Social buffering and contact transmission: network connections have beneficial and detrimental effects on Shigella infection risk among captive rhesus macaques

Abstract

In social animals, group living may impact the risk of infectious disease acquisition in two ways. On the one hand, social connectedness puts individuals at greater risk or susceptibility for acquiring enteric pathogens via contact-mediated transmission. Yet conversely, in strongly bonded societies like humans and some nonhuman primates, having close connections and strong social ties of support can also socially buffer individuals against susceptibility or transmissibility of infectious agents. Using social network analyses, we assessed the potentially competing roles of contact-mediated transmission and social buffering on the risk of infection from an enteric bacterial pathogen (Shigella flexneri) among captive groups of rhesus macaques (Macaca mulatta). Our results indicate that, within two macaque groups, individuals possessing more direct and especially indirect connections in their grooming and huddling social networks were less susceptible to infection. These results are in sharp contrast to several previous studies that indicate that increased (direct) contact-mediated transmission facilitates infectious disease transmission, including our own findings in a third macaque group in which individuals central in their huddling network and/or which initiated more fights were more likely to be infected. In summary, our findings reveal that an individual's social connections may increase or decrease its chances of acquiring infectious agents. They extend the applicability of the social buffering hypothesis, beyond just stress and immune-function-related health benefits, to the additional health outcome of infectious disease resistance. Finally, we speculate that the circumstances under which social buffering versus contact-mediated transmission may occur could depend on multiple factors, such as living condition, pathogen-specific transmission routes, and/or an overall social context such as a group's social stability.

Keywords: Contact-mediated transmission; Infectious disease risk; Nonhuman primate; Social buffering; Social networks.

Figure 1. Diagnostic box-plots of between-group similarities and differences in relationships between network metrics and Shigella infection.

Each plot shows the mean and distribution of a network metric (Y axis) plotted for infected and uninfected individuals in each group (X Axis). Red box-plots represent comparisons for Groups I and II, and Yellow comparisons for Group III.

Figure 2. Group I: grooming social network and Shigella infection risk.

Social Network graph of grooming relationships indicating the effect of social buffering on Shigella infection in Group I. Nodes are sized proportional to the eigenvector centralites of individual macaques. Infection (yellow nodes: N = 23 out of 101) is mostly restricted to individuals with the lowest eigenvector centralities.

Figure 3. Group III: aggression social network and Shigella infection risk.

Social Network graph of aggression relationships indicating the effect of social contact on Shigella infection in Group III. Nodes are sized proportional to the aggression out-degrees of individual macaques. Infection (yellow nodes: N = 7 out of 100) is prevalent among individuals with moderate-to-high out-degree.

Figure 4. Group III: aggression, huddling, and Shigella infection risk.

Scatter-plot showing the effect of social contact on Shigella flexneri infection for Group III. Black dots are infected macaques (N = 7 out of 100), and are concentrated among those with moderate-to-high huddle betweenness and aggression out-degree.