A small area analysis of acute exposure to temperatures and mental health in North Carolina

Increasing evidence suggests that temperatures adversely impact mental and behavioral disorders (MBD). This study explores the effects of temperatures on mental health outcomes using over 5.9 million MBD-related emergency department (ED) visits across three geographical regions of North Carolina (i.e., Mountains, Piedmont, and Coast) from 2016 to 2019. A distributed lag non-linear model (DLNM) with a generalized linear model and quasi-Poisson distribution adjusted for humidity, long-term seasonal time trends, and day of the week examined the acute impact (i.e., 7-day) of temperature on daily MBD-related ED visits at zip code tabulation area (ZCTA) locations. Results were pooled at the region and state levels and reported in reference to the median temperature using a case-time series design for the analysis of small-area data. Stratified analyses were conducted for age, sex, and specific mental-health related ED visits (substance use, mood disorders, anxiety disorders). At the state level, we found significant positive associations between high temperatures (97.5th percentile) and an increase in relative risk (RR) for total MBDs (RR:1.04, 95% CI,1.03-1.05) and psychoactive substance use (RR:1.04, 95% CI, 1.02-1.06). Low air temperatures (2.5th percentile) only increased risk for the elderly (i.e., 65 and older) and predominantly white communities (RR: 1.03, CI: 1.03-1.05). During high temperatures (97.5th percentile), majority-white communities (RR:1.06, CI: 1.01-1.10) and low-income communities had the highest risk for MBDs (RR: 1.05, CI: 1.03-1.07). Our findings suggest there is a positive association between exposure to high temperatures and increased MBD-related ED visits, modified by patient age and place-based sociodemographic (ie., race and income) context.