RESUMO
BACKGROUND: Legionella species are abundant in the environment and are increasingly recognized as a cause of severe pneumonia. Increases in cases of community-acquired legionellosis in the greater Philadelphia metropolitan area (GPMA) led to concern that changing environmental factors could influence occurrence of disease. METHODS: We evaluated the association between weather patterns and occurrence of legionellosis in the GPMA, using both traditional Poisson regression analysis and a case-crossover study approach. The latter approach controls for seasonal factors that could confound the relationship between weather and occurrence of disease and permits the identification of acute weather patterns associated with disease. RESULTS: A total of 240 cases of legionellosis were reported between 1995 and 2003. Cases occurred with striking summertime seasonality. Occurrence of cases was associated with monthly average temperature (incidence rate ratio [IRR] per degree Celsius, 1.07 [95% confidence interval [CI], 1.05-1.09]) and relative humidity (IRR per 1% increase in relative humidity, 1.09 [95% CI, 1.06-1.12]) by Poisson regression analysis. However, case-crossover analysis identified an acute association with precipitation (odds ratio [OR], 2.48 [95% CI, 1.30-3.12]) and increased humidity (OR per 1% increase in relative humidity, 1.08 [95% CI, 1.05-1.11]) 6-10 days before occurrence of cases. A significant dose-response relationship for occurrence of cases was seen with both precipitation and increased humidity. CONCLUSIONS: Although, in the GPMA, legionellosis occurred predominantly during summertime, the acute occurrence of disease is best predicted by wet, humid weather. This finding is consistent with the current understanding of the ecological profile of this pathogen and supports the contention that sporadic legionellosis occurs through contamination of water sources.
