Exposure patterns and the risk factors of Crimean Congo hemorrhagic fever virus amongst humans, livestock and selected wild animals at the human/livestock/wildlife interface in Isiolo County, upper eastern Kenya

Abstract

Crimean Congo hemorrhagic fever (CCHF) is a tick-borne zoonotic disease caused by CCHF virus (CCHFV). The disease has a complex transmission cycle that involves a wide range of hosts including mammalian and some species of birds. We implemented a sero-epidemiological study in Isiolo County, Kenya, to determine relative seroprevalences of CCHFV in humans, livestock and in wild animals. A seropositive herd was defined as having at least one seropositive animal. In addition, we identified subject and environment level factors that could promote exposure to CCHFV. Humans (n = 580) and livestock species (n = 2,137) were recruited into the study through a multistage random sampling technique, and in addition, various species of wild animals (n = 87) were also sampled conveniently. Serum samples from all recruited humans and animals were collected and screened for CCHFV antibodies using ID Screen multispecies, double-antigen IgG enzyme-linked immunosorbent assay (ELISA). The overall anti-CCHFV IgG seroprevalences in humans, cattle, goats, sheep and camels were 7.2% [95% CI: 3.1–15.8%], 53.9% [95% CI: 30.7–50.9%], 11.6% [95% CI: 7.2–22.5%], 8.6% [95% CI: 3–14%] and 89.7% [95% CI: 78–94%], respectively. On average, the sampled wild animals had CCHFV seroprevalence of 41.0% [95% CI: 29.1–49.4%]; giraffes had the highest mean CCHF seroprevalence followed by buffaloes, while impala had very low exposure levels. Statistical analyses using mixed effects logistic regression models showed that CCHFV exposure in humans was significantly associated with male gender, being over 30 years of age and belonging to a household with a seropositive herd. In livestock, a combination of animal- and environment level factors including older animals, high normalized difference vegetation indices (NDVI) and high vapour pressure deficit were significantly associated with CCHFV infection. Age, sex and species of wild animals were considered the key risk factors in the analysis, but none of these variables was significant (P-value = 0.891, 0.401 and 0.664, respectively). Additionally, RT-qPCR analysis revealed the presence of CCHFV RNA in camels (30%), cattle (14.3%), and goats (3.8%), but not in humans, sheep, or wild animals. This study demonstrates that environmental factors, such as NDVI and vapor pressure deficit, affect CCHFV exposure in livestock, while the presence of infected livestock is the key determinant of human exposure at the household level. These findings underscore the importance of using One Health approaches to control the disease in human-livestock-wildlife interfaces. For instance, the existing CCHF surveillance measures could be enhanced by incorporating algorithms that simulate disease risk based on the environmental factors identified in the study. Additionally, tick control in livestock, such as the use of acaricides, could reduce CCHFV exposure in livestock and, consequently, in humans.