A time-series study of the association of rainfall, relative humidity and ambient temperature with hospitalizations for rotavirus and norovirus infection among children in Hong Kong.

BACKGROUND

Rotavirus and norovirus are infectious pathogens primarily affecting children under 5 years old. The impact of rainfall on diarrheal diseases remains inconclusive. This study aimed to evaluate the association between short-term variation in rainfall, temperature and humidity, and rotavirus and norovirus hospitalizations among young children in Hong Kong.

METHODS

Generalized additive negative binomial regression models with distributed lag non-linear terms, were fit with daily counts of hospital admissions due to rotavirus and norovirus infection as the outcomes and daily total rainfall and other meteorological variables as predictors, adjusting for seasonality and trend.

RESULTS

Generally, greater rainfall was associated with fewer rotavirus, but more norovirus hospitalizations. Extreme precipitation (99.5 mm, 99th percentile) was found to be associated with 0.40 (95% confidence interval (CI) 0.20-0.79) and 1.93 (95% CI 1.21-3.09) times the risk of hospitalization due to rotavirus and norovirus infection respectively, relative to trace rainfall. Stronger associations were observed in winter for rotavirus and in summer for norovirus. The duration of association with rotavirus was notably longer than norovirus. Higher temperatures were found to be associated with fewer hospitalizations for both rotavirus and norovirus infection, while higher relative humidity was generally associated with more norovirus, but fewer rotavirus, hospitalizations.

CONCLUSIONS

Both rotavirus and norovirus hospitalizations were strongly associated with recent precipitation variation but in opposite directions. With the introduction of the rotavirus vaccine norovirus is likely to become a greater threat than rotavirus and thus greater precipitation may become more clearly associated with more childhood diarrhea.