Patterns and variation in the mammal parasite-glucocorticoid relationship.

Parasites are ubiquitous and can strongly affect their hosts through mechanisms such as behavioural changes, increased energetic costs and/or immunomodulation. When parasites are detrimental to their hosts, they should act as physiological stressors and elicit the release of glucocorticoids. Alternatively, previously elevated glucocorticoid levels could facilitate parasite infection due to neuroimmunomodulation. However, results are equivocal, with studies showing either positive, negative or no relationship between parasite infection and glucocorticoid levels. Since factors such as parasite type, infection severity or host age and sex can influence the parasite-glucocorticoid relationship, we review the main mechanisms driving this relationship. We then perform a phylogenetic meta-analysis of 110 records from 65 studies in mammalian hosts from experimental and observational studies to quantify the general direction of this relationship and to identify ecological and methodological drivers of the observed variability. Our review produced equivocal results concerning the direction of the relationship, but there was stronger support for a positive relationship, although causality remained unclear. Mechanisms such as host manipulation for parasite survival, host response to infection, cumulative effects of multiple stressors, and neuro-immunomodulatory effects of glucocorticoids could explain the positive relationship. Our meta-analysis results revealed an overall positive relationship between glucocorticoids and parasitism among both experimental and observational studies. Because all experimental studies included were parasite manipulations, we conclude that parasites caused in general an increase in glucocorticoid levels. To obtain a better understanding of the directionality of this link, experimental manipulation of glucocorticoid levels is now required to assess the causal effects of high glucocorticoid levels on parasite infection. Neither parasite type, the method used to assess parasite infection nor phylogeny influenced the relationship, and there was no evidence for publication bias. Future studies should attempt to be as comprehensive as possible, including moderators potentially influencing the parasite-glucocorticoid relationship. We particularly emphasise the importance of testing hosts of a broad age range, concomitantly measuring sex hormone levels or at least reproductive status, and for observational studies, also considering food availability, host body condition and social stressors to obtain a better understanding of the parasite-glucocorticoid relationship.