Background: Pneumonia is a common cause of childhood hospitalization. Both host and pathogen factors are associated with environmental conditions, but the associations between childhood pneumonia and meteorological variables are unclear. This study investigated the short-term associations between childhood pneumonia admissions and meteorological variables.
Methods: A retrospective time-series analysis was conducted using distributed lagged nonlinear models. Daily pneumonia admissions among children <15-year-old in Hong Kong during 2004-2011 were regressed on daily meteorological variables with air pollutants, influenza admissions, seasonal trend, and long-term trend adjusted. Analyses were stratified by age group.
Results: There were 34,303 admissions during 2004-2011. Temperature had a u-shaped association with childhood pneumonia admissions with minimum morbidity temperature at 25°C and with a long lagged effect up to 45 days. The cumulative relative risk (cum RR [95% confidence interval]) (vs. 25°C; lag 0-45 days) at 30°C was 1.41 (1.11, 1.79) and was 1.58 (1.21, 2.06) at 12°C. Relative humidity (RH) showed a u-shaped association with minimum risk at 65% and lagged effect up to 45 days. The cum RR (vs. 65%; lag 0-45 days) at 53% was 1.26 (1.04, 1.54) and was 2.22 (1.78, 2.77) at 94%. Children 5- to 14-year-olds were highly sensitive to temperature and RH while admissions among children <5-year-old were only associated with high RH.
Conclusions: Childhood pneumonia admissions were very strongly associated with both high and low temperature and RH for children 5-14-year-olds. Efforts to reduce exposure of children to extreme temperatures and RH may have the potential to reduce the burden of pediatric pneumonia.