High seroprevalence of antibodies to Dengue, Chikungunya, and Zika viruses in Dire Dawa, Ethiopia: A cross-sectional survey in 2024

Abstract

Background: Aedes-borne diseases infect millions of people each year. In the last decade several arbovirus outbreaks have been reported in Ethiopia. Arbovirus diagnosis and surveillance is lacking, and the true burden is unknown.

Methods: In this study we conducted a seroprevalence survey using a commercially available test kit that tests for immunological responses to IgM and IgG for dengue (DENV), Zika (ZIKV), and chikungunya (CHIKV) viruses in Dire Dawa city, eastern Ethiopia. A total of 339 individuals were sampled using a household-based clustered design. As a contextual comparison, a secondary survey was conducted among 180 individuals in Addis Ababa, where no Aedes-borne virus outbreaks have been reported.

Findings: We found a high IgG seroprevalence for DENV (76%), CHIKV (44%), and ZIKV (38%), and <20% IgM seropositivity across all viruses. In contrast, minimal seropositivity was detected in Addis Ababa (where the highest seropositivity we found was to IgM for DENV at approximately 3%.) Age-specific trends showed early and widespread DENV exposure, with over half of the population seropositive by age 10. Quantitative antibody levels indicated strong correlation between DENV and ZIKV IgG, suggesting potential cross-reactivity. However, higher DENV IgG titers among ZIKV-positive individuals raise the possibility of true prior co-exposure. Intraclass correlation analyses revealed household-level clustering for DENV and CHIKV responses but not for ZIKV.

Conclusions: These results suggest intense and possibly ongoing transmission of Aedes-borne viruses in Dire Dawa, particularly dengue and chikungunya. Apparent ZIKV exposure warrants cautious interpretation given the potential for cross-reactivity, but cannot be ruled out. Our findings underscore the need for improved arbovirus surveillance and diagnostic capacity in Ethiopia, especially in urban centers where competent vectors are established.

Fig 1. Age-specific seroprevalence of IgG and IgM antibodies to dengue, Zika, and chikungunya viruses in Dire Dawa, Ethiopia (March 2024).

Bars represent the proportion of individuals in each age group who tested positive for each virus. Error bars indicate 95% confidence intervals calculated using the Wilson score method. IgG results reflect past exposure, while IgM results suggest more recent or acute exposure. A total of 339 individuals were included in the survey. Fig A in S1 Text presents age specific estimates with smaller age groups.

Fig 2. Predicted probability of IgG seropositivity by age based on GEE models (Dire Dawa, March 2024).

Predicted probabilities of seropositivity for dengue, Zika, and chikungunya IgG are shown as smoothed curves generated from generalized estimating equations (GEEs), incorporating age as a natural spline (df = 3) and adjusting for household-level clustering. Shaded areas represent 95% confidence intervals. Estimates reflect modeled age-specific trends in seropositivity, rather than crude age group proportions.

Fig 3. Spearman correlation matrix of quantitative IgG and IgM levels for each arbovirus.