Disgust as an adaptive system for disease avoidance behaviour

1. Introduction: the problem of parasites

Parasites are ubiquitous; in some ecosystems their biomass rivals that of predators [1]. Parasitic viruses, bacteria, protozoa, nematodes, helminthes and arthropods live in durable relationships with their hosts, from whom they draw energy, shelter, transport and reproductive opportunity. They damage their host's inclusive fitness by producing toxins, manipulating behaviour to their own ends, and spreading to kin and community [2]. The costs of infection constitute an important selection pressure, which all animals face. As a result, natural selection has designed elegant and interlocking solutions to protect animals from parasite damage, including a range of physiological barriers and a complex immune system [3]. Beyond these physiological defences, animals also defend themselves from infection through behaviour that functions as a ‘behavioural immune system’ [4].

Pathogen avoidance behaviour has a long evolutionary history, and can be found in a broad range of taxa. For example, eusocial insects manage faecal wastes [5], crustaceans avoid diseased conspecifics [6], herbivores forage selectively to avoid faeces [7,8] and grooming behaviour is found in a range of species [9]. The disgust system is a psychological mechanism for producing pathogen avoidant behaviour [10–12]. In previous work, we have stressed the universality of disgust, showing that there is much that is similar about disgust responses between animals and humans, between humans and over historical time [11,13]. However, there is also much that differs between individuals and between social groups. In this paper, we look at disgust and disease avoidance behaviour in human individuals and in human social groups as an adaptive system. Natural selection has produced a solution to the problem of hard-to-detect parasites by designing a system that is sensitive to local information about infection risk. This system responds to parasite pressure not just over evolutionary time, but over lifetimes, using what cues it can. This may be information about an individual's current state, its history of sickness and exposure to disgusting experiences, or what it has learnt from the local culture and from the hygiene practices of others.

In this paper, we begin by recapping evidence that disgust is a universal driver of pathogen avoidance behaviour in humans. We then turn to sources of variation in the disgust system at the individual and group level. We set out the links between the hygiene behaviour of individuals and of groups and between individual disgust and the content of cultures. This allows us to integrate perspectives from psychology, ecology and cultural evolution, as well as epidemiological perspectives on disease prevalence, anthropological perspectives on manners and morality, and the symbolism of purity and pollution. Our conclusions stress the importance of using an evolutionary perspective in combination with interdisciplinary sources to provide an integrated understanding of the set of human behaviours that have a foundation in pathogen avoidance. This, may, in turn, offer insights that are important to the practice of public health.

2. The universality of disgust

Disgust is a fundamental part of human nature. Darwin was the first to propose that disgust is expressed universally [14] and many studies since then have supported this proposal [15,16]. Though there has been no systematic cross-cultural survey of the objects and events that elicit disgust in humans, the available data suggest that there is a universal set of disgust cues. These include bodily wastes, body contents, sick, deformed, dead or unhygienic people, some sexual behaviour, dirty environments, certain foods—especially if spoiled or unfamiliar—and certain animals [11,17,18]. Objects that have contacted any of the above can also become disgusting. Further, certain types of immoral acts are widely described as disgusting. Contact with disgust elicitors, real or imagined, is associated with (i) a characteristic facial expression that is recognizable across cultures [16,19], (ii) behaviour patterns that include withdrawal, distancing, stopping or dropping the object of disgust and shuddering [20,21], (iii) physiological changes including lowered blood pressure and galvanic skin response, recruitment of serotonin pathways, increased immune strength [22,23], and (iv) reports of negative affect including nausea.

Pathogen avoidance behaviour is universal across cultures, with all societies demonstrating individual and group-level hygiene behaviours. These include bodily, domestic and communal cleansing, avoidance of close contact or exchange of bodily fluids with others (with exceptions for mates and kin), and the avoidance of foods that are spoilt, contaminated or unfamiliar.

Explanations of disgust in the philosophical, anthropological, humanities and psychological literatures have been varied and inconsistent, reflecting this broad range of phenomena to be explained. The phenomenological philosopher Aurel Kolnai thought that disgust resulted from excess and surfeit: ‘A surplus of life, … an indifference to quality … a desire towards death’ [24, pp. 72–73]. Freud considered disgust a learned reaction formation that could be cultivated towards any activity through development [25]. The social anthropologist Mary Douglas argued that dirt and disgust are a product of culture, such that anomalous objects and events that do not fit the local cosmology have to be rejected, so as not to threaten social order [26]. In psychology, the dominant Rozin–Haidt school has disgust originating in the rejection of spoiled foods, but also serving to cope with the existential terror of being an animal and hence mortal [21,27]. A recent cultural study of disgust labels it ‘the Hydra’ since it seems too complex to explain [28].

