Social Vulnerability and Prevalence and Treatment for Mental Health and Substance Use Disorders
Robert D Gibbons,Mark Olfson,Loren Saulsberry,Mark J Edlund,Sahar Zangeneh,Natalie Bareis,Lydia Chwastiak,Jason B Gibbons,Ronald C Kessler
DOI: https://doi.org/10.1001/jamapsychiatry.2024.1870
2024-10-01
JAMA Psychiatry
Abstract:Importance: Community-level social vulnerability (SV) is associated with physical illness and premature mortality. Its association with mental health (MH) and substance use disorders (SUDs) needs further study. Objective: To study associations of SV with clinical diagnoses of MH disorders, SUDs, and related treatments in the US noninstitutionalized population of adults aged 18 years and older. Design, setting, and participants: A survey of adults in a national sample of US households between October 2020 and October 2022. Participants drawn from a multistage, clustered, and stratified area probability sample of US households were included, excluding adults older than 65 years because of the difficulty of differentiating mental disorders from symptoms of dementia. The sample also included adults living in prisons, state psychiatric hospitals, and homeless shelters who were excluded from the sample of US households used in these analyses. Each sample household was sent a letter explaining the study and offering the option to complete the household roster online, by phone, or by email. Of the 12 906 adults selected for clinical interviewing in the household sample, 4674 completed clinical interviews. Main outcomes and measures: Main outcomes were Structured Clinical Interview for DSM-5 past-year diagnoses of MH disorders and SUDs and responses to survey questions regarding treatment received. The Social Vulnerability Metric (SVM) and the Area Deprivation Index (ADI) were used to determine SV at the residential zip code level. Results: The analysis involved 4674 participants (2904 [62.13%] female and 1770 [37.87%] male; mean [SD] age, 41.51 [13.41] years). Controlling for measured confounders, the SVM was significantly associated with diagnoses of schizophrenia spectrum disorder (SSD; adjusted odds ratio [aOR], 17.22; 95% CI, 3.05-97.29), opioid use disorder (OUD; aOR, 9.47; 95% CI, 2.30-39.02), stimulant use disorder (aOR, 6.60; 95% CI, 2.01-21.67), bipolar I disorder (aOR, 2.39; 95% CI, 1.19-4.80), posttraumatic stress disorder (aOR, 1.63; 95% CI, 1.06-2.50), and any MH disorder (aOR, 1.44; 95% CI, 1.14-1.83), but not major depressive disorder (MDD), generalized anxiety disorder (GAD), or any SUD. Results were similar for the ADI but generally of lower magnitude (SSD aOR, 11.38; 95% CI, 1.61-80.58; OUD aOR, 2.05; 95% CI, 0.30-14.10; stimulant use disorder aOR, 2.18; 95% CI, 0.52-9.18). Among participants with SSDs, SV was associated with reduced MH treatment (aOR, 0.001; 95% CI, 0.00-0.18) and reduced SUD treatment in participants with OUD or stimulant use disorder (aOR, 0.24; 95% CI, 0.02-2.80). Conclusions and relevance: In contrast to previous studies using nonclinical symptom-based survey data, we found no association between SV and GAD or MDD. By contrast, there were associations of SV with prevalence of SSD, stimulant use disorder, and OUD with corresponding decreases in treatment. These results suggest that the SVM might assist in developing more comprehensive care models that integrate medical and social care for MH disorders and SUDs.