Future directions and concluding remarks
Our analyses included only experimental studies, with hosts exposed to a single parasite species and a single stressor. This approach, although easier to interpret and valuable to tease apart stressor effects in host-pathogen interactions, is difficult to translate to the natural world, where populations are likely exposed to multiple pathogens and a combination of stressors. When considering co-infections, for instance, stressors may compromise one arm of immune defense, making hosts more vulnerable to pathogens that require such response. For example, food restriction increased levels of eosinophils in capybaras (a Th2 immune response) and consequently reduced nematode burden (where resistance relies on the Th2 response), but coccidian infection intensity increased due to inadequate Th1 immune response (Eberhardt et al. 2013). Future studies should use a combination of field and laboratory experiments to perturb processes that covary with stressors to determine how and why results vary comparing laboratory and real-world conditions.
As a next level of complexity, host-pathogen systems do not occur in isolation, and some other biotic stressors and interactions can indirectly affect disease dynamics. For example, hosts compete for resources with other species and are consumed by predators. Consequently, stressors can affect other community members in ways that could enhance or negate epidemiological effects on hosts and pathogens (Strauss et al. 2015, 2016). Furthermore, most known pathogens infect multiple host species (Woolhouse et al. 2001), but some host species are disproportionately responsible for parasite transmission (Haydon et al. 2002). Generally, ecologically resilient species exhibit fast life histories and invest less in immune defense compared to more disturbance-sensitive species (Johnson et al. 2012; Previtali et al. 2012; Pap et al. 2015), predicting that resilient species will have insufficient immune response to prevent pathogen replication and transmission, resulting in higher transmission rates. Therefore, future research is sorely needed to evaluate the effects stressors have on different host species and their relative contribution to community disease transmission.
Moreover, combining experimental and modeling approaches is needed to move beyond associational patterns toward a mechanistic understanding of how stressors affect hosts and pathogens due to the common occurrence of multiple simultaneous stressors. Approaches are available for incorporating stressors into epidemiological models, such as examining variation in R0, the basic reproductive number of a parasite (Anderson & May 1991). Pinpointing when and how stressors increase or decrease R0 is crucial to understanding their roles in infectious disease dynamics. Though multiple mechanisms (including changes in host contact rates and per-contact probability of transmission) are often subsumed in the transmission parameter β, these need not be fixed, as we have illustrated with our models. The same applies to birth and death rates, and even to pathogen virulence, given that variation in host immune defenses alters per-contact transmission probabilities and the duration of the infectious period. As a next step, integrating a series of models with empirical results will inform the generality of predicted patterns.
Finally, our study highlights the need to expand empirical research at the interface of stress and infectious disease in highly relevant systems for zoonotic disease emergence. The studies included in our meta-analysis had low coverage of both vertebrates and terrestrial systems, yet terrestrial vertebrates such as rodents and bats have been linked repeatedly to zoonotic diseases affecting humans and livestock (Luis et al. 2013; Han et al. 2016). However, only one rodent study rodents provided sufficient data to be included in our meta-analysis (Eze et al. 2013).
As anthropogenic activities continue to alter ecosystems in ways that facilitate disease emergence worldwide, we must consider stressor effects on disease dynamics. Our findings improve our understanding of this interplay and provide insights for predicting and mitigating the impacts of stressor-pathogen synergies on human, animal, and planetary health.