No evidence for a classic transmission-duration tradeoff in human malaria infections

Pathogenic organisms are typically thought to be constrained by a tradeoff between the rate and duration of transmission, an assumption that underpins a considerable body of evolutionary theory. Here we test for a transmission-duration tradeoff using detailed historical malaria infection data from an era prior to widespread use of antibiotics when humans were deliberately infected with malaria parasites as treatment for neurosyphilis (malariatherapy). These time series follow individual human infections until recovery or treatment with antimalarial drugs due to acute need (a proxy for virulence), and include data on the abundance of specialized transmission stages that can be used to project parasite fitness. We fit a model to estimate initial parasite multiplication rates (PMRs) and find that faster within-host multiplication extends infection duration (time until recovery) and enhances parasite fitness without a discernible cost, such as increased virulence. Initial PMRs exhibit strain-specific differences, a feature required for evolution by natural selection, but our results contradict the idea that the evolution of human malaria parasites is constrained by a transmission-duration tradeoff. Significance statementPathogenic organisms are usually assumed to face a tradeoff such that aggressively exploiting host resources enables more efficient transmission but at the cost of shorter infections. If such a classic transmission-duration tradeoff is not general, then it is not clear what prevents pathogenic organisms from evolving to exploit their hosts ever more aggressively. We use historical data from human malaria infections to show a remarkable lack of evidence for a transmission-duration tradeoff, since faster parasite multiplication tends to prolong infections and enable more efficient transmission. Therefore, classic predictions regarding the evolution of infection-induced harm to hosts may not apply to human malaria parasites, and efforts to locate general evolutionary constraints on pathogenic organisms should look beyond a classic tradeoff. FundingThis work was supported by the Cornell University College of Agriculture & Life Sciences (M.A.G.). L.M.C. was partially supported by the National Science Foundation Grant # 2144680. Data availabilityAll supporting data and code are provided as supplemental files. NoteFor ease of reviewing, this MS includes all elements (including figures) embedded in the text. If accepted, we would be happy to provide elements as separate files.