Air Conditioning Use and Ambulance Transports for Heat-Related Illness Across Summer Months in the Tokyo Metropolitan Area.

Heat is a growing public health threat, yet seasonal variations in adaptation including air conditioning (AC) use remain poorly understood. We examined the modifying role of electricity consumption (EC), a proxy for air conditioning (AC) use, in the association between maximum temperature and ambulance transports for heat-related illness in the Tokyo Metropolitan Area. Daily ambulance transport data (June-September, 2015-2019), alongside maximum temperature and EC data, were analyzed across seven prefectures. Using a two-stage approach, we fitted quasi-Poisson regression models to estimate relative risks (RRs) per 1 °C increase in maximum temperature under high and low EC conditions, and pooled prefecture-month-specific estimates through multivariate meta-analysis. Attributable fractions (AFs) and attributable numbers (ANs) were also calculated. A total of 88,893 transports were analyzed. In June, higher risks were observed under high EC (RR = 1.37, 95%CI: 1.32-1.42) compared with low EC (RR = 1.28, 95%CI: 1.25-1.32). In July, the pattern reversed, with lower risks under high EC (RR = 1.33, 95%CI: 1.30-1.36) than low EC (RR = 1.39, 95%CI: 1.35-1.44). EC modified the association between heat and ambulance transport in both early summer months (p = 0.01 and 0.03, respectively), but not in August or September. Across months and EC strata, heat-attributable fractions ranged from 22.1% to 34.0%. The higher risks under high EC in June may have reflected delayed or insufficient air conditioning use, whereas the lower risks in July are consistent with a possible protective role of AC emerging as the season progresses. These findings underscore the need for early season adaptation measures.