Geographical variation in the impact of extreme heat events on respiratory hospitalizations among older adults in 120 U.S. metropolitan areas

Abstract ISEE – 2023 (Taiwan) Topic: Climate Subtopic: Temperature Type: Oral Presentation   Authors: Cassandra R. O’Lenick1, Mallory Turner2, William Steinhardt2, Stephanie Cleland1,2, Lloyd Hill2, Lucas Neas2, Meghan E. Rebuli1, Ilona Jaspers1, Ana G. Rappold2,1   University of North Carolina at Chapel Hill, Chapel Hill, NC United States Environmental Protection Agency, Chapel Hill, NC   Abstract title (20 word max): Geographical variation in the impact of extreme heat events on respiratory hospitalizations among older adults in 120 U.S. metropolitan areas   Abstract  Objectives: Extreme heat exposure is a well-known cause of mortality, particularly among older adults. However, evidence of the effect of extreme heat on morbidity is inconsistent. We applied a uniform modeling approach across 120 of the largest US metropolitan areas to assess heat-related hospitalizations for respiratory disease among older adults. Methods. We obtained ZIP code-level daily counts of respiratory disease hospitalizations using Medicare billing claims from short-stay, inpatient hospitalizations. We estimated ambient heat exposure using spatially interpolated, population weighted weather station data. To estimate heat-health impacts during the warm season (June-Sept), we explored lag days 0-6, and used conditional Poisson with a distributed lag non-linear model, matched on ZIP code of residence, month, and day of hospitalization. We fit separate models for three heat exposure metrics; all models were adjusted for temporal trends and meteorology. Results. From 2000 to 2017, we identified 9,141,138 respiratory disease hospitalizations during the warm season. Associations between respiratory hospitalizations and extreme heat days - defined as the 95th percentile of the city-specific warm season temperature - exhibited substantial geographic variation.  When comparing the 95th percentile to the 50th percentile of temperature, cumulative odds ratios ranged from 1.09 (95% CI: 1.03, 1.15) in Palm Bay, Florida to 0.95 (95% CI: 0.89, 1.00) in Boise City, Idaho. Geographical variation was not explained by U.S. region/division or Köppen classification of climate zones. Associations were generally larger on lag days 0-2 compared to later lag periods. Conclusion. Considerable geographical variation in the magnitude and direction of odds ratios suggest that drivers of heat-health impacts are occurring at the local level. Future work should investigate area-level factors that modify risk. This work was supported by a cooperative agreement between the U.S. EPA and University of North Carolina-Chapel Hill. The abstract does not necessarily reflect the views and policies of the U.S. EPA.   Key words: Climate, Temperature, Built environment, Temperature extremes, Respiratory outcomes, Environmental epidemiology

Impact/Purpose

Objectives: Extreme heat exposure is a well-known cause of mortality, particularly among older adults. However, evidence of the effect of extreme heat on morbidity is inconsistent. We applied a uniform modeling approach across 120 of the largest US metropolitan areas to assess heat-related hospitalizations for respiratory disease among older adults. Methods. We obtained ZIP code-level daily counts of respiratory disease hospitalizations using Medicare billing claims from short-stay, inpatient hospitalizations. We estimated ambient heat exposure using spatially interpolated, population weighted weather station data. To estimate heat-health impacts during the warm season (June-Sept), we explored lag days 0-6, and used conditional Poisson with a distributed lag non-linear model, matched on ZIP code of residence, month, and day of hospitalization. We fit separate models for three heat exposure metrics; all models were adjusted for temporal trends and meteorology. Results. From 2000 to 2017, we identified 9,141,138 respiratory disease hospitalizations during the warm season. Associations between respiratory hospitalizations and extreme heat days - defined as the 95th percentile of the city-specific warm season temperature - exhibited substantial geographic variation.  When comparing the 95th percentile to the 50th percentile of temperature, cumulative odds ratios ranged from 1.09 (95% CI: 1.03, 1.15) in Palm Bay, Florida to 0.95 (95% CI: 0.89, 1.00) in Boise City, Idaho. Geographical variation was not explained by U.S. region/division or Köppen classification of climate zones. Associations were generally larger on lag days 0-2 compared to later lag periods. Conclusion. Considerable geographical variation in the magnitude and direction of odds ratios suggest that drivers of heat-health impacts are occurring at the local level. Future work should investigate area-level factors that modify risk. This work was supported by a cooperative agreement between the U.S. EPA and University of North Carolina-Chapel Hill. The abstract does not necessarily reflect the views and policies of the U.S. EPA.

Citation

O'Lenick, C., M. Turner, W. Steinhardt, S. Cleland, K. Hill, L. Neas, M. Rebuli, I. jaspers, AND A. Rappold. Geographical variation in the impact of extreme heat events on respiratory hospitalizations among older adults in 120 U.S. metropolitan areas. 35th Annual Conference of the International Society for Environmental Epidemiology, Kaohsiung, TAIWAN, PROVINCE OF CHINA, September 17 - 21, 2023.