Social Mixing Patterns and Chikungunya Reemergence Risk in French Polynesia.

Importance: Chikungunya virus has reemerged in some French overseas territories and mainland France, raising concern for its reintroduction in French Polynesia. Objective: To assess current population immunity to chikungunya virus and estimate future outbreak risk. Design, Setting, and Participants: This cross-sectional study included data from 2 serosurveys and used a modeling approach that integrated seroprevalence, contact matrices, and demographic characteristics. The first serosurvey was conducted on the island of Tahiti (June 2018) among schoolchildren aged 6 to 16 years. The second serosurvey (November 2019 to December 2021) included adults aged 18 to 69 years across French Polynesia, a French overseas territory in the Southeast Pacific that comprises 75 inhabited islands grouped into 5 subdivisions with 279 000 people according to the 2022 census. Data analysis was conducted from April to June 2025. Main Outcomes and Measures: Main outcomes were seroprevalence rates and factors associated with seropositivity; age-stratified infection proportions for the 2014 to 2015 outbreak, pre-outbreak the basic reproduction number (R0), the effective reproduction number (Reff) for 2025, and Reff projections were estimated. Results: Serological data were available for 2363 participants (457 schoolchildren; median [IQR] age, 11 [8-13] years; 244 [53.3%] female, and 1906 adults; 38 [28-52] years; 1003 [52.6%] female). Seroprevalence was 62.8% (95% CI, 58.0%-67.0%) in schoolchildren, and the weighted seroprevalence was 67.6% (95% CI, 64.8%-70.3%) in adults. In adults, lower odds of seropositivity were associated with male sex (adjusted odds ratio [aOR], 0.71; 95% CI, 0.55-0.92), age groups 30 to 44 years and 45 to 69 years compared with 18 to 29 years (ages 30-44 years: aOR, 0.46; 95% CI, 0.32-0.66; ages 45-69 years: aOR, 0.33; 95% CI, 0.22-0.48), living in the Austral subdivision (aOR, 0.38; 95% CI, 0.27-0.55), and higher education (ie, university or more education: aOR, 0.51; 95% CI, 0.32-0.81). Living in larger households and being unpartnered were associated with higher odds of seropositivity (larger households: aOR, 1.69; 95% CI, 1.12-2.56; being unpartnered: aOR, 1.95; 95% CI, 1.03-3.69). Modeled infection rates for 2014 to 2015 ranged from 52.6% (95% credible interval [CrI], 43.6%-60.0%) in children aged 0 to 4 years to 80.8% (95% CrI, 73.3%-86.2%) in those aged 10 to 14 years. Inferred pre-outbreak R0 was 1.78 (95% CrI, 1.73-1.82), and Reff for 2025 was estimated at 0.95 (95% CrI, 0.85-1.04). Assuming assortative mixing, projections suggest Reff would exceed 1 (lower 95% CrI >1) by 2028, but assuming random mixing, Reff would exceed 1 in 2051, more than 20 years later than the other estimate. Conclusions and Relevance: In this cross-sectional study, high chikungunya seroprevalence was observed, with heterogeneity between subdivisions, age groups, and other sociodemographic characteristics. Modeling analysis estimated Reff near 1 in 2025, suggesting sustained transmission could occur if chikungunya is reintroduced. These findings suggest that incorporating contact matrices could improve outbreak risk assessments, while assuming random mixing might underestimate reemergence timing.