Screen Time and Children’s Health

There has been concern, for some time, about the impact of screen time on children’s health; in particular, does time spent on activities involving a monitor or screen (including television) affect young people’s mental and physical health? A number of governmental and professional bodies, across the world, have suggested parental caution when allowing children access to digital devices, in the light of these concerns^1,2. In contrast, a recent report from the UK Royal College of Paediatrics and Child Health³ suggests that these concerns are, in effect, overstated: “...there is little evidence that this is the way that screen time works in influencing children’s well-being…”. On the bases of such reports, the Chief Medical Officer for England, said,⁴ “Time spent online can be of great benefit to children and young people…allowing young people to find support and information,” but also that, “We need to take a precautionary approach and our advice will support children to reap these benefits and protect them from harm.” Given these apparent contradictions, what should parents believe?

The report from the Royal College of Paediatrics and Child Health³ seems to be the work most ‘out of line’ with current findings, so let’s be clear about their conclusions. They noted clear relationships between screen time and poor diet, less exercise, and predictors of obesity. Children spending more than two hours a day using a screen (and that means almost all of them!) are at risk of depressive symptoms—a commonly noted finding in the literature. But this finding was juxtaposed quickly against a statement about other data suggesting benefits from screen time. Finally, this report pointed to trends for screen time being associated with poorer educational outcomes, less sleep, and worse fitness, but it rejected evidence for direct associations with ill health and poor well-being, pointing to other more important predictors of such negative outcomes.

Given that the conclusions from the UK Royal College of Paediatrics and Child Health and, potentially, the subsequent advice from the Chief Medical Officer for England seem to fly in the face of so much other data, it seems worthwhile to unpack the evidence on which these conclusions are based. The Royal College report is based heavily on a previous review of a set of twelve other reviews of the literature, along with some interviews conducted with children. Despite the suggestions that the evidence relating screen time to negative outcomes for children might be overstated, the report is merely noting a failure to find any consistent relationships across the previous reviews sampled. This raises a number of questions about the strength of the Royal College’s suggestions that are important for parents to consider.

Firstly, basing conclusions on a review of previous reviews is not particularly good practice. It would be different if the authors had used the pre-existing reviews to identify previous articles and had analyzed the data from those original articles, but this was not the case. In fact, some of the reviews, included in this report³, came to more salutary conclusions about negative relationships involving screen time than did the Royal College of Paediatrics and Child Health. A key problem when conducting analyses in this way is that the new review has to rely on the selection criteria from the preceding reviews, which may be highly variable and may not even be suitable for current purposes. In addition, pooling data across reviews may hide subtle trends in the individual reviews, or, indeed, in the individual reports, that get lost in the aggregating process.

Secondly, the original reviews sampled were conducted between 2011 and 2017, with most being published between 2013 and 2016. It might be asked whether these reviews contain data that are relevant to the rapid advance of social media use by children, some three years later—especially, remembering that the original articles sampled in the original reviews will have been published sometime before those reviews.

Thirdly, the concept of screen time itself is actually very wide, encompassing both traditional television watching and social media usage. As the authors of the Royal College report note, the data they examined largely focused on television watching. It is extremely unclear whether the effects derived from watching television, especially more traditional forms of viewing, are at all comparable with those emerging from new social media use. More traditional approaches to television watching include a shared common experience, discussed the next day, which is not necessarily true of newer individualized watch-on-demand viewing. A common rhetorical device used to dispel concerns about new social media technology is to point to previous worries about older technology, and say that this older technology didn’t have the devastating impact predicted. Indeed, in a similarly "upbeat" piece, Benjamin Rosen⁵ used this very device—noting concerns about the telephone, raised in 1887. This very analogy illustrates the problem perfectly! There is good evidence that the effects of communicating using the telephone through speech and through digital/textual means are quite different from one another—with the latter producing more, not less, isolation for some.⁶

Fourthly, an issue needing consideration, when examining the suggestion that screen time does not produce many negative results³, is that the authors are claiming that they failed to find a relationship—they are reporting a ‘null’ result. Null results are important to know about—their publication overcomes what is known as the ‘file drawer problem,’ where the failures to find an effect are ‘filed away,’ and only studies showing an effect are reported. This can skew the knowledge base. However, null findings are very difficult to interpret because they can be a product of a study’s insufficient power—the ability to detect a relationship if one is there. With no new quantitative analyses provided and a sample size of only 12, this statistical problem may well be the reason for the Royal College of Paediatrics and Child Health’s failure³. Moreover, this problem of power is especially great when the data are highly variable—which, when considering a diverse set of included studies, with different sampling techniques and orientations from one another, is doubtless the case.

The Royal College’s report was published at the same time as another very interesting study about the impact of screen time on psychological functioning and brain structure⁷. This investigation demonstrated that screen time has a range of associations with psychological function—especially externalizing psychopathologies and intelligence, and with altered brain structure—notably alterations to the prefrontal cortex (involved with planning and thinking), and with the posterior cingulate (involved with inhibition). The authors were very clear to suggest that the variety of impacts that they discovered means that there can be no clear one-to-one mapping of screen time to function and that some screen time may improve functioning, depending on what the screen time was used for. However, the potential for negative impacts was clear and not mentioned in the report from the Royal College of Paediatrics and Child Health.

On closer examination, the Royal College of Paediatrics and Child Health’s³ conclusions may need to be taken with a very large grain of salt—especially, as they stand in contrast to a vast amount of evidence suggesting harms derive from many aspects of social media use. Of course, screen time is a nebulous concept, and a much more important concept to examine might be the function of the usage. But the advice to parents should remain that of the American Psychiatric Association: limit screen time for children, and work with them to produce good habits. Anything else, at this stage, may verge on the irresponsible.