Screen viewing is increasingly prevalent and has become a part of common sedentary behaviour globally. Excessive screen viewing time has been reported to be associated with poorer health outcomes in children, including increased obesity risk, reduced motor and cognitive development, and worse psychosocial health.1 Considering the detrimental effects of excessive sedentary behaviour on health, WHO2 and several countries have developed and issued recommendations to limit screen time to 1 h per day or less among children aged 2–5 years. However, concerns about the recommendation for a universal cutoff for children's overall screen time have been raised,3, 4 considering the relatively weak evidence and the differential effects of different types of screen time.
Research in context
Evidence before this study
Early childhood is a crucial period for developing behavioural habits that persist into adolescence and adulthood. Screen viewing has been suggested to have detrimental effects on child health, including an increased risk for obesity, reduced motor and cognitive development, and poorer psychosocial health. According to an estimation from UNICEF, WHO, and the World Bank, in 2016, more than 41 million children aged younger than 5 years were overweight and obese; the number has continued to increase during the past few years and is estimated to reach more than 50 million by 2030. Possible mechanisms linking screen viewing time with obesity include the inverse associations between screen viewing time and physical activity and sleep. However, the longitudinal effects of screen viewing on movement behaviours within a finite 24 h period are unknown. To identify studies that have assessed the associations between screen viewing time and physical activity, sedentary behaviour, and sleep in children, we searched PubMed from database inception to Sept 17, 2019, for studies published in English using the primary search terms “children”, “screen viewing”, “media use”, “television viewing”, “physical activity”, “exercise”, “fitness”, “sleep” and “movement behaviour”. Most studies assessing associations between screen viewing time and physical activity and sleep included school-aged children (age 6–12 years) and adolescents (age 12–19 years). A systematic review concluded that associations between screen time and physical activity were weak and inconsistent in school-aged children and adolescents. The literature consistently supports the importance of reducing screen viewing time for adequate sleep duration among children and adolescents. We further observed that most of the existing studies reporting on the association between screen viewing time and movement behaviours used a cross-sectional design. No previous study has assessed the associations between screen viewing time and movement behaviours using composition techniques, which take into account the compositional nature of movement behaviours within a defined 24 h period.
Added value of this study
By using analyses that account for compositional time use, we found that longer total and device-specific screen viewing time at age 2–3 years was associated with a higher proportion of time spent on sedentary behaviours and a lower level of moderate-to-vigorous physical activity and light physical activity at age 5·5 years. These findings extend existing research in children of preschool and school age (age 0–12 years) through the use of a longitudinal study design and the objective assessments of all movement behaviours, thereby strengthening the existing evidence linking screen viewing time with later child health.
Implications of all the available evidence
Our analysis addresses an important research gap and establishes a platform for research to further assess the effect of content type, timing, and parental co-use of media with their children on unfavourable movement behaviours. The findings suggest that further research on the longitudinal effects of screen viewing on other movement behaviours is warranted, and support public health efforts to reduce screen viewing time in young children to minimise the possible negative impact on children's health behaviours and general health.
One of the suggested mechanisms through which screen viewing might influence health is by displacing time that would otherwise be spent engaged in physical activity. This proposed hypothesis has, to our knowledge, rarely been investigated in children of preschool age (age 0–6 years), but has been supported by evidence from studies in adolescents. On one hand, some studies have suggested that long periods of screen viewing are strongly associated with decreases in physical activity.5, 6 On the other hand, a previous meta-analysis and a 2013 study reported that such associations are either weak or non-existent.7, 8 As such, evidence on the association between screen viewing time and physical activity is mixed and inconclusive. Inadequate sleep is another possible mechanism linking screen viewing time and poor health. School-aged children (age 6–12 years) and adolescents who spend excessive time watching screens might be more likely to have inadequate sleep, which is usually measured by a lower total sleep time and later bedtimes.9 Such associations are concerning since they develop during infancy and can persist in mid-childhood.10
However, studies investigating the associations between screen viewing time and other behaviours, such as physical activity and sleep, have mainly been done in school-aged children and adolescents (age 12–19 years), and adequate empirical evidence is scarce in young children. Moreover, most previous studies have been largely cross-sectional and thus findings could be affected by reverse causation. Previous studies have focused on conventional electronic devices such as televisions and computers, and little evidence is available regarding newer portable devices such as smartphones and tablets that have broad capabilities (eg, internet). Newer types of screen devices, which allow real-time interaction and potentially continuous stimulation for children, provide a different type of exposure and have become increasingly common in young children.11 Studies have also been criticised for investigating physical activity, sedentary behaviour, and sleep in isolation, or with only partial adjustment for time spent on other behaviours.12, 13 This approach does not account for the fact that an increase in one behaviour might lead to a decrease in at least one of the remaining behaviours.
Physical activity, sedentary behaviour, and sleep represent the movement spectrum across 24 h and are referred to as movement behaviours.13 Since 2016, several individual countries and WHO2 have developed integrated movement guidelines for children, which highlight the importance of targeting all movement behaviours to maximize health benefits. Previous evidence from our GUSTO mother-offspring cohort in Singapore showed that the majority of children aged 5·5 years engaged in excessive screen viewing time, had inadequate sleep, and low levels of physical activity; only one in 20 children met all the integrated guideline targets.14 Such a high prevalence of unfavourable behaviours highlights the necessity of identifying opportunities to improve all these behaviours in young children. Previous studies have repeatedly suggested the conceptualisation of individuals' daily activity data as compositions, consisting of time spent asleep, sedentary behaviour, and different intensities of physical activity (ie, light physical activity and moderate-to-vigorous physical activity [MVPA]), which sum to 24 h.13, 15 24 h movement data require different analytical methods than traditional regressions, to account for their compositional nature. Although composition techniques have been proposed to be valid approaches for analysing a set of variables that sum to a constant16 and implemented in a number of research areas,17, 18 few studies have implemented these techniques in movement behaviour research,15, 19 and no studies have explored how screen viewing behaviour affects these activity components.
To address existing research gaps, we aimed to investigate the associations of total and device-specific screen viewing time at ages 2–3 years with accelerometer-measured movement components, including sleep, sedentary behaviour, light physical activity, and MVPA at age 5·5 years.