RESUMO
BACKGROUND: Evidence-based interventions addressing loneliness and social isolation are needed, including among low-income, community-dwelling older adults of diverse racial and ethnic backgrounds. Our objective was to assess the effect of a peer intervention in addressing loneliness, isolation, and behavioral health needs in this population. METHODS: We conducted a mixed-method, two-year longitudinal study of a peer-outreach intervention in 74 low-income older adults recruited via an urban senior center in San Francisco. Structured participant surveys were conducted at baseline and every 6 months for up to 2 years. Outcomes included loneliness (3-item UCLA loneliness scale), social interaction (10-item Duke index), self-perceived socializing barriers (range: 0-10), and depression (PHQ-2 screen). Data were analyzed using mixed-effects linear and logistic regression adjusted for age and gender. Qualitative, semi-structured interviews with participants (N = 15) and peers (N = 6) were conducted in English and Spanish and analyzed thematically. RESULTS: Participants were on average 71 years old (range: 59-96 years), with 58% male, 15% LGBT, 18% African American, 19% Latinx, 8% Asian, 86% living alone, and 36% with an ADL impairment. On average, 43 contact visits (IQR: 31-97 visits) between participants and peers occurred over the first year. Loneliness scores decreased by, on average, 0.8 points over 24 months (p = 0.015). Participants reported reduced depression (38%-16%, p < 0.001) and fewer barriers to socializing (1.5 fewer, p < 0.001). Because of the longitudinal relationship and matching of characteristics of peers to participants, participants reported strong feelings of kinship, motivations to reach out in other areas of life, and improved mood. CONCLUSION: Diverse older adults in an urban setting participating in a longitudinal peer program experienced reduced loneliness, depression, and barriers to socializing. Matching by shared backgrounds facilitated rapport and bonding between participants and peers.