Seroprevalence of chikungunya virus infection in India, 2017: a cross-sectional population-based serosurvey.

Background: Since its re-emergence in 2005, chikungunya virus (CHIKV) transmission has been documented in most Indian states. Information is scarce regarding the seroprevalence of CHIKV in India. We aimed to estimate the age-specific seroprevalence, force of infection (FOI), and proportion of the population susceptible to CHIKV infection.

Methods: We did a nationally representative, cross-sectional serosurvey, in which we randomly selected individuals in three age groups (5-8, 9-17, and 18-45 years), covering 240 clusters from 60 selected districts of 15 Indian states spread across all five geographical regions of India (north, northeast, east, south, and west). Age was the only inclusion criterion. We tested serum samples for IgG antibodies against CHIKV. We estimated the weighted age-group-specific seroprevalence of CHIKV infection for each region using the design weight (ie, the inverse of the overall probability of selection of state, district, village or ward, census enumeration block, and individual), adjusting for non-response. We constructed catalytic models to estimate the FOI and the proportion of the population susceptible to CHIKV in each region.

Findings: From June 19, 2017, to April 12, 2018, we enumerated 117 675 individuals, of whom 77 640 were in the age group of 5-45 years. Of 17 930 randomly selected individuals, 12 300 individuals participated and their samples were used for estimation of CHIKV seroprevalence. The overall prevalence of IgG antibodies against CHIKV in the study population was 18·1% (95% CI 14·2-22·6). The overall seroprevalence was 9·2% (5·4-15·1) among individuals aged 5-8 years, 14·0% (8·8-21·4) among individuals aged 9-17 years, and 21·6% (15·9-28·5) among individuals aged 18-45 years. The seroprevalence was lowest in the northeast region (0·3% [95% CI 0·1-0·8]) and highest in the southern region (43·1% [34·3-52·3]). There was a significant difference in seroprevalence between rural (11·5% [8·8-15·0]) and urban (40·2% [31·7-49·3]) areas (p<0·0001). The seroprevalence did not differ by sex (male 18·8% [95% CI 15·2-23·0] vs female 17·6% [13·2-23·1]; p=0·50). Heterogeneous FOI models suggested that the FOI was higher during 2003-07 in the southern and western region and 2013-17 in the northern region. FOI was lowest in the eastern and northeastern regions. The estimated proportion of the population susceptible to CHIKV in 2017 was lowest in the southern region (56·3%) and highest in the northeastern region (98·0%).

Interpretation: CHIKV transmission was higher in the southern, western, and northern regions of India than in the eastern and northeastern regions. However, a higher proportion of the population susceptible to CHIKV in the eastern and northeastern regions suggests a susceptibility of these regions to outbreaks in the future. Our survey findings will be useful in identifying appropriate target age groups and sites for setting up surveillance and for future CHIKV vaccine trials.

Funding: Indian Council of Medical Research.