An evolutionary perspective, however, provides a parsimonious explanation for the multiple elicitors and the behavioural tendencies of the disgust system. According to Darwinian thinking, disgust should be considered an adaptive system that drives the behavioural avoidance of infectious disease [11] Constant selection pressure from the ubiquitous presence of pathogenic parasites in animal and human ancestral environments [29] would have selected for those individuals with alleles disposing towards a ‘behavioural immune system’ preventing contact with, and incorporation of, pathogens [4]. Setting aside, for the present, the issue of moral disgust, it can be seen that all of disgust's basic elicitors (listed above) are implicated in the risk of transmission of infectious disease [11] and paired stimuli with, and without, disease risk show significant differences in disgust response [10,12]. This relationship between disgust elicitors and disease sources appears consistently across cultures and through the historical record [14,20,30]. Rats and other mammals display the characteristic gape expression after eating noxious food, and, as in humans, this reaction is dependent on the insular cortex [31] suggesting that disgust may be at least a pan-mammalian adaptation. Because behavioural immune systems are ubiquitous in animals and predate the evolution of modern humans, all humans should come equipped with a disgust system, rather than learning disgust as a product of culture (as Douglas, Freud and their followers have argued). It is also unlikely that disgust in humans originated in food distaste, because of the pan-vertebral need to avoid pathogens of all types, not just those using food as a vector of infection.

Brain imaging studies also show that there is a specific network associated with disgust. Viewing images of disgusting stimuli, or videos of people with disgusted expressions results in robust and recurring activation in specific brain areas—a neural network including the anterior insular cortex, basal ganglia, ventrolateral prefrontal cortex, anterior temporal cortex, medial prefrontal cortex and visual cortex [25,32,33]. Autobiographical recall of disgusting episodes activates the insular cortex and the basal ganglia [34], as does exposure to disgusting smells [33]. Two recent meta-analyses of functional neuroimaging studies of emotion found that activation of the basal ganglia was reliably associated with disgust [35].

3. The variability of disgust

Although disgust has an ancient and universal function, the disgust system reacts with different levels of activation to the same stimulus between individuals and over the lifetime of the same individual [27,36].

Figure 1 shows global variation in disgust sensitivity from a web survey with 38 845 participants. They were asked to rate how disgusting they found a series of disease-relevant images such as a sick person, a plate of what looked like bodily fluids and a crowded underground train, on a Likert scale of 0–5 (see [10] for details of methods). The results showed a high degree of variability between individuals (average standard deviation = 0.83). Differences in variability between continents were not statistically significant, varying only between 3.77 and 3.94 (s.d. = 0.05). Though a multivariate ANOVA model controlling for age, gender and occupational differences found cultural region to be a significant predictor of disgust sensitivity, this result was largely explained by the low average disgust sensitivity in the Australia/Oceania sample.

How do differences in disgust sensitivity arise? Evolutionary psychological approaches assume that psychological systems—including emotions—represent solutions to adaptive problems repeatedly encountered during evolutionary history. Plasticity in these systems should reflect the degree of variability in these problems: low levels of variability over the lifespan and between generations will favour the evolution of highly constrained systems. Such systems reliably develop in a broad range of environments with little phenotypic variation and have the benefit of efficiency and rapid and reliable development [37]. Greater environmental variation, on the other hand, should favour more plastic systems that use environmental—including social—information to adaptively shape behaviour according to the challenges of the local environment [38].

Pathogen pressure has led to the selection of a disgust/hygiene behaviour system in individuals that is both universal and plastic to local environmental variation. Figure 2 schematizes the factors that influence the human disease avoidance system. While disgust motivates hygiene behaviour, disgust sensitivity varies between individuals as a trait and within individuals by their state and also through individual learning. Disgust and hygiene in ultrasocial humans facing shared pressure from pathogens are not, however, simply a matter for individuals. The lower half of the diagram depicts group effects in the adaptive system. Culture—the socially acquired information shared by a particular group—affects the individual system through social learning and group hygiene behaviour through norms about manners. Group hygiene can be symbolically extended into cultural ideas about purity and pollution and can also affect public health by influencing the prevalence and virulence of pathogens in the environment. Group hygiene also influences the hygiene behaviour of individuals via imitation. The content of culture is, itself, a product of the individual brains that support it and so reflects the predispositions of those brains. Finally, pathogens exert selective pressure on the whole system.

All of the components of the system, whether in brains (shown on the left of the diagram) or in behaviour (shown on the right) are affected by environmental factors extraneous to the disease avoidance system, including seasonality, climate, ecosystem, habitat and the particular structure of host populations (not shown). In the following sections of this paper, we describe the elements of this adaptive system and how it works. (For ease of reference, numbers in the figure refer to the relevant sections of the paper.)

