RESUMEN
Ross River virus (RRV) is the most important vector-borne disease in Australia. The National Notifiable Diseases Surveillance System has confirmed that its incidence is often greatest in the state of Queensland, where there is a clear seasonal pattern as well as interannual variability. Previous studies have examined relationships between large-scale climate fluctuations (such as El Niño Southern Oscillation) and vector-borne disease. No previous study has examined such relationships with the Quasi-Biennial Oscillation (QBO), another large-scale climate fluctuation. We employ time-series analysis techniques to investigate cycles inherent in monthly RRV incidence in Queensland, Australia, from January 1991 to December 1997 inclusive. The presence of a quasi-biennial cycle in the RRV time series that is out of phase with the climatic QBO is described. Quantitative analyses using correlograms and periodograms demonstrate that the quasi-biennial cycle in the RRV time series is statistically significant, at the 95% level, above the noise. Together with the seasonal cycle, the quasi-biennial cycle accounts for 77% of the variance in Queensland RRV cases. Regression analysis of QBO and summer rainfall in three climatic zones of Queensland indicates a significant association between QBO and rainfall in the subtropical southeastern part of the state. These results suggest an indirect influence of the QBO on RRV incidence in Queensland, via its influence on climate in this region. Our findings indicate that the QBO may be a useful predictor of RRV at several months lead, and might be used by public health authorities in the management and prevention of this disease.