ASSOCIATION BETWEEN SOCIAL MEDIA USE AND DEPRESSION AMONG U.S. YOUNG ADULTS

Abstract

Background: Social media (SM) use is increasing among U.S. young adults, and its association with mental well-being remains unclear. This study assessed the association between SM use and depression in a nationally representative sample of young adults.

Methods: We surveyed 1,787 adults ages 19 to 32 about SM use and depression. Participants were recruited via random digit dialing and address-based sampling. SM use was assessed by self-reported total time per day spent on SM, visits per week, and a global frequency score based on the Pew Internet Research Questionnaire. Depression was assessed using the Patient-Reported Outcomes Measurement Information System (PROMIS) Depression Scale Short Form. Chi-squared tests and ordered logistic regressions were performed with sample weights.

Results: The weighted sample was 50.3% female and 57.5% White. Compared to those in the lowest quartile of total time per day spent on SM, participants in the highest quartile had significantly increased odds of depression (AOR = 1.66, 95% CI = 1.14-2.42) after controlling for all covariates. Compared with those in the lowest quartile, individuals in the highest quartile of SM site visits per week and those with a higher global frequency score had significantly increased odds of depression (AOR = 2.74, 95% CI = 1.86-4.04; AOR = 3.05, 95% CI = 2.03-4.59, respectively). All associations between independent variables and depression had strong, linear, dose-response trends. Results were robust to all sensitivity analyses.

Conclusions: SM use was significantly associated with increased depression. Given the proliferation of SM, identifying the mechanisms and direction of this association is critical for informing interventions that address SM use and depression.

Keywords: communications media; depression; internet; social media; young adult.

Figure. Multivariable Associations between Depression and Social Media Use Variables

Each social media use variable is divided into quartiles from lowest (Q1) to highest (Q4). Vertical bars represent 95% confidence interval and point estimates of adjusted odds ratio. P value for overall linear effect was .002, <.001 and <.001 respectively for each social media use variable. The multivariable model adjusted for age, sex, race, relationship status, living situation, household income, and education level.