4. Disgust in brains: sources of diversity

Figure 2 shows pathogens exerting selection pressure on individual disgust and the hygiene behaviour it motivates. However, selection has not produced a constant level of disgust sensitivity. Variation can be accounted for by three kinds of phenomena: (i) innate trait differences; (ii) by plasticity, which allows for adaptation to current states; and (iii) the ability to learn from changes in the environment over the lifespan. These types of influence will be treated in turn.

5. Hygiene behaviour

In figure 2 we show individual hygiene behaviour as a product of the disgust system in individual brains and also a product of the copying of the hygiene behaviour of others.

6. Group hygiene behaviour

So far we have looked at the effects of proximate factors on disgust sensitivity in brains and the hygienic behaviour of individual humans. Yet, humans are a social species; our behaviour reflects social as well as individual considerations. The lower half of figure 2 schematizes group-based effects on the disease avoidance system.

7. The effects of culture on disgust and hygiene

Culture consists of the pool of skills, attitudes, beliefs and values which have been socially learned by the set of individuals in a population or group [93]. Figure 2 shows culture affecting disgust via social learning (which we discussed in §4b). Culture also affects the hygiene behaviours that are adopted by groups (and hence, the behaviour of individuals) through the power of norms and manners. We also show group culture as, at least partially, a product of the biases of individual brains through predispositions to take on board certain ideas and not others. In modern societies, culture has also come to contain scientific ideas about pathogens and disease, and microscopic pathogens have finally become ‘visible’. We will discuss the consequences of these developments for the adaptive system at the end of §8.

8. Discussion: the disease avoidance system

In this paper, we have set out the components of the adaptive system that produces disease avoidance behaviour. We have seen that natural selection has produced a system that has taught individuals how to behave hygienically. Pathogens are, however, not just an issue for individuals, but for groups, and as a highly social animal, humans have group responses to diseases. Hence we have postulated a parallel system whereby group brains (culture) affect group behaviour as regards hygiene, and vice versa, and these can have important effects on the prevalence of pathogens. These group effects affect the brain and behavioural responses of individuals.

While we have set out evidence from many types of source about the individual links, is there any evidence of it acting as a complete system? Some intriguing new studies provide evidence that this is the case. It has been shown that societies faced with high pathogen pressure have higher average scores on personality traits such as extraversion and openness to experience [97], higher average scores on collectivism versus individualism scales [98] and higher numbers of religions [99]. These effects could not be explained by latitude, climate or socio-economic status. Further, in the first two studies, historical disease prevalence was a better predictor of these psychological variables than contemporary prevalence, suggesting that disease risk is a plausible cause, rather than a consequence, of the cultural differences observed. Taken together, these results suggest that societies that have faced high pathogen threats tend to become more inwardly focused, leading to less mixing and hence less contact with potential pathogen threats in the form of individuals from other groups.

Figure 2 suggests three ways in which the system may be operating to generate these effects. First of all, natural selection may have been in operation in the past history of the individuals in these groups, leading those under high pathogen stress to evolve higher levels of hygiene, driven by higher disgust sensitivity. This could have translated into lower willingness to engage with other cultures, and greater conservatism (which has been shown to be associated with conservative values [100]). However, when we compared the same disease prevalence data with our own disgust sensitivity scores from the dataset shown in figure 1, we found no significant correlations, which argues against genetic differences between groups owing to varying histories of pathogen prevalence.

The second way of explaining the observed relationships between pathogens and group ‘openness’ is that individuals are responding to the perception that there is a lot of disease about, i.e. sickness in others causing a risk, which drives individual learning, which modifies disgust responses, which shows up in lower openness scores and enhanced xenophobia. Navarrete & Fessler [101] observed that, not only does sensitivity to disgust predict more negative attitudes towards foreign peoples (xenophobia), but also predicts more positive attitudes towards one's own cultural in-group (ethnocentrism). They also showed that another disease-relevant individual difference variable—PVD—predicts both xenophobia and ethnocentrism. Further, xenophobic reactions to foreigners have been found to be stronger among people who feel personally vulnerable to germs and disease (as measured by the PVD scale) [102]. More particularly, results showed that higher levels of PVD predicted stronger anti-immigrant attitudes towards those from subjectively foreign locations. There was no such effect on attitudes towards culturally familiar immigrant populations. It thus seems probable that the intervening variable that can explain the observed relationship between pathogens and cultural characteristics is, in fact, disgust. A similar effect was observed in World of Warcraft (a web-based multi-player game). When a plague of ‘Corrupt blood’ began to kill up to half of the players, they began to avoid big cities [103].

A final possibility is that selection is operating on social groups. Those groups with the strongest collective anti-pathogen defences, such as strong norms about manners or avoiding strangers, outcompete those groups without. Using a model which carefully spells out the links in this adaptive system thus helps reasoning about the ways in which a complex set of results can be better understood.

