Environment/Climate Change
Impact of extreme heat on mortality in a medically vulnerable population of kidney transplant recipients: a national cohort study Ting Lu, MS* Ting Lu Lu Department of Surgery, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA
Background
Extreme heat increases mortality risk in the general population, but data for vulnerable groups like kidney transplant (KT) recipients who are immunosuppressed with high comorbidity burdens are limited. We evaluated associations between weekly ambient heat index (HI) and post-transplant mortality and death-censored graft failure. We used Mundlak decomposition to separately examine the impact of short-term HI deviations from average norms versus longer-term place-based differences in typical heat exposure.
Methods
We studied 286,522 adult KT recipients (USRDS, 2000–2021) linking to recipients’ time-varying residence. Daily ZCTA-level HI was computed from PRISM 800-m gridded climate inputs using spatially area-weighted and summarized as weekly mean HI. We fit negative binomial mixed models with ZCTA random effect, adjusting for calendar trends and seasonality. We compared: (i) single-term weekly HI models and (ii) Mundlak decomposition separating within-ZCTA weekly deviations from between-ZCTA season-specific mean HI differences. Effects are reported per 1°C HI increase.
Results
Analyses included 17,613,719 ZCTA-weeks (86,001 deaths; 58,828 graft failures) full-season and 7,383,358 ZCTA-weeks (32,914 deaths; 22,786 graft failures) warm-season across 23,379 ZCTAs. Single-term models showed higher weekly HI linked to mortality and graft failure, strongest in warm season (mortality IRR=1.009 [1.007–1.011], p<0.001). Mundlak models showed no association with short-term within-ZCTA deviations; however, between-ZCTA season-specific mean HI showed consistent positive associations, stronger in warm season (mortality IRR=1.015 [1.012–1.018], p<0.001).
Conclusions
Higher weekly heat exposure increases mortality and graft failure risk, particularly in warm seasons. Long-term place-based heat differences dominate associations, suggesting a need for local temperature context in risk assessment and targeted heat-preparedness for KT recipients during warmer months.
