March 17, 2021

expert reaction to observational study on electronic media use in 5 year olds and psychosocial symptoms

An observational study published in BMJ Open looks at extensive electronic media use in five-year-olds and risk of emotional and behaviour issues.

Dr Amy Orben, College Research Fellow at Emmanuel College and Research Fellow, MRC Cognition and Brain Sciences Unit, University of Cambridge, said:

“When considering this study it is important to remember that it is purely observational, i.e. it cannot establish a causal link between time spent on electronic media and developmental outcomes in small children. Many factors that might influence whether a child spends a lot of time on electronic media (e.g. whether both parents have to work, what sort of neighbourhood the child lives in), also influence their psychosocial symptoms: this means that an association can exist even if no causal link is present. As the study analyses either include no control variables to account for this, or only minimal controls, this study does not provide any evidence of such a causal link.

“It is important to note that many analyses were conducted to link screen time at age 18 months and 5 years to psychosocial symptoms at age 5, and only a couple of those came out statistically significant. Most of these significant links were found in models when no proper control variables were included. In models with minimal control variables only 1 out of 16 analyses linking screen time to psychosocial symptoms were significant. This could well be noise, as using normal p values one would expect 1 in 20 analyses to be significant if no effect is present in the data. I would predict there would be even fewer significant links if more extensive control variables were included (as mentioned in the above paragraph). 

“Further, the outcome variables were dichotomised in a way that was not explained, and it is therefore difficult to judge the robustness of this analysis.”

‘High-dose electronic media use in five-year-olds and its association with their psychosocial symptoms: a cohort study’ by Janette Niiranen et al. was published in BMJ Open at 23:30 UK time on Wednesday 17 March.

DOI: doi:10.1136/bmjopen-2020-040848

Declared interests

Dr Amy Orben: “No COIs.”