Associations of substance use, psychosis, and mortality among people living in precarious housing or homelessness: A longitudinal, community-based study in Vancouver, Canada
Andrea A. Jones,Kristina M. Gicas,Sam Seyedin,Taylor S. Willi,Olga Leonova,Fidel Vila-Rodriguez,Ric M. Procyshyn,Geoffrey N. Smith,Toby A. Schmitt,A. Talia Vertinsky,Tari Buchanan,Alex Rauscher,Donna J. Lang,G. William MacEwan,Viviane D. Lima,Julio S. G. Montaner,William J. Panenka,Alasdair M. Barr,Allen E. Thornton,William G. Honer
DOI: https://doi.org/10.1371/journal.pmed.1003172
2020-07-06
PLoS Medicine
Abstract:The "trimorbidity" of substance use disorder and mental and physical illness is associated with living in precarious housing or homelessness. The extent to which substance use increases risk of psychosis and both contribute to mortality needs investigation in longitudinal studies.A community-based sample of 437 adults (330 men, mean [SD] age 40.6 [11.2] years) living in Vancouver, Canada, completed baseline assessments between November 2008 and October 2015. Follow-up was monthly for a median 6.3 years (interquartile range 3.1–8.6). Use of tobacco, alcohol, cannabis, cocaine, methamphetamine, and opioids was assessed by interview and urine drug screen; severity of psychosis was also assessed. Mortality (up to November 15, 2018) was assessed from coroner's reports and hospital records. Using data from monthly visits (mean 9.8, SD 3.6) over the first year after study entry, mixed-effects logistic regression analysis examined relationships between risk factors and psychotic features. A past history of psychotic disorder was common (60.9%). Nonprescribed substance use included tobacco (89.0%), alcohol (77.5%), cocaine (73.2%), cannabis (72.8%), opioids (51.0%), and methamphetamine (46.5%). During the same year, 79.3% of participants reported psychotic features at least once. Greater risk was associated with number of days using methamphetamine (adjusted odds ratio [aOR] 1.14, 95% confidence interval [CI] 1.05–1.24, <em>p</em> = 0.001), alcohol (aOR 1.09, 95% CI 1.01–1.18, <em>p</em> = 0.04), and cannabis (aOR 1.08, 95% CI 1.02–1.14, <em>p</em> = 0.008), adjusted for demographic factors and history of past psychotic disorder. Greater exposure to concurrent month trauma was associated with increased odds of psychosis (adjusted model aOR 1.54, 95% CI 1.19–2.00, <em>p</em> = 0.001). There was no evidence for interactions or reverse associations between psychotic features and time-varying risk factors. During 2,481 total person years of observation, 79 participants died (18.1%). Causes of death were physical illness (40.5%), accidental overdose (35.4%), trauma (5.1%), suicide (1.3%), and unknown (17.7%). A multivariable Cox proportional hazard model indicated baseline alcohol dependence (adjusted hazard ratio [aHR] 1.83, 95% CI 1.09–3.07, <em>p</em> = 0.02), and evidence of hepatic fibrosis (aHR 1.81, 95% CI 1.08–3.03, <em>p</em> = 0.02) were risk factors for mortality. Among those under age 55 years, a history of a psychotic disorder was a risk factor for mortality (aHR 2.38, 95% CI 1.03–5.51, <em>p</em> = 0.04, adjusted for alcohol dependence at baseline, human immunodeficiency virus [HIV], and hepatic fibrosis). The primary study limitation concerns generalizability: conclusions from a community-based, diagnostically heterogeneous sample may not apply to specific diagnostic groups in a clinical setting. Because one-third of participants grew up in foster care or were adopted, useful family history information was not obtainable.In this study, we found methamphetamine, alcohol, and cannabis use were associated with higher risk for psychotic features, as were a past history of psychotic disorder, and experiencing traumatic events. We found that alcohol dependence, hepatic fibrosis, and, only among participants <55 years of age, history of a psychotic disorder were associated with greater risk for mortality. Modifiable risk factors in people living in precarious housing or homelessness can be a focus for interventions.
medicine, general & internal