The Indian Ocean Dipole and Cholera Incidence in Bangladesh: A Time-Series Analysis

BACKGROUND: It has been reported that the El Nino-Southern Oscillation (ENSO) influences the interannual variation of endemic cholera in Bangladesh. There is increased interest in the influence of the Indian Ocean dipole (IOD), a climate mode of coupled ocean-atmosphere variability, on regional ocean climate in the Bay of Bengal and on Indian monsoon rainfall. OBJECTIVES: We explored the relationship between the IOD and the number of cholera patients in Bangladesh, controlling for the effects of ENSO. METHODS: Time-series regression was performed. Negative binomial models were used to estimate associations between the monthly number of hospital visits for cholera in Dhaka and Matlab (1993-2007) and the dipole mode index (DMI) controlling for ENSO index [NINO3, a measure of the average sea surface temperature (SST) in the Nino 3 region], seasonal, and interannual variations. Associations between cholera cases and SST and sea surface height (SSH) of the northern Bay of Bengal were also examined. RESULTS: A 0.1-unit increase in average DMI during the current month through 3 months before was associated with an increase in cholera incidence of 2.6% [(95% confidence interval (CI), 0.0-5.2; p = 0.05] in Dhaka and 6.9% (95% CI, 3.2-10.8; p < 0.01) in Matlab. Cholera incidence in Dhaka increased by 2.4% (95% CI, 0.0-5.0; p = 0.06) after a 0.1-unit decrease in DMI 4-7 months before. Hospital visits for cholera in both areas were positively associated with SST 0-3 months before, after adjusting for SSH (p < 0.01). CONCLUSIONS: These findings suggest that both negative and positive dipole events are associated with an increased incidence of cholera in Bangladesh with varying time lags.