BackgroundThe desiccation of the Great Salt Lake (UT, USA), driven by water use and diversion, has reduced water inflow and exposed vast areas of lakebed. Exposed lakebed can release airborne PM2·5, degrading air quality in surrounding communities. Previous research has established links between air pollution and mental health outcomes, but there is little research on the specific mental health effects of declining lake levels and associated dust exposure. We aimed to examine the association between Great Salt Lake water levels, PM2·5 concentration, and major depressive episodes in the surrounding population.MethodsIn this ecological study, we investigated relationships between Great Salt Lake decline, PM2·5 exposure, and mental health outcomes across all Utah counties using hydrological, atmospheric, and epidemiological datasets between 2006 and 2018. These data included in-situ lake measurements (from the US Geological Survey), PM2·5 from ground-based monitoring and reanalysis datasets (from MERRA-2 and AirNow network stations), the Social Vulnerability Index (from US Centers for Disease Control and Prevention), and records of major depressive episodes (from the Substance Abuse and Mental Health Services Administration). All datasets underwent quality control, variable preparation, and calculation of Z scores, and were harmonised both spatially and temporally. We calculated PM2·5 exceedance days based on US Environmental Protection Agency and WHO thresholds for harms, namely more than 15 μg/m3 and more than 35 μg/m3. Statistical analyses included Kruskal-Wallis tests for non-parametric group comparisons, followed by Dunn's post-hoc tests for pairwise comparisons. An ANOVA examined direct (main) and indirect (interaction) effects between factors, with F statistics measuring between-group versus within-group variance ratios.FindingsGreat Salt Lake shrinkage was linked to poorer air quality, with decreasing lake area (r=-0·28; p<0·0001) and volume (r=-0·31; p<0·0001) associated with more PM2·5 exceedance days. A dose-response relationship was found between depression severity and pollution exposure: individuals who had fewer high PM2·5 exceedance days had very low depression scores, whereas those exposed to more exceedance days had high and very high scores (H 28·9574; p<0·0001). This relationship showed nuanced differences across seasons (H 152·4771; p<0·0001) and age groups (H 51·8269; p<0·0001). Finally, our analysis showed a direct association between PM2·5 exceedance days and depression severity (F 12·341; p=0·0005), whereas social vulnerability acted as a significant moderator (F 6·979; p=0·0084). This interaction indicates that for a given level of PM2·5 exposure, socially vulnerable populations have a disproportionately higher prevalence of major depressive episodes.InterpretationThis study found associations between PM2·5 exposure and depressive episodes are concerning and warrant further longitudinal investigations, including more precise exposure measurements. Our findings underscore the complex interconnections between environmental degradation, air quality deterioration, and population mental health. By establishing this desiccated air pollution to mental health association, our results show that preventing lakebed exposure is not only an ecological necessity but a crucial public health intervention. Public health protections should focus on proactive watershed management alongside targeted mental health support for the most affected downwind and socially vulnerable communities.FundingUS National Aeronautics and Space Administration.