Ultra-short-term triggering of acute heart failure by heat following the monsoon rainy season: an hourly resolution case-crossover study in older adults

Background/Introduction Acute heart failure (AHF) is a leading cause of emergency care in older adults. Conventional daily temperature metrics may miss rapid heat-related physiological stress occurring within hours. In East Asia, the end of the rainy season marks an abrupt transition to hot, humid midsummer, potentially creating a short period of heightened vulnerability to acute cardiovascular decompensation. Purpose To quantify phase-specific nonlinear associations between ambient temperature and AHF and to identify the most relevant hourly lag windows during the early post-rainy period, when heat exposure may act as an immediate trigger. Methods We conducted a time-stratified case-crossover study of 1,107 emergency transports for physician-diagnosed AHF among residents aged >=65 years (2012-2019) in Japan. Control hours were matched by year, month, day of week, and hour. Seasonal phases were predefined as rainy season (RS) and early (EP), mid (MP), and late (LP) post-rainy months. Conditional logistic regression modelled temperature using a four-knot cubic spline (reference 25.8 C), adjusting for barometric pressure, relative humidity, and PM2.5. During EP, we additionally examined hourly moving-average windows (lag 0 to 0-11 h) and day-level windows to identify the most relevant exposure period. Exploratory subgroup analyses assessed effect modification by sex, age category, calendar strata, and year group. Results Exposure-response patterns were phase-specific, with the steepest increase and significant nonlinearity during EP (P for nonlinearity = 0.013), whereas associations in RS, MP, and LP were flatter or less consistent (Figure 1). In EP, AHF risk increased above about 28 C and remained elevated at higher temperatures. Hourly lag analyses showed the strongest associations from the event hour through the previous 3 hours (lag 0-3 h), with attenuation after 5-6 hours (Figure 1). At 35 C (lag 0), odds ratios (ORs) were 20.08 (95% CI 5.95-67.82) in men and 11.27 (3.20-39.74) in women. Day-level windows were consistent with elevated risks on the same day (0-23 h) and the following day (24-47 h). In sensitivity analyses, the Heat Index showed nonlinear amplification (P = 0.042), suggesting an added contribution of humidity. Exploratory subgroup analyses (Figure 2) showed broadly similar heat-related increases across sex, age, calendar strata, and year groups; ORs at 35 C were highest in adults aged >=85 years (19.95, 4.82-82.66), although confidence intervals were wide in some strata. Interaction tests did not show strong evidence of effect modification. Conclusion(s) Early post-rainy heat appears to be a rapid, nonlinear trigger of AHF within hours in older adults. These findings support sub-daily heat surveillance and targeted clinical preparedness during the monsoon-to-summer transition, particularly in ageing populations exposed to increasingly intense humid heat.