Sleep

Sociodemographic and maternal-related correlates of children's movement behaviours from preschool to adolescence in Singapore: a longitudinal cohort study.

TL;DR

Children become less physically active and have longer screen-viewing as they transition into adolescence, with the greatest declines in physical activity and increases in inactivity and screen-viewing occurring between ages 8 to 10 years, and girls, Malay ethnicity, lower socioeconomic status, and less favourable maternal movement behaviours being key correlates.

Key Findings

Sleep declined by 3% from ages 5.5 to 12 years in a Singapore cohort of 837 children.

  • Sleep decreased from 9.1 to 8.8 hours/day between ages 5.5 and 12 years.
  • This represents a 3% reduction over the study period.
  • Sleep was measured using wrist-worn accelerometry.
  • Data were collected at ages 5.5, 8, 10, and 12 years as part of the Growing Up in Singapore Towards healthy Outcomes (GUSTO) study.

Light physical activity (LPA) declined by 24% from ages 5.5 to 12 years.

  • LPA decreased from 5.8 to 4.4 hours/day between ages 5.5 and 12 years.
  • The greatest annual decrease in LPA (0.3 hour/annum) occurred from ages 8 to 10 years.
  • LPA was assessed using wrist-worn accelerometry across four time points.
  • Sample included 837 children from the GUSTO longitudinal cohort.

Moderate-to-vigorous physical activity (MVPA) declined by 44% from ages 5.5 to 12 years.

  • MVPA decreased from 71.3 to 40.1 minutes/day between ages 5.5 and 12 years.
  • The greatest annual decrease in MVPA (10.4 min/annum) occurred from ages 8 to 10 years.
  • This represents the largest proportional decline of any movement behaviour measured.
  • MVPA was measured using wrist-worn accelerometry.

Inactivity increased by 26% from ages 5.5 to 12 years.

  • Inactivity increased from 8.0 to 10.1 hours/day between ages 5.5 and 12 years.
  • The greatest annual increase in inactivity (0.6 hour/annum) occurred from ages 8 to 10 years.
  • Inactivity was assessed via wrist-worn accelerometry.
  • Linear regression models with generalised estimating equations were used to analyse longitudinal changes.

Screen-viewing increased by 155% from ages 5.5 to 12 years.

  • Screen-viewing increased from 1.8 to 4.6 hours/day between ages 5.5 and 12 years.
  • The greatest annual increase in screen-viewing (0.8 hour/annum) occurred from ages 8 to 10 years.
  • Screen-viewing data were obtained via proxy-reported measures.
  • This was the largest proportional increase of any behaviour measured in the study.

The period from ages 8 to 10 years represented the most pronounced phase of unfavourable change across multiple movement behaviours.

  • The greatest annual increase in inactivity (0.6 hour/annum) occurred between ages 8 and 10 years.
  • The greatest annual increase in screen-viewing (0.8 hour/annum) also occurred between ages 8 and 10 years.
  • The greatest annual decreases in both LPA (0.3 hour/annum) and MVPA (10.4 min/annum) occurred during this same age interval.
  • Four measurement time points were used: ages 5.5, 8, 10, and 12 years.

Being female was associated with less favourable movement behaviour profiles, including less sleep, higher inactivity and screen-viewing, and/or lower physical activity.

  • Girls had significantly less sleep, higher inactivity and screen-viewing, and/or lower PA compared to boys.
  • Sex was identified as a key sociodemographic correlate across the study period.
  • Analyses used linear regression models with generalised estimating equations.
  • The association was consistent across the age range studied (5.5 to 12 years).

Malay ethnicity was associated with less favourable movement behaviour profiles in children.

  • Children of Malay ethnicity had significantly less sleep, higher inactivity and screen-viewing, and/or lower PA.
  • Ethnicity was identified as a significant sociodemographic correlate in the Singapore cohort.
  • The GUSTO study includes a multiethnic Singaporean sample, enabling ethnic comparisons.
  • These findings highlight ethnicity as a characteristic to be prioritised in future interventions.

Lower socioeconomic status was associated with less favourable movement behaviour profiles in children.

  • Children of lower socioeconomic status had significantly less sleep, higher inactivity and screen-viewing, and/or lower PA.
  • Socioeconomic status was identified as a key sociodemographic correlate alongside sex and ethnicity.
  • The study used proxy-reported data for sociodemographic variables.
  • Authors highlighted lower SES as a characteristic to be prioritised in future interventions.

Maternal physical activity levels and sitting time were associated with children's sleep, inactivity, and MVPA up to age 8 years.

  • Maternal PA levels and/or sitting time were significantly associated with children's sleep, inactivity, and MVPA.
  • These associations were observed up to age 8 years but not consistently beyond this age.
  • Maternal lifestyle-related correlates were examined alongside sociodemographic factors.
  • Analyses used linear regression models with generalised estimating equations applied to the 837-child cohort.

Maternal sitting and screen-viewing behaviours were associated with children's screen-viewing at all ages studied.

  • Maternal sitting and screen-viewing behaviours were significantly associated with children's screen-viewing from ages 5.5 to 12 years.
  • This association persisted across all four measurement time points, unlike associations for other movement behaviours.
  • Screen-viewing data were proxy-reported.
  • This finding suggests a persistent modelling effect of maternal sedentary screen behaviours on children's screen time throughout childhood into early adolescence.

What This Means

This research followed 837 children in Singapore from age 5.5 to 12 years, measuring how much they slept, moved, and sat still at four different points in time. The study found that as children grew older, they became substantially less active and spent far more time being inactive and watching screens. By age 12, children were doing about 44% less moderate-to-vigorous physical activity than they did at age 5.5 (dropping from about 71 to 40 minutes per day), their screen time had more than doubled (rising from about 1.8 to 4.6 hours per day, a 155% increase), and they were inactive for roughly 2 more hours each day. Notably, the sharpest and most rapid changes in all of these behaviours happened between ages 8 and 10, suggesting this is a particularly critical window. The study also identified groups of children who tended to fare worse. Girls, children of Malay ethnicity, and children from lower socioeconomic backgrounds consistently showed less sleep, more inactivity, more screen time, and less physical activity. Additionally, mothers' own habits mattered: when mothers were more physically active or sat less, their children tended to have better movement behaviours up to age 8, and mothers who spent more time sitting or watching screens tended to have children with higher screen time throughout the entire age range studied. This research suggests that efforts to keep children active and limit screen time may need to focus especially on the transition from primary school around ages 8 to 10, and should prioritise support for girls, lower-income families, and specific ethnic groups. The findings also highlight that encouraging mothers to be more physically active and reduce their own screen time could have a lasting positive effect on their children's habits, particularly in early childhood.

Have a question about this study?

Citation

Tan S, Edney S, Padmapriya N, Tan S, Chong Y, Tan K, et al.. (2026). Sociodemographic and maternal-related correlates of children's movement behaviours from preschool to adolescence in Singapore: a longitudinal cohort study.. BMJ open. https://doi.org/10.1136/bmjopen-2025-105140