Social Media Use and Well-Being Across Adolescent Development
Among Australian students, well-being was highest with moderate after-school social media use, while heavy use and (in older teens) no use were linked to poorer well-being.
*Longitudinal cohort (repeated cross-sectional); Level 2b (OCEBM).
*Longitudinal cohort (repeated cross-sectional); Level 2b (OCEBM).
Citation
Singh B, Zhou M, Curtis R, Maher C, Dumuid D. Social Media Use and Well-Being Across Adolescent Development. JAMA Pediatrics. 2026;180(3):288-297. doi:10.1001/jamapediatrics.2025.5619
Background
Prior studies disagree on whether social media harms or helps adolescents. This study tested whether associations with well-being change by age and sex over early-to-late adolescence.
Patients
Australian government-school students in grades 4-12 (2019-2022). Excluded: non-government schools; nonbinary/unknown sex not analyzed due to data restrictions.
Intervention
After-school social media use (weekdays 3-6 PM): none (0 hours/week), moderate (>0 to <12.5), highest (≥12.5).
Control
Moderate use.
Outcome
(Primary) Low overall well-being: mean of 8 domains scored 1-5; low defined as <3. (Secondary) Each domain separately.
Follow-up Period
Up to 3 years per student; main analyses pooled 2020-2022.
Results
100,991 adolescents contributed 173,533 observations (mean age 13.5 years; 49.9% female). Associations were nonlinear: moderate use generally aligned with the best well-being.
| Group (school stage) | Exposure vs moderate use | Adjusted odds ratio for low well-being (95% CI) |
|---|---|---|
| Girls (grades 7-9) | Highest use | 3.13 (2.88-3.39) |
| Boys (grades 7-9) | Highest use | 2.25 (1.86-2.72) |
| Girls (grades 10-12) | No use | 1.79 (1.41-2.27) |
| Boys (grades 10-12) | No use | 3.00 (2.01-4.46) |
Odds ratio: relative odds of low well-being compared with moderate use.
Analyses used mixed-effects models adjusted for region, parent education, grade, and year.
Limitations
Observational design cannot prove cause and effect. Social media use was self-reported and limited to weekdays after school, not total use. Well-being was simplified into high vs low, and older nonusers were uncommon, increasing uncertainty.
Funding
Australian governmental research grants; funders had no role.
Clinical Application
In counseling teens, avoid “all-or-nothing” rules; screen for heavy use, but also ask whether complete avoidance limits social connection, especially in older boys.
Discussion
Sign in to join the discussion.
This is an important subject. Why does the summary list "moderate" as both an Intervention and an Outcome? If the study were repeated I'd want to know about total social medial use. Most of my patients with higher screen time hours seem to be on a device late into the night. And what happens after school? For lots of kids, sports, clubs, activities, etc. Australia banned social medial use for kids under 16 effective December 2025.
In this cohort study of self-reported social media use, the researcher choose to compare the low and high users of social media to the moderate users for the purpose of comparative analysis. There was no true intervention but rather self-identified groupings which included all three groups. Then the researches decided that the moderate use group would be the control group.
In this longitudinal cohort study, highest after-school social media use was associated with low well-being vs moderate use (eg, girls grades 7-9: OR 3.13; 95% CI, 2.88-3.39). Given the observational design, how should concerns about confounding and applicability affect practice change? This study reports adjusted odds ratios from mixed-effects logistic models comparing social media use categories with low well-being. Which interpretation is most appropriate for an OR of 3.13 for highest vs moderate use among girls in grades 7-9?