Epidemiological characteristics and intervention-associated transmission dynamics of the largest chikungunya fever outbreak in China

<title>Abstract</title> <p>Background The largest documented chikungunya fever (CF) outbreak in China occurred in Foshan, yet evidence on outbreak-wide epidemiological characteristics and intervention-associated effects has remained limited. Methods We conducted a longitudinal analysis from 16 June to 10 August 2025, integrating case and vector epidemiology, and generalised estimating equation (GEE) models. Results A total of 8,964 cases were reported, with no deaths. Females had a higher incidence than males (1,027.12 vs 919.71 per 100,000), and males had lower infection odds (OR 0.89, 95% CI = 0.86–0.93). Adults older than 60 years showed the highest incidence (2371.62 vs. &lt;1402.00 per 100,000) and infection risk (OR &lt; 1 in other age groups). Cases were concentrated in Shunde District (82.42%). The onset-to-diagnosis interval shortened after interventions. 2.60% were asymptomatic; fever, rash and arthralgia were the predominant symptoms. The effective reproduction number (\(\:{R}_{t}\)) rose rapidly (3.12, 95% CI = 0.08–11.51) and declined below 1 by 1 August (0.87, 95% CI = 0.80–0.94). Breteau index (BI) and adult mosquito density Index (ADI) were decreased and high in the border areas between districts. For non-lagged GEE, one-unit changes in ADI and BI were associated with 9.28% (95% CI= -13.40–37.92; P = 0.45) and 0.60% (95% CI =-1.48–2.73; p = 0.57) changes in case numbers. One-unit changes in ADI were associated with a 10.73% change in \(\:{R}_{t}\) (95% CI = 3.23–18.78; p = 0.004), and 7.50% in BI (95% CI =-5.15–21.84; p = 0.26). For lagged GEE, lagged ADI remained associated with \(\:{R}_{t}\) (8.74%, 95% CI = 1.14–16.91; p = 0.023), while lagged tOD showed no significant association with case numbers (2.68%, 95% =CI -8.21 to 14.86; p = 0.64) and \(\:{R}_{t}\) (2.17%, 95% CI= -1.70 to 6.20; p = 0.28). Conclusions This outbreak showed high transmissibility, concentration in Shunde, and elevated risk in females and older adults. The key area for vector control was the border of different districts. Prioritising routine and targeted adult-mosquito surveillance and control, while maintaining larval source management and accelerating case detection, not only for this outbreak but also for other non-endemic urban settings facing comparable arboviral threats.</p>