Parasites, and the arms races between parasites and other organisms, have long been associated with theories of sexual selection because organisms benefit from choosing parasite-free and parasite-resistant mates (e.g., Hamilton & Zuk Reference Hamilton and Zuk1982). With rising parasite load, organisms should channel their energies away from display to fight parasites and so may be less able to invest in attractive traits (Folstad & Karter Reference Folstad and Karter1992). Because parasites are generally detrimental to health and survival, high parasite load increases the importance of selection of mates who possess traits indicating resistance to parasites; and so we can predict that relative parasite stress will affect human preferences for traits associated with health and disease resistance.

Several previous studies in humans have reported positive correlations between various measures of parasite stress and preferences for putative cues of long-term health, suggesting that individuals place greater emphasis on potential mates' health cues in regions where pathogens impose a greater selection pressure (Gangestad & Buss Reference Gangestad and Buss1993; Low Reference Low1990). For example, in a study of 29 cultures, Gangestand et al. (2006) observed positive correlations between a measure of pathogen prevalence and the importance placed on potential mates' physical attractiveness, health, and intelligence.

Human faces contain several potential cues to parasite resistance and have been the focus of much research in attractiveness (e.g., Thornhill & Gangestad Reference Thornhill and Gangestad1999). A recent study of 30 Westernized countries showed that regional differences in women's preferences for masculine traits in men's faces, a cue of men's long-term health (Rhodes et al. Reference Rhodes, Chan, Zebrowitz and Simmons2003; Thornhill & Gangestad Reference Thornhill and Gangestad2006), were correlated with a potential proxy measure for parasite stress: a composite health measure derived from various World Health Organization statistics on mortality and life expectancy (std β=−0.515, t=−3.18, p=0.004; DeBruine et al. Reference DeBruine, Jones, Crawford, Welling and Little2010). The relationship between women's masculinity preferences and this health measure remained even after controlling for regional variation in wealth and mating strategies (DeBruine et al. Reference DeBruine, Jones, Crawford, Welling and Little2010) or controlling for homicide rates (DeBruine et al. Reference DeBruine, Jones, Little, Crawford and Welling2011), a potential indicator of intrasexual competition (Brooks et al. Reference Brooks, Scott, Maklakov, Kasumovic, Clark and Penton-Voak2011). A similar correlation (std β=−0.478, t=−3.77, p<0.001) was also observed in a U.S. sample using a composite health measure derived from the United States Centers for Disease Control statistics on mortality due to illness and disease across 50 states, even after controlling for regional variation in wealth, income inequality, fertility, and homicide rates (DeBruine et al. Reference DeBruine, Jones, Little, Crawford and Welling2011). These data indicate that women in regions with lower health exhibit stronger preferences for secondary sexual traits associated with long-term heath in male faces.

Here, we have re-analyzed both samples of data using Fincher & Thornhill's (F&T's) measures of parasite stress. Linear regression using the weighted least squares (WLS) method to control for number of participants per country showed that, across countries, there was a significant positive relationship between parasite stress and women's preferences for masculine men (std β=0.654, t=4.58, p<0.001). Using the same analysis, across the states of the United States, a significant positive relationship was also observed between these variables (std β=0.443, t=3.43, p=0.001). These re-analyses show that F&T's parasite stress measures generate the same results as our previously used composite health measures; parasite stress predicted regional variation in women's preference for male facial masculinity in a nearly identical way to the health measures. Indeed, our health measures and these parasite stress measures were highly correlated in both samples of 30 Westernized countries (r=−0.880, p<0.001) and 50 US states (r=−0.668, p<0.001).

Although many studies have implicated pathogen stress in regional variation in behavior, the correlational nature of these studies and the large number of covarying ecological factors mean that it is not possible to draw firm conclusions about the effects of pathogen stress on behavior (Brooks et al. Reference Brooks, Scott, Maklakov, Kasumovic, Clark and Penton-Voak2011; DeBruine et al. Reference DeBruine, Jones, Little, Crawford and Welling2011; Lee & Zeitsch Reference Lee and Zietsch2011; Little et al. Reference Little, DeBruine and Jones2011). F&T address this limitation of correlational studies by noting experimental research in which viewing cues to disease salience affected behavior in ways predicted by their parasite-stress theory of social behavior, whereas viewing control images did not (e.g., Mortensen et al. Reference Mortensen, Becker, Ackerman, Neuberg and Kenrick2010). Similarly, we have also found that viewing cues to disease salience increased preferences for facial cues of long-term health in potential mates, but not same-sex individuals (Little et al. Reference Little, DeBruine and Jones2011). Similarly, after women are primed with questions about pathogen prevalence, their preferences for traits indicating genetic quality (e.g., intelligence and muscularity) increase in relation to preferences for traits indicating parental quality (e.g., kindness and nurturing; Lee & Zeitsch in press).

Collectively, these correlational and experimental findings, together with our new analyses, suggest that F&T's parasite-stress theory of social behaviors and attitudes can be usefully extended to understand regional variation in mating behaviors and preferences, further underlining the importance of pathogens in shaping human behavior.