Skip to content

Genetic Factors May Substantially Account for Associations Between Screen Time and Psychiatric Problems in Preadolescents

Key findings

  • This population-based cohort study evaluated the extent of genetic confounding in causal associations between screen time and attention problems or internalizing problems in 4,262 children of European ancestry, ages 9 to 11
  • Each standard deviation (SD) of child-reported screen time was associated with a 0.10-SD increase in attention problem score and a 0.03-SD increase in internalizing problem score
  • In a model adjusted for a screen time polygenic risk score (PRS) and attention problem PRS, genetic confounding explained 10% of the association between child screen time and attention problems or internalizing problems
  • In a model adjusted for PRSs using single nucleotide variant–based heritability, genetic confounding explained 43% of the link between screen time and internalizing problems, and a model using twin-based heritability fully accounted for the association
  • The potential for genetic confounding should be considered in sociobehavioral studies of modifiable factors for youth mental health

Excessive "screen time" is widely associated with mood and attention problems among children and adolescents. Scientists are finding that genetic factors can affect not just health problems but also behavioral traits, and now Harvard researchers have determined genetic variations may largely explain associations between screen time and certain psychiatric problems in preadolescents.

The authors of their report in JAMA Network Open are Yingzhe Zhang, MS, of the Harvard T.H. Chan School of Public Health, Karmel W. Choi, PhD, clinical psychologist and researcher in the Center for Precision Psychiatry at Massachusetts General Hospital, Henning Tiemeier, MD, PhD, Sumner and Esther Feldberg professor of Maternal and Child Health at the Chan School of Public Health, and colleagues.


The researchers obtained 2016–2019 data from the Adolescent Brain Cognitive Development (ABCD) Study, a long-term, nationally representative U.S. study that recruits primarily from schools. They included 4,262 children of European ancestry, ages 9 to 11 at baseline, participating at 21 sites.

At baseline, children completed a 14-question questionnaire about daily screen time and a parent or other caregiver completed a shorter questionnaire about their child's total screen use.

One year later, parents completed the Achenbach Child Behavior Checklist for Ages 6–18 to assess attention problems (attention problem subscale; score range, 0–20) and internalizing problems (anxious or depressed, somatic complaints and withdrawn or depressed subscales; score range, 0–64).

The research team applied a genetic sensitivity analysis method called Gsens, which combines information from:

  • Polygenic risk scores (PRSs)—calculated in this study for leisure television watching time, attention-deficit/hyperactivity disorder (ADHD), and major depression based on genome-wide studies of adults
  • Single nucleotide variant (SNV)-based heritability—In this study, calculated for screen time, attention problems, and internalizing problems using the GCTA–GREML software, adjusting for age, sex, study site, and top 10 principal components of ancestry
  • Twin-based heritability using 216 pairs of European monozygotic twins and 333 pairs of dizygotic twins from the ABCD Study

Heritability Estimates

Children reported a mean daily screen time of 3.2 hours and parents reported a mean daily time of 1.2 hours. Mean scores at the one-year follow-up were 2.9 points for attention problems and 5.4 points for internalizing problems.

Adjusted SNV-based heritability estimates were 0.08 for child-reported screen time, 0.18 for attention problem scores, and 0.07 for internalizing problem scores. Twin-based heritability estimates were 0.58 for screen time, 0.88 for attention problems, and 0.48 for internalizing problems.

Each 1-standard deviation (SD) increase in screen time was associated with a 0.10-SD increase in attention problem score and a 0.03-SD increase in internalizing problem score.

Genetic Risk Scores

A 1-SD higher television time PRS was associated with a 0.18-SD longer child-reported screen time. The PRSs of ADHD and depression were also associated with child-reported screen time, although to a lesser extent (β = 0.14 SD and 0.07 SD, respectively).

Parent-reported child screen time was also associated with all three PRSs, although with smaller magnitudes.

Genetic Confounding

In a model adjusted for the screen time PRS and attention problem PRS, the researchers found that 10% of the association between screen time and attention problems was explained by genetic confounding—an individual genotype was a common cause. Modeling the association between screen time and internalizing problems yielded the same figure, 10%.

In a model adjusted for PRSs using SNV-based heritability, the magnitude of genetic confounding for internalizing problems was much larger, 42.7%. In models adjusting for PRSs using twin-based heritability, genetic confounding fully accounted for the associations between screen time and attention problems or internalizing problems.


Many policymakers and scientists view child screen time as a modifiable risk factor. However, if genetic factors account for a large part of the observed association between screen time and mental health, then interventions restricting screen time could be less effective than expected in preventing child attention and internalizing problems.

Parents and clinicians should not interpret the study findings as suggesting they should adopt a lenient attitude toward children who use electronic devices excessively. Increased screen time could be a factor in other risks, including reduced physical activity and poor academic achievement.

Learn more about the Center for Precision Psychiatry

Learn more about the Psychiatric and Neurodevelopmental Genetics Unit

Related topics


Karmel W. Choi, PhD, Karestan C. Koenen, PhD, and Jordan W. Smoller, MD, ScD, of the Department of Psychiatry, and colleagues used a novel two-stage approach, an exposure-wide association scan followed by Mendelian randomization, to validate actionable targets for efforts to prevent depression.


Dylan E. Hughes, Joshua L. Roffman, MD, MMSc and colleagues found that psychopathology in children is most strongly related to a neurodevelopmental polygenic score comprising overlapping variants across ADHD, autism spectrum disorder, MDD and Tourette's, rather than to disease-specific polygenic scores as in adults.