Rotavirus infections in a community based cohort in Vellore, India.

INTRODUCTION The burden of infection in communities determines the spread of rotavirus infection and disease in susceptible populations. This study reports rotavirus infection and disease in a community based birth cohort in Vellore. METHODS Bimonthly surveillance and diarrheal stool were collected from 452 children enrolled at birth, of whom 373 completed three years of follow up. Samples were screened for rotavirus by an ELISA and genotyped by reverse transcription polymerase chain reaction for VP7 and VP4 genes. Rotavirus incidence rates were calculated using Poisson regression equations. Risk factors associated with symptomatic and asymptomatic rotavirus infections were compared using multiple logistic regression. RESULTS A total of 1149 episodes of rotavirus infections occurred in 94.4% children in the cohort. Incidence of rotavirus infection was 1.04 (0.97-1.1) per child-year with 0.75 asymptomatic and 0.29 symptomatic infections per child-year. About 18% of the children were infected in the first month, mainly with the G10P[11] strain. Rotavirus infections were more prevalent during October-March, but seasonality was not as marked in rotavirus disease. Rotavirus was associated with 15.1% of mild diarrhea, 38.9% of moderate/severe diarrhea and 66.7% of very severe diarrhea. Four common G types - G1 (26.8%), G2 (16%), G10 (11.2%) and G9 (9.6%) were seen, with high rates of mixed infections and untypable samples. Male gender, presence of siblings and low maternal education were associated with rotavirus disease. CONCLUSION This study demonstrates that rotavirus is the most common cause of gastroenteritis in the community, and indicates that since rotavirus caused the greatest proportion of moderate and severe disease, targeted interventions such as vaccines are needed for rotavirus, in addition to health education, sanitation and appropriate treatment to decrease diarrheal disease in communities.