Projecting and valuing climate change impacts on anxiety and depression in the contiguous USA: a damage function approach

Background

Links between climate change and mental health concerns in the USA are recognised, but research is underdeveloped. Many studies rely on self-reported exposures, non-representative samples, or inconsistent definitions of mental health outcomes, limiting their utility in supporting robust, population-scale projections to inform the public. Few studies have attempted to quantify future impacts, and those that do have largely focused on suicide rather than broader mental health morbidity. To address this gap, we projected the impact of 1–6°C contiguous US warming and the associated precipitation changes on self-reported mental health difficulties, anxiety, and depression symptom-days among adults relative to 2005 baseline climate and 2015 baseline health data.

Methods

We combined epidemiologically derived exposure–response relationships with projections from five CMIP6 climate models to estimate mental health impacts (mental health difficulties, anxiety, and depression symptom-days) in adults under present-day (2022) and end-of-century (2095) sociodemographic scenarios. We used data from the US Centers for Disease Control & Prevention Behavioral Risk Factor Surveillance System (BRFSS) to inform baseline symptom-day incidence rates (2013–23 datasets) and to estimate sex-specific and age group-specific mental health difficulty day allocation ratios for anxiety and depression (2018 Depression and Anxiety Module for Oregon and Tennessee). Analyses covered acute (short-term, over the past month) temperature and precipitation exposures by sex and income, and both acute and chronic (multi-year average maximum temperature) exposures for urban populations. Baseline symptom-day incidence rates were estimated using negative binomial regression analysis of BRFSS 2013–23 data, stratified by month, state, sex, and age group. Economic valuation was based on an original analysis of the Medical Expenditure Panel Survey 2018–21 data to develop fixed-effects regression-based estimates of health-related quality of life losses from anxiety and depression symptom-days. These daily losses are monetised using a scaled value per quality-adjusted life-year, calculated by dividing the United States Environmental Protection Agency Value of a Statistical Life by quality-adjusted life expectancy. We used Monte Carlo simulations to propagate uncertainty across health, climate, and valuation inputs.

Findings

Assuming present-day sociodemographics, warming of 1–6°C would result in 401 million to 1·8 billion and 329 million to 1·4 billion excess annual self-reported anxiety and depression symptom-days in adults, respectively, representing a 5–23% increase from baseline. Corresponding annual values of excess anxiety and depression burden are US$13 billion to $57 billion and $11 billion to $47 billion, respectively (2023 US dollars, undiscounted). We estimate an additional two to seven anxiety and one to six depression symptom-days per person-year, with larger burdens—four to 15 and three to 14 symptom-days, respectively—among low-income subpopulations. The greatest impacts are projected to occur in Appalachia. Using 2095 sociodemographics resulted in an increase in symptom days of almost 30% and an increase in monetised impacts of almost 90%.

Interpretation

These findings underscore the need for mental health investment in regions with restricted adaptive capacity due to economic hardship. Promoting individual and community resilience is crucial.