Zika Phylogeography Finds Shifting Source Hubs Across Two Epidemic Eras
A phylogeographic network analysis of Zika virus genomes traces a changing pattern of cross-country spread from 2012 through 2025, with the earliest inferred movement emerging from Thailand before the virus expanded through the Americas and later reappeared from Asian source hubs. The study reconstructed a time-scaled tree from 1,043 viral genomes and converted branch-to-branch location changes into 81 source–destination–year events spanning 39 locations. The result is not a simple map of where cases occurred; it is a dynamic picture of where the virus appears to have been exported, when those exports clustered, and which countries repeatedly acted as transmission origins.
The temporal arc matters. Early activity began in Asia, but most identified cross-border events clustered in the Americas during 2014–2016, when Brazil became the principal source hub. After that epidemic peak receded, the geography changed again. All six events identified from 2019 onward originated from Thailand or India and reached destinations in Asia, Europe, and Africa. In practical surveillance terms, the network no longer points to a single enduring center of gravity. It points to a virus capable of resuming export from geographically distinct source regions when ecological and travel conditions align.
Why Air Travel, Incidence, and Distance Aligned With Export Events
The analysis linked identified transmission events with several country-level factors. High source-country incidence was associated with greater odds of a cross-country event, with an odds ratio of 2.66. Greater air travel was also associated with export events, with an odds ratio of 2.52. Higher destination-country gross national income per capita showed a positive association as well, with an odds ratio of 1.70. In contrast, greater geographic distance sharply reduced the odds of an event, with an odds ratio of 0.11, and higher source-country gross national income per capita was inversely associated, with an odds ratio of 0.64.
These are not abstract modeling outputs; they fit the biology of a mosquito-borne virus that still depends on human movement to bridge borders. High incidence in the source country increases the pool of infections available for export. Air travel can move infected travelers faster than local surveillance systems can react, especially when viremic individuals reach settings with competent Aedes vectors. Geographic distance behaves as a barrier because longer routes require more opportunities for transport interruption, behavioral friction, and time for detection. The income signal is more complicated, but the direction of association suggests that lower-income source settings may be more vulnerable to sustained local transmission, while higher-income destinations may simply be more connected and more likely to receive imported lineages that are detected and sampled.
The network structure also exposed persistence beneath the headline epidemic phases. Fourteen source countries showed elevated residual source propensities, with Brazil and Thailand among the highest. Residual propensity here suggests that even after accounting for the measured covariates, some countries still contributed more export events than expected. In outbreak intelligence, those residuals matter because they point to structural features not fully captured by incidence, travel volume, income, or distance alone—perhaps local vector ecology, surveillance intensity, population movement patterns, or the timing of outbreaks relative to seasonal conditions.
Thailand’s Long Source Activity and the Return of Asian Exportation
One of the most striking findings is the duration of Thailand’s inferred source activity, which spanned 2012–2025. That time window stretches across the earliest and the most recent segments of the network, linking the pre-American phase with the post-2019 pattern of renewed Asian exportation. The country appears less like a transient waypoint and more like a sustained node within the global transmission system.
Brazil, by contrast, dominated the middle years. The 2014–2016 concentration of events in the Americas matches the period when the virus expanded explosively across that region, and the network analysis places Brazil at the center of that movement. The shift back toward Asia in recent years signals that Zika’s transnational spread is not locked into one hemisphere or one historical phase. The source hubs move with changing epidemiology, travel patterns, and perhaps seasonal vector suitability.
This is where phylogeography offers value beyond routine case counting. Reported cases describe where disease was recognized. Viral genomes describe where lineages moved. When those two views are combined, cross-border spread becomes visible as a network rather than a set of isolated national outbreaks.
Surveillance Implications for 2026 and Beyond
The study’s conclusions point toward a surveillance strategy built around country pairs rather than single-country alerts. High-incidence source settings, heavy air-travel links, and short geographic separations identify routes where importation is more plausible and where public health systems may need faster genomic confirmation. The Americas remain an established corridor for Zika movement, but persistent or reemerging Asian sources deserve equal attention.
The final caution in the paper reaches beyond the historical dataset. El Niño may increase vector-borne transmission potential in 2026–2027. If warmer or wetter conditions expand mosquito breeding and transmission suitability, a revived export pattern could become more consequential, especially where travel ties connect endemic or resurgent source countries to receptive destinations. The biology is familiar: when vector abundance rises and infected travelers move across borders, the virus finds new ecological openings.
The evidence still comes from a preprint and has not undergone peer review. It also depends on the genomes that happened to be sampled and shared, which means unsampled circulation can alter the apparent shape of any transmission tree. Even so, the signal is strong enough to suggest that Zika’s source geography has shifted over time rather than remaining fixed, and that surveillance systems should be built to detect those shifts early, before imported lineages seed local transmission in receptive settings.
- • medRxiv preprint DOI 10.64898/2026.11.12.26367925: https://doi.org/10.64898/2026.11.12.26367925