Premature Deaths Among Urban Residents with Psychosis: An Analysis of Contributing Environmental Factors

Background/Purpose: Psychosis prevalence varies according to geographical location: a substantial number of studies show a strong association between living in an urban setting, and psychosis onset (Vassos et al., 2012). Although various factors associated with urban living have been theorized to contribute to psychosis development, the specific mediating variables are unclear (Winz et al., 2022). This study aimed to investigate some of the most common theorized variables, such as heat, greenspace, poverty, and crowding using geographical information systems (GIS). Another variable that has not been considered among the others is light pollution. Since blue light exposure at night has been linked to various psychiatric disorders (Burns et al., 2023), light pollution was hypothesized to be one of the primary mediating effects on the relationship between urbanity and psychosis.

Methods: The 2023 National Violent Death Reporting System (NVDRS) dataset for Arizona was utilized, and the subset of deaths that were psychosis-related were found through a text keyword search of officer case reports. The keywords included “psychosis”, “psychoses”, “schizophren*”, “schizo*”, “psychotic”, “delusion*”, and “hallucinat*”. Independent variables included in the analysis were population density, percent below the poverty line, urban designation by census (yes/no), proximity to greenspace, and light pollution. Population density, percent in poverty, and urban designation were gathered from the American Community Survey (ACS) data from the U.S. Census Bureau. Greenspace was determined through cleaning the Protected Areas Database (PAD-US) 3.0 dataset. Light pollution data was gathered from the Visible infrared Imaging Radiometer Suite (VIIRS) Day/Night Band (DNB) from the Earth Observations Group.
Univariate and bivariate tests were conducted for spatial autocorrelation using the Moran’s I to examine the degree and significance of spatial effects. All Moran’s I analyses were performed using ESRI’s ArcGIS with an inverse distance spatial weights matrix. Additionally, a geographically weighted Poisson regression between psychosis deaths and the independent variables was conducted using the OLS modeling tool in ArcGIS (see Table 1).

Results: With inclusion of the weights matrix, the multiple regression analysis showed that crowding (B = 0.001, p = .002), percent below the poverty line (B = 0.003, p = .046), and light pollution (B = 0.004, p < .001) were statistically significant predictors of higher counts of psychosis-related deaths. Greenspace proximity was not statistically significant.

Conclusions and Implications: In summary, light pollution is a potential risk factor for psychosis, and merits further investigation. Blue light exposure can have adverse effects on mania and other symptomatology. As the primary providers of mental healthcare for people with psychosis, social workers can gain insight from geographic trends in mental health. While clinical and individual factors must be considered in the biopsychosocial model of care, the environment is equally crucial in treatment planning, particularly in considering preventative care for people at risk of psychosis. Interestingly poverty and crowding were also risk factors for premature deaths due to psychosis. The findings highlight the significance of urban living space equity for prevention of mental health issues and suggest specific reasons urban dwellers are more likely to have psychotic episodes.