Heat stroke related hospitalizations in older adults: disproportionally high impact of the first heat wave

Older adults are vulnerable to extreme temperature events, which have been documented to be increasing in intensity and frequency. As the US continues on its aging trajectory, it is crucial to characterize the effects of extreme temperature events on older adults' hospitalizations, so that we can optimize preventive efforts.. This study examines the effects of high ambient temperature on hospitalizations directly related to heat exposure among older adults aged 65 or above, and to compare the effects caused by the first and subsequent heat waves. We abstracted all 701 cases of hospitalizations due to heat stroke (HSH), defined as ICD-9-CM 992.0 – 992.9, occurred in older adults living in Boston Metropolitan Statistical Area (MA-NH MSA) between January 1st 1991 and December 31st 2006 from the Centers for Medicare and Medicaid Services databases. Daily temperature records of the area were obtained from the National Oceanographic and Atmospheric Administration for the study period. The dataset includes daily maximal and minimal temperature from over 50 stations within 120 miles radius from the area's centroid. The daily temperature data were interpolated for each ZIP code using an inverse distance weighting method, which allows for multivariate interpolation by assigning the values to unknown locations calculated with a weighted average of the values available at the known points. To characterize temporal variations of daily counts of HSH and their relationships with maximum and minimum ambient temperatures a time series study was conducted using Poisson regression applied. The seasonality of HSH and temperatures was through harmonic regression with four terms to account for two annual peaks. The peak timing for HSH and daytime temperature were estimated using d–method. The non-linear effect of high temperatures was captured by a specially designed function allowing for exponential growth above thresholds of 95th percentiles for daily minimum and maximum temperatures to reflect the peak of daytime and nighttime exposure to heat. The heat wave periods in the Boston MSA was defined as 3 consecutive days with daily maximum temperature above the threshold. This definition of a heat wave offered the highest explanatory power (as compared to five other commonly used definitions) and was selected for the final model. The impacts of the first and subsequent heat waves on HSH were investigated. The regression model with the temporal elements: long-term trend, effects of individual years, seasonality, and the effects of weekday and holidays) explained 30.8% variability in the HSH. On average HSH peaked on July 8th with median (IQR) of 190 (185.5; 198) days, 15 days earlier than the peak in maximum temperature, observed on July 24th with median (IQR) of 205 (203; 206.6) days. The regression model including a binary indicator for heat wave episodes and adjustments for temporal features explained 43.5% variability in HSH. The relative risk of HSH associated with a heat wave episode was 9.119[95%CI: 7.584 - 10.964]. While the average maximum and minimum temperature values were relatively stable across 36 heat wave periods, however the rate of HSH occurred within the first heat wave period was twice higher as compared to fourth wave (9.64±11.62 vs 4.20±2.79 HSH/per heat wave episode, respectively). Similarly, a number of HSH per day significantly declined from 2.7±4.13 during 14 first heat waves to 1.05±1.68 during 5 observed fourth heat wave. The estimated annual peaks of HSH closely correlate with the occurrence of the first heat wave each year (r=0.557, p<0.038). Based on a heat wave definition generated for this study, we found that a heat wave resulted in ten-fold increase in hospitalizations due to heat stroke among the older adults, residing in temperate climate of Boston, MA. In a given season, the first heat wave causes more severe health impact than the subsequent heat waves. We advocate for increasing the awareness of detrimental health effects of heat exposure in older adults and for systematic improvement of living conditions, infrastructure and medical support for most vulnerable population to ensure their wellbeing and reduce the cost of health care.