Does Alcohol Use Affect the Relationship Between Social Support and Depression? An Examination of Permanent Supportive Housing (PSH) Residents in Fort Worth, Texas

Background: Social support tends to be protective against depression in a wide variety of groups. However, relatively little is known about how alcohol use might affect this relationship over time, especially among supportive housing residents, who are at risk for both depression and poor social support. Purpose: This study examined whether the association between social support and depression was modified by alcohol consumption among permanent supportive housing (PSH) residents in Fort Worth, Texas. Methods: We used the baseline and 6-month follow-up data from the Mobile Community Health Assistance for Tenants (m.chat) program, collected during 2014-2017. m.chat was a technology-assisted health coaching program for people with mental health conditions residing in PSH in Fort Worth, Texas. Participants’ current levels of depression and social support were measured using Patient Health Questionnaire-9 (PHQ-9) and Interpersonal Support Evaluation List (ISEL), respectively; alcohol consumption was measured via a self-report frequency measure that ranged from “0” (never) to “4” (more than 4 times a week). The association between social support and depression was studied using generalized estimating equation (GEE) models including alcohol use as an effect modifier and sex, age, race, marital status and perceived physical health as a priori covariates. This analysis included 567 participants. Results: At baseline, about 38.2% of participants reported some alcohol consumption in the past 90 days. Greater baseline social support was associated with improvements in depression severity. One unit increase in the baseline ISEL score predicted a 4.3% reduction (95%CI: 2.6%, 6.1%) in depression scores over time. Greater baseline alcohol consumption predicted an 11.9% increase (95%CI: 0.8%, 21.6%) in depression scores over time. We did not find a significant interaction between alcohol consumption and social support and changes in depression severity after adjusting for sex, age, race, marital status and perceived physical health. Conclusions: Among PSH residents in Fort Worth, Texas, greater social support was associated with a reduction in depression scores. Higher alcohol consumption was associated with an increase in depression scores. The protective effect of social support did not differ by alcohol consumption level. These findings can be used to design more robust health coaching programs for formerly homeless persons that integrate positive social support.