Understanding how temperature shifts could impact infectious disease

Abstract

Climate change is expected to have complex effects on infectious diseases, causing some to increase, others to decrease, and many to shift their distributions. There have been several important advances in understanding the role of climate and climate change on wildlife and human infectious disease dynamics over the past several years. This essay examines 3 major areas of advancement, which include improvements to mechanistic disease models, investigations into the importance of climate variability to disease dynamics, and understanding the consequences of thermal mismatches between host and parasites. Applying the new information derived from these advances to climate-disease models and addressing the pressing knowledge gaps that we identify should improve the capacity to predict how climate change will affect disease risk for both wildlife and humans.

Fig 1. Several important advances to understanding how temperature shifts affect infectious disease have been made in since 2013, when some of the most recent reviews on the topic were published.

These advances include (a) improvements to mechanistic disease models; (b) investigations into the importance of climate variability to disease dynamics; and (c) an improved understanding of how thermal mismatches affect host–parasite interactions. Primary knowledge gaps in each of these 3 areas are also provided. SIER, susceptible–exposed–infected–recovered; SIR, susceptible–infected–recovered. The image in panel b was obtained from a Google Image Creative Commons License Search (https://webstockreview.net/image/cold-clipart-cold-climate/2528297.html?text=DMCA+Report+-+https%3A%2F%2Fwebstockreview.net%2Fimage%2Fcold-clipart-cold-climate%2F2528297.html). The frog image in panel c was provided by Jeremy Cohen, and the chytrid fungus image below is available at https://www.commons.wikimedia.org/wiki/File:Batrachochytrium_species_morphology_in_culture.

Fig 2. Graphical representation of a simple compartment model for a vector-borne disease.

Red diamond compartments represent the vectors and blue squares the host population. S–I–R indicates susceptible, infected, and recovered, respectively, for the hosts. S_v, E_v, and I_v indicate susceptible, exposed, and infected for the vectors. Solid lines indicate individuals moving between compartments, dashed lines the route of infection, and dotted lines demographic processes in the vectors. We present an example of a vector-borne disease model because it is more complex than models of directly transmitted parasites.

Fig 3. Depiction of the TMH.

(a) Cold-adapted parasites in isolation most outperform cold-adapted noninfected hosts at warm temperatures (double-sided arrow), whereas (b) warm-adapted parasites most outperform warm-adapted hosts at cool temperatures. Hence, when cold- and warm-adapted parasites grow in or on hosts, the TMH predicts that their prevalence will peak at warm (c) and cool temperatures (d) and be left (c) and right skewed (d), respectively. Given that parasites require a minimum, threshold density of hosts to persist [50], the hypothesis assume that there are little to no parasites at temperatures where hosts perform very poorly. Thus, within the threshold performance/density of hosts (purple bands), much of the performance of the parasite growing in or on the host can often be characterized using a linear relationship (blue and red arrows in panels c and d), despite the relationship being non-monotonic (both increasing and decreasing, i.e., hump shaped) across the entire temperature range. This is because the relationship within this range is predicted to be monotonic and, with error, might often be well fit with a line or some other monotonic function. Given this, the TMH has often been tested 1 of 2 ways: (e) by testing for the effect of a statistical interaction between the temperature during disease sampling and the temperature to which host are adapted (usually measured as the 50-year mean temperature at the location of their collection) on infection probability or (f) by testing for a negative relationship between the 50-year mean temperature of hosts and the temperature of peak prevalence minus the 50-year mean temperature of hosts. TMH, thermal mismatch hypothesis.