When dengue case reports from Brazil, Colombia, and Southeast Asia are normalized by total outbreak size and plotted against age, the resulting curves collapse onto a single shape. The same holds for Zika and chikungunya. Three different diseases, across different continents, different population sizes, and different years, produce effectively the same age-dependent proportion of cases. The distribution does not scale with outbreak magnitude or shift with geography. It is invariant.
The invariance breaks cleanly at the boundary between vector species. Malaria and leishmaniasis --- transmitted by Anopheles and sandflies respectively --- produce age distributions entirely unlike the Aedes-transmitted trio. The shape is not a property of the disease, the pathogen, or the human population. It is a property of the vector. The mosquito species determines which age groups get bitten in what proportions, and this ecological fact dominates every epidemiological variable that might be expected to matter more.
Note (2026-11-10): I checked the abstract of arXiv 2603.16773. It reports that the age distribution of the proportion of reported cases is essentially the same for the three Aedes-related infections (dengue, Zika, chikungunya) regardless of outbreak size and region, and entirely different for infections transmitted by other vectors, and says the proportion is probably determined by ecological interactions between vectors and hosts. The essay states flatly that the shape is a property of the vector and that this dominates every other epidemiological variable; the abstract says probably, and the account of biting times and habitats is the essay's own.
This is a strange inversion. Epidemiological models typically treat transmission rates, population immunity, and reporting biases as the primary shapers of case distributions. The finding suggests that something more primitive --- the behavioral ecology of the vector, its preferred biting times, its habitat preferences, its host-seeking patterns --- imposes a signature so strong that it persists through all the noise of human variation. The mosquito's daily routine is more epidemiologically stable than the health system reporting it.
The broader implication is that in any system where an intermediary connects cause to effect, the intermediary's behavioral pattern may dominate the observable outcome more than either the cause or the effect themselves. The messenger shapes the message more than the sender.