Rotavirus infections and climate variability in Dhaka, Bangladesh: a time-series analysis

Abstract

SUMMARYAttempts to explain the clear seasonality of rotavirus infections have been made by relating disease incidence to climate factors; however, few studies have disentangled the effects of weather from other factors that might cause seasonality. We investigated the relationships between hospital visits for rotavirus diarrhoea and temperature, humidity and river level, in Dhaka, Bangladesh, using time-series analysis adjusting for other confounding seasonal factors. There was strong evidence for an increase in rotavirus diarrhoea at high temperatures, by 40·2% for each 1°C increase above a threshold (29°C). Relative humidity had a linear inverse relationship with the number of cases of rotavirus diarrhoea. River level, above a threshold (4·8 m), was associated with an increase in cases of rotavirus diarrhoea, by 5·5% per 10-cm river-level rise. Our findings provide evidence that factors associated with high temperature, low humidity and high river-level increase the incidence of rotavirus diarrhoea in Dhaka.