More time on social media today, more depressive symptoms in a year

In recent years, both the time spent on social media and the prevalence of persistent sadness/hopelessness have increased among teenagers. The “social media → depression” theme is common in public discourse, but scientific data have long been mostly “snapshots” of a single moment in time, and have been difficult to discern.

What was already known

Cross-sectional studies have yielded mixed results, ranging from weak positive associations between screen time and depressive symptoms to null effects. Even in longitudinal studies, confounding is a key methodological issue:

• interpersonal differences (some people are generally more online and feel sad more often),
• and intrapersonal fluctuations (this year a specific teenager spends more time on social networks than usual - what will happen in a year?).

Without dividing them, it is easy to mistake “differences between people” for “changes in one person over time.” Plus, the opposite sequence is possible: it is not social networks that increase symptoms, but a worsening mood that pushes online activity to increase.

Why is age 9-12 important?

It's early puberty: the brain's systems of motivation and sensitivity to social cues are accelerating, while control and self-regulation are still maturing. At the same time, sleep patterns, daily routines, and social circles are changing, all of which increase vulnerability to behavioral "swings."

In children aged 9–12, bursts of time on social networks predict increased depressive symptoms a year later. No reverse sequence — “first depression, then increased online time” — was found. The findings were based on data from 11,876 participants in the national ABCD project (USA), observation — 4 annual waves. The study was published in the journal JAMA Network Open.

What's new

• Within a single adolescent, if in a given year they spent more time on social media than usual, then a year later they had higher depressive symptoms (standardized effects β=0.07 and β=0.09 in two consecutive intervals - small but stable).
• Depressive symptoms did not lead to subsequent increases in time spent on social media in any time frame.
• There were no consistent differences in depressive symptom levels between different adolescents (those who "on average" sedentary more vs. less) after controlling for gender, race/ethnicity, income, parental education, and family context.

How it was studied

The researchers used data from the Adolescent Brain Cognitive Development (ABCD) project, the largest longitudinal study of brain development and health in adolescents in the United States (21 centers). Participants joined the study at ages 9–10 and completed surveys annually for three years:

• Social networks: self-report of average daily time spent on social networks (minutes on weekdays and weekends).
• Depressive symptoms: scores on the CBCL Depressive Problems Scale (parent version), which measures the frequency of symptoms in daily life.

The key tool of the analysis is RI-CLPM (random-intercept cross-lagged panel model). In simple terms, it divides the variation into two parts:

1. Differences between people (some people are generally more online or more sad).
2. Fluctuations within one person from year to year (this year he sits more than usual - what will happen next?).

This approach allows us to catch the time sequence specifically within the teenager, and not confuse it with the fact that “some people are generally more on their phones and are sad more often.”

The model described the data well (according to the fit criteria), which increases confidence in the results.

Why is this important?

• In recent years, both screen time and the proportion of teens with persistent sadness/hopelessness have been on the rise. Until now, many studies have been “snapshots” (one point in time) and have not allowed us to understand what follows.
• This shows a temporal order: a surge in social media → more symptoms a year later. This is not yet proof of causality, but it is a much stronger argument than simple correlations.

What it doesn't mean

• The study is observational. We see consistency and connection, but we cannot name a specific causal mechanism.
• They measured time, not content: passive scrolling, comparing yourself to others, cyberbullying, toxic topics - all of these could play a role, but were not taken into account separately.
• No diagnoses were made: we are talking about symptoms according to a validated questionnaire.

Practical implications for parents, schools and doctors

A "preemptive" signal. If a child aged 9-12 suddenly spends more time on social networks than usual, this is a reason to prevent mood problems over the next year.

What can be done without panic and prohibitions:

• Family media plan: agreements on time and “screen-free windows” (dinner, getting ready for bed, first hour after waking up).
• Night mode: silent notifications and no gadgets in the bedroom.
• Conscious consumption: unsubscribing from “trigger” content, adding supportive communities, reflecting on “how do I feel about this content now?”
• Age restrictions: Most platforms are 13 and up; parental guidance and privacy filters are especially important before this age.
• A conversation about risks: comparing yourself to “perfect” feeds, FOMO, cyberbullying, “fool challenges”, algorithms.
• Alternatives to dopamine: sports/movement, offline communication, creativity, short mindfulness practices.

For clinicians: add 2–3 simple screening questions about social media to your early teen visits and discuss realistic steps—not “ban everything,” but reduce peaks and reinforce helpful practices.

How strong is the effect?

The effects are small in magnitude but consistent. In public health, it is these “small but massive” effects that often drive significant changes at the population level—especially when millions of children are involved.

What is not closed and where to go next

• Mechanisms: passive scrolling, social comparison, rumination, sleep deprivation, cyberbullying? More frequent measurements are needed (diaries, EMA, smartphone sensors).
• Content instead of minutes: which formats protect (social support, learning, creativity), and which increase the risk.
• Individual differences: who social networks help and who they harm more (personality characteristics, family climate, stressful events).
• The role of platforms: design without “hooks”, with transparent feed settings and restrictions on night pushes for teenagers.