next to a sick person in a primary care waiting room, or after having had a really bad cold. We are excitedly awaiting research showing whether one can learn more effectively from experts, and whether artificial intelligence can be used to detect potentially contagious people.

Ethics

Information of the ethics can be found in the original article [2].

Data accessibility

Data from the original article are available at https://osf.io/btc7p/.

Authors' contributions

J.A. and M.L. drafted the reply. All authors revised and accepted the final version.

Competing interests

We declare we have no competing interests.

Funding

We received no funding for this study.