The interrelated adaptive system we have described is not, of course, static. It exhibits positive feedback loops that can lead to interesting dynamics. For example, the proportion of hygiene cooperators in a population may fluctuate cyclically because as defectors from group-level hygiene activity become frequent, so will pathogens. When pathogen stress is high, it once again pays to cooperate, and the proportion of defectors should again decrease. The proportion of hygiene defectors may thus go up and down in response to fluctuations in the costs of cooperation. This dynamic has been formally modelled in the context of risky versus safe sexual behaviour: as the proportion of ‘safe’ strategists increases, the prevalence of sexually transmitted diseases (STDs) decreases. In this changed environment, ‘risky’ individuals may have some advantage. Oscillations of high disease/safe behaviour, low disease/risky behaviour thus occur [104].

There is also a positive feedback loop between culture and group hygiene by which cultural rules with real biological import (manners) can become subject to the symbolization principle we call ‘extension’. When a set of cultural rules becomes symbolic, they can extend beyond behaviour with direct relevance to biological pathogens. An obvious extension is from rules governing contact with biological parasites to rules governing contact with ‘social parasites’ (i.e. individuals that claim an unfair proportion of social resources). In this way, systems of manners can be extended to become systems of moral rules. Violators of rules for the apportionment of socially produced resources can then be labelled as disgusting, and sanctioned accordingly.

It is probable that there have also been major shifts in the way the system has operated since the dawn of Homo sapiens. Changes in environment, social structures and technology have led to changes in pathogen prevalence, as well as in individual and group hygiene. While the system operated ‘blindly’ for most of human history, Leewenhoek's invention of the microscope 300 years ago finally allowed us to ‘see’ pathogens. Pasteur & Koch popularized the idea that invisible germs caused disease, and the idea has now caught on globally, perhaps because the notion of an invading parasitic life form is so exquisitely disgusting [13]. The scientific method has also allowed us to determine the most effective and cost-effective methods of disease prevention, so as to inform health policy internationally [105]. We have thus moved from the instinctive practice of safe hygiene, to more educated approaches. However, the ancient emotion of disgust retains its power; a recent hygiene promotion campaign in Ghana successfully elicited disgust to increase rates of handwashing nationally [106]. We now plan to test whether promoting hygiene as good manners is an effective disease control strategy.

9. Conclusions

We have seen that disgust, hygiene behaviour and culture form an interlinked adaptive system, which has long served to reduce the dangers of disease. We have argued that disgust in the brain and the disease avoidance behaviour that it motivates is universal in humans (and in other animals) and is a product of the selection pressure of pathogens in the environment. However, disgust is also plastic, being able to retune according to signals from within the body and from the social and biological environment. Heightened disgust sensitivity leads to heightened disease avoidance behaviour (hygiene). However, as a social species, group hygiene behaviour is not just an aggregate of individual behaviour but the consequence of individuals using each other as models to be imitated, and as resources for cooperation in healthy behaviours. Changes in group hygiene behaviour, including cooperative public health activities, can reduce pathogen prevalence. The tendency to defect on responsibilities to the group can be countered by norms imposed by culture. The content of culture is itself biased by learning mechanisms based in the disgust system in individual brains.

We have shown that humans are equipped with a series of highly specialized learning mechanisms and biases. Learning about disease (and disgust) is not the product of domain-general systems; features of these learning mechanisms clearly reflect the nature of the disease avoidance problem they interact with. But the informational inputs to these systems differ, both between people and across groups, as a consequence of local disease problems and the learning histories of the individuals in one's social group. Individual differences are therefore inevitable. This variation, however, is frequently presented as evidence that evolutionary explanations are inappropriate (e.g. [107,108]). On the contrary, variable responses in the disgust system are indicative of a rich evolved psychology capable of absorbing and integrating information from diverse sources, and generating adaptive behaviour in a wide range of environments.

The adaptive system depicted in figure 2 highlights the intricate interrelations between the content of individual brains and group brains (culture) and individual and group behaviour. We have used it to explore one human need—the avoidance of disease. However, humans have multiple evolved needs including a need for food, to avoid predators, to rear young, to maintain social affiliations, etc. [109]. Our framework should prove useful in integrating perspectives from across disciplines for other similar adaptive systems.

Infectious disease still remains a leading cause of mortality worldwide [110] and promoting safe hygiene may be one, if not the most, cost-effective means of preventing disease [105]. Evolutionarily informed work that can explain the causes of disease avoidant behaviour may offer vital clues as to how best to change environments and cultures so as to favour changes in group and individual behaviour, and hence to prevent this annual toll of infectious disease. We hope that those who seek to promote disease avoidant behaviours will recognize the power of using the disgust system's natural affinity for producing adaptive responses to disease threats when developing future public health programmes.