Family-Based Screen Time Break for Child Psychiatric Symptoms
Henning Tiemeier
DOI: https://doi.org/10.1001/jamanetworkopen.2024.19824
2024-07-13
JAMA Network Open
Abstract:The current surge in the prevalence of psychiatric problems among children and adolescents is poorly understood. When I talk about this enigma at a party, at work, or to strangers while traveling, most parents believe they know the answer: youths' excessive screen time caused the increase in psychiatric problems. The scientific evidence for an association between screen use and child psychiatric problems, however, is not clear. Many observational studies have been published, but evidence from randomized clinical trials remains scarce and hardly justifies the claims that screen time is the culprit behind psychiatric problems and a possible mental health crisis. The work of Schmidt-Persson et al 1 makes an important contribution to this discussion as it tests the crucial hypothesis that the relationship is causal. In a secondary analysis of a small randomized clinical trial, they show that a 2-week break in screen time use can reduce the psychological problems of children aged 4 to 17 years. Selected from a large pool of 1420 eligible families, which in turn come from a sample of 6820 families that responded to a survey, they found 89 families eligible and willing to be randomized to a family intervention; 45 families agreed to reduce their leisure-time screen use to less than 3 hours for 2 weeks and, 44 families served as controls. In total, 181 children participated. Two weeks after the intervention, children had a 1.67-point decreased score on the Strength and Difficulties Questionnaire, a relatively modest but significant change. 2 This questionnaire is a well-validated and commonly used measure in general population studies. Several factors make this intervention family based. First, families participated with all their children. Second, families were selected that had a moderately to severely high level of parental recreational screen use per day (>2.4 hours). Third, and most important, both children and parents had to hand in smartphones and tablets for 2 weeks; television monitors were installed in the homes to check compliance. Hence, the intervention targeted both children and their parents. This likely enhanced compliance, which was indeed very good (97% used screens <7 hours per week), and resulted in a joint family effort. Also, the extra time freed up could be spent together. Interestingly, in a prior analysis, the authors showed that parental mood also improved. 3 The authors, however, did not investigate in the present study whether this parental mood change accounted for the observed improvement in children's psychiatric problems. The study by Schmidt-Persson et al 1 is methodologically sophisticated; the tobit regression model accounts for the generally quite low levels (ie, the floor effects) of the children's scores on the on the Strength and Difficulties Questionnaire. The randomization was performed independently of the investigator team, so that the researchers were blinded as to who was in the intervention group (obviously parents and children were not blinded). The outcome is modeled as a change from baseline, done under the assumption that any difference between the groups can be attributed to random chance and that there were no problems with the randomization process and no significant measurement error. 4 In addition, the outcomes were imputed in sensitivity analyses, assuming that all missing outcome data (ie, the behavioral and emotional problems) can be characterized by observable variables that must be well measured and used for the imputation. 4 This assumption is less obvious, because if families that were frustrated and unhappy with the intervention had a much higher chance to drop out, this assumption would likely not be met. These assumptions are relevant, as the effect of the intervention would not be significant if only the 162 children with complete follow-up were included and baseline problems had not been accounted for. The scientific relevance of this trial of screen time reduction is that observational studies of screen time may be biased. A recent publication from my group suggests that at least some of the association between screen time and psychiatric problems may be due to shared genetic risk factors that explain both the propensity for screen use and the vulnerability for attention and, to a lesser degree, for internalizing problems. 5 Seminal work by Orben and Przybylski 6 shows that methodological flexibility may explain some of the reported associations between screen use; their analyses suggest that psychiatric problems and digital technology likely explain less than 0.4% of the variation. Like parents, many clinicians and policymakers view screen time as a modifiable risk factor for behavioral problems. What makes the work of Schmidt-Persson et al 1 so important is that it is a proof of principle: not only can -Abstract Truncated-
medicine, general & internal