Abstract

Background Environmental heat exposure can negatively impact mental health. Evidence for its effect on mood disorder symptoms is inconsistent. Current studies are limited by poor temporal and geographical resolution. Methods We used ecological momentary assessment (EMA) data from the smartphone app juli to investigate the association between real-time mean and maximum ambient temperature collected from smartphone geolocation, and depressive and manic symptom scales, every two weeks, in adults with depression and bipolar disorder. We used negative binomial mixed-effects regression models, controlled for demographic and weather variables, and stratified by season. Results We analysed data from 4,000 participants with depressive symptom scores and 2,132 with manic symptom scores, between 2021 and 2023. We found that each 1{degrees}C increase in mean daily temperature in the preceding two weeks was associated with a 0.2% reduction in depressive symptom scores (coeff 0.998, 95%CI 0.997-0.999) and a 0.4% increase in manic symptom scores (coeff 1.004, 95%CI 1.001-1.007). Associations between maximum temperature and symptom scores followed a similar pattern. Limitations We were unable to capture several socio-demographic covariates, had limited geographical information due to privacy regulations, and included a non-random sample. Conclusions We found evidence that higher temperatures were associated with increased manic symptoms and decreased depressive symptoms, indicating an important relationship between temperature and the mood disorder continuum. With global heating, there is a need to understand the impact of temperature on mood symptoms, to provide targeted clinical prevention and support. This study demonstrates potential for EMA methods to inform our understanding of these links.