An Analysis of Travel and Spatial Risk Factors for Incident Dengue in Kenya

Dengue (DENV) is currently the most prevalent mosquito-borne disease, seen in a myriad of tropical countries. Travel has long been associated with infectious diseases, and thus, traveling to areas where Aedes aegypti and Aedes albopictus proliferate may be a risk factor for contracting dengue fever. This travel and spatial analysis may be informative to citizens of dengue-endemic countries, particularly Kenya, and to public health officials who have a goal of searching for optimal regions to implement dengue-related interventions.
The goal of this secondary data analysis was to assess the association between travel, various travel patterns, and incident DENV, as well as a spatial analysis of the distribution of dengue serotypes and dengue coinfections in four sites in Kenya. Data were obtained from an ongoing cohort study from the LaBeaud lab. These data were taken from four clinical sites in Kenya, designed to be a representative sample of the overall Kenyan population. This was done by having two clinical sites in Coastal Kenya and two sites in Western Kenya, with one site in each region being rural and the other being urban. The objective of the parent study was to uncover the burden of arboviruses amongst children in Kenya, studying both prevalence and transmission patterns. From this, travel information was collected, including whether or not someone traveled, and if so, where they traveled. Odds ratios were estimated from logistic regression models to evaluate whether incident dengue was associated with history of travel, region of travel, and distance traveled. Both crude and adjusted odds ratios of incident dengue were reported. Further analyses were done using ArcGIS software in order to gain spatial awareness of DENV serotypes, as well as to utilize hotspot analysis tools on subjects’ travel points in order to ascertain if there would be ‘hotspots’ for dengue based on travel location. Originally, we found an increased risk for dengue based on prior travel history (OR: 1.25, 95% CI: [1.01, 1.53]). However, after stratifying by study sites, this effect was not statistically significant. After further adjustment for study site, our results did not support this proposed dengue-travel association. This loss of statistical significance after adjustment may be due to the characteristics of our sample. It appeared as though participants living in or traveling to Western Kenya may be at increased risk of acute DENV infection, and our optimized hotspot analysis shows hotspots for DENV in travel points in Western Kenya. Yet in our cohort, the majority of those traveling are from Western Kenya, and the majority of those traveling to Western Kenya are from Western Kenya. After controlling for participant region, the associations between and travel to Western Kenya were also no longer statistically significant. However, there is evidence of increased serotype diversity and dengue coinfections in Western Kenya, indicating that potential public health interventions targeting dengue would be optimized in Western regions of the country. Further analyses are needed to better understand the mechanism behind the relationship between risk factors relating to travel and incident dengue.