Vulnerability to Renal, Heat and Respiratory Hospitalizations During Extreme Heat Among U.S. Elderly
Published Date:Mar 7 2016
Source:Clim Change. 136(3):631-645.
Pubmed Central ID:PMC4956383
Funding:T42 OH008455/OH/NIOSH CDC HHS/United States
R00 ES015774/ES/NIEHS NIH HHS/United States
T32 AG027708/AG/NIA NIH HHS/United States
R21 ES024012/ES/NIEHS NIH HHS/United States
R18 EH000348/EH/NCEH CDC HHS/United States
R21 ES020695/ES/NIEHS NIH HHS/United States
R21 ES020156/ES/NIEHS NIH HHS/United States
K99 ES015774/ES/NIEHS NIH HHS/United States
Extreme heat (EH) is a growing concern with climate change, and protecting human health requires knowledge of vulnerability factors. We evaluated whether associations between EH (maximum temperature > 97th percentile) and hospitalization for renal, heat and respiratory diseases among people ≥ 65 years differed by individual and area-level characteristics.
We used Medicare billing records, airport weather data, U.S. Census data and satellite land cover imagery in 109 US cities, May-September, 1992–2006, in a time-stratified case-crossover design. Interaction terms between EH and individual (> 78 years, black race, sex) and home ZIP-code (percentages of non-green space, high school education, housing built before 1940) characteristics were incorporated in a single model. Next, we pooled city-specific effect estimates or regressed them on quartiles of air conditioning prevalence (ACP) in a multivariate random effects meta-analysis.
EH and combined renal/heat/respiratory hospitalization associations were stronger among blacks, the very old, in ZIP codes with lower educational attainment or older housing and in cities with lower ACP. For example, for EH versus non-heat days, we found a 15% (95% CI 11%–19%) increase in renal/heat/respiratory hospitalizations among individuals in ZIP codes with higher percent of older homes in contrast to a 9% (95% CI 6%–12%) increase in hospitalizations in ZIP codes with lower percent older homes.
Vulnerability to EH-associated hospitalization may be influenced by age, educational attainment, housing age and ACP.
Supporting Files:No Additional Files
You May Also Like: