Associations between social media-based emotional support and internalizing symptoms: a longitudinal study in a national sample of U.S. adolescents and emerging adults

Abstract

The mental health crisis in young people persists, with adolescents and emerging adults from minority backgrounds facing elevated risk for internalizing symptoms. While social media (SM) use is ubiquitous among U.S. young people, research has primarily focused on time spent on SM and potential harms, with limited attention to protective factors such as emotional support. This study examined longitudinal associations between perceived SM-based emotional support and internalizing mental health symptoms in young people from marginalized backgrounds. Between November 2023 and March 2024, 689 U.S. persons aged 13–21 from minority backgrounds (racial/ethnic, LGBTQ+, or rural-living) were recruited through SM advertisements. Participants completed web-based surveys at baseline and two follow-up waves (4–5 months and 8–9 months post-baseline). Depression (PHQ-9), anxiety (GAD-7), and loneliness (UCLA-3) were assessed as outcomes. SM-based emotional support was measured using a modified 4-item PROMIS Emotional Support scale. Multi-level models examined associations between SM-based emotional support and mental health outcomes, adjusting for baseline demographic and psychosocial covariates including in-person peer support, number of SM platforms used, and everyday discrimination. The sample was predominantly female birth sex (76.9%), White non-Hispanic (52.2%), with mean age 19 years. Nearly half (46.9%) identified as lesbian, gay, or bisexual. After adjusting for all covariates, higher SM-based emotional support at Wave 1 was significantly associated with lower depression (β = −0.09, p < 0.05), anxiety (β = −0.11, p < 0.01), and loneliness (β = −0.07, p < 0.01) scores. These associations remained stable across the 10-month study period, as indicated by non-significant interaction terms between wave and SM-based emotional support. Loneliness scores decreased significantly over time by 0.26 points per wave ( p < 0.01), while depression and anxiety scores showed no significant temporal changes. Perceived availability of emotional support on social media was associated with lower internalizing symptoms in young people from marginalized backgrounds, independent of in-person peer support. These associations remained stable over time, suggesting SM-based emotional support may complement traditional forms of social support, particularly for young people experiencing marginalization or discrimination.