Yoshitaka Wakabayashi, Shin Nakayama, Ai Yamamoto, Takatoshi Kitazawa
Seasonal variations in bloodstream infections (BSIs) have been reported for several bacterial species. The independent effects of weather variables, such as humidity, temperature, precipitation, and bright sunshine duration have not been fully considered. We analyzed monthly BSI incidence rates at Teikyo University Hospital, Tokyo, from January 2014 through December 2024. Incidence was defined as the number of new BSI cases per 10,000 inpatient-days each month. Meteorological data were obtained from the Japan Meteorological Agency. Absolute humidity (AH) was residualized against temperature and calendar-month fixed effects (AHresid) to isolate the humidity component independent of temperature and seasonality. Monthly BSI counts were analyzed using negative binomial regression models with a log link and an offset for monthly inpatient-days to assess associations with weather variables. During the 11-year period, AHresid demonstrated a significant association with Escherichia coli (E. coli)-specific BSI incidence (IRR 1.11 [95% CI 1.01-1.21], p = 0.029). In contrast, other weather variables, such as temperature, precipitation and bright sunshine duration showed no association with all-cause and E. coli-specific BSIs. We found that rising humidity was associated with increased E. coli-specific BSI incidence. Because increments in temperature and humidity are both expected with global warming, continued surveillance of climate-related trends in bloodstream infections is warranted.