Associations of adolescent social media use trajectories with spatial and verbal memory: a prospective cohort study

Background: Evidence on screen time and cognition is mixed, with few longitudinal studies on social media patterns and memory. This study aimed to examine how social media trajectories relate to cognitive performance in early adolescence. Methods: We analyzed a prospective cohort (N = 7528, 51.1% male, mean age: 10 years (8-13 years), 41.8% non-White) from the Adolescent Brain Cognitive Development (ABCD) Study (baseline (2016-2018) to Year 2 (2018-2020)). Group-based trajectory modeling estimated patterns of daily social media use from baseline-Year 2. Three social media time trajectories: (1) no or very low use, (2) low but increasing use, and (3) high and increasing use were identified. Cognitive functioning was measured using the Little Man Task (LMT) and the Rey Auditory Verbal Learning Test (RAVLT). Linear regression models estimated the association between social media time trajectories and cognitive functioning at Year 2, adjusting for baseline age, sex, race, ethnicity, household income, parent education, attention-deficit/hyperactivity symptoms, depressive symptoms, respective baseline cognitive measures, non-social media screen use, and study site. Findings: Compared to no or very low social media use, low increasing social media use was associated with lower performance on the RAVLT Initial Trials (B: -1.38; 95% CI: -1.82, -0.94), RAVLT Retroactive Interference Trial (B: -0.38; 95% CI: -0.52, -0.25), and RAVLT Long Delay Trial (B: -0.41; 95% CI: -0.55, -0.26). Compared to no or very low social media use, high increasing social media use was associated with lower accuracy on the LMT (B: -0.03; 95% CI: -0.05, -0.01), and lower performance on the RAVLT Initial Trials (B: -1.90; 95% CI: -2.76, -1.04), RAVLT Retroactive Interference Trial (B: -0.61; 95% CI: -0.89, -0.32), and RAVLT Long Delay Trial (B: -0.55; 95% CI: -0.84, -0.27). Interpretation: Increases in social media time were prospectively associated with lower cognitive performance two years later. Monitoring digital use, implementing a Family Media Use Plan, and balancing screen time with cognitively enriching activities may help mitigate these effects. Future studies should examine the effects of various contemporary media on cognitive functioning. Funding: The research was supported by the Doris Duke Foundation (2022056).