Family factors associated differently with preschool activity and motor ratings

A national household survey of 10,967 children aged 3 to 6 found that family and child characteristics associated with reported physical activity were not the same as those associated with parents’ ratings of motor competence. After adjustment for 19 predictors, child sociability, the availability of sports equipment at home, and caregiver participation in activity were positively associated with daily physical activity. Parental support was more strongly associated with parent-rated motor competence, although that result weakened when support items overlapping with co-activity were removed. The cross-sectional study, conducted in all 31 province-level divisions of mainland China in 2020, describes statistical associations rather than causal effects.

The researchers used the 2020 China National Fitness Activity Status Survey, employing a three-stage, stratified probability-proportional-to-size design. The first stage selected 486 county-level units, with population size determining selection probabilities and socioeconomic development and geographic location used for stratification. The second stage sampled residential and village committees, with the urban and rural numbers fixed according to each county’s population composition. The third stage drew families, including two children aged 3 to 6 per selected committee where possible, and parents or primary caregivers served as proxy respondents. Trained surveyors completed questionnaires during home visits, with electronic upload, double data entry and telephone verification. Survey weights accounted for the sampling design, while regression variance estimates treated counties as sampling units nested within provinces. The broader sample included 10,967 children from 470 counties, but listwise deletion reduced the fully adjusted analytic samples to 8,059 for physical activity and 8,824 for motor competence.

Caregivers reported children’s activity in three daily periods on weekdays and weekends. For weekdays, the questions covered total activity inside and outside kindergarten; for weekends, they covered out-of-home activity, including movement-based classes. Reported minutes were converted to category midpoints, with the open-ended “more than 60 minutes” category assigned 75 minutes for the primary analysis, then combined into a daily average with a possible range of 0 to 225 minutes. The weighted mean was 114.5 minutes per day among the 9,665 children for whom activity could be scored. Motor competence was not measured with an observed performance test. Instead, parents rated children against same-age peers on overall athletic ability, speed in learning new movements, and a combined strength-endurance-agility measure, producing a three-to-15-point scale with a weighted average of 11.4 and an internal consistency of 0.81. This distinction is crucial: the two outcomes measured reported activity volume and caregiver perception rather than directly observed movement performance.

Parental support was assessed with a six-item scale combining beliefs, encouragement and enacted support. Examples included viewing activity as important, encouraging outdoor play, taking children to sport activities and playing actively with them at home. The scale, developed for the surveillance program through a three-round Delphi process with 83 senior experts, had an internal consistency of 0.76, and higher scores indicated stronger support. The researchers also included parenting orientation, persistence when children resisted activities, parental exercise, parental screen time, caregiver co-activity, parental obesity, home sports equipment, perceived neighborhood sport environment, child sociability and five demographic covariates: child age, sex, urban or rural residence, household income and parental education. All predictors were entered simultaneously in survey-weighted linear regressions, producing semistandardized coefficients and exact 95% confidence intervals.

For physical activity, each standard-deviation increase in child sociability was associated with 8.3 additional minutes per day, with a 95% confidence interval of 6.7 to 9.9. A standard-deviation increase in home sports equipment was associated with 8.0 additional minutes, with a 95% confidence interval of 6.5 to 9.5. Caregiver co-activity was associated with 5.9 additional minutes, and parental support with 4.5 minutes, with confidence intervals of 4.4 to 7.4 and 2.8 to 6.1, respectively. Direct comparisons did not distinguish sociability from equipment, or co-activity from support, because the relevant confidence intervals included zero. In the model’s relative-importance analysis, sociability, equipment and co-activity accounted for 27%, 27% and 17% of the variance assigned to the focal predictors. The full model explained 11.5% of the variance in reported activity.

Parent-rated motor competence showed a different adjusted pattern. Each standard-deviation increase in parental support was associated with 0.32 points on the three-to-15 scale, with a 95% confidence interval of 0.26 to 0.38. Child sociability was associated with 0.19 points, home sports equipment with 0.18 points and perceived neighborhood sport environment with 0.16 points. The support estimate exceeded the sociability estimate by 0.13 points and the equipment estimate by 0.14 points in direct contrasts. However, the six-item support scale included two enacted-support items that conceptually overlapped with the separate co-activity measure. When those items were removed, the support coefficient fell by about 30% to 0.22 points, with a 95% confidence interval of 0.17 to 0.28, and its differences from sociability and home equipment no longer excluded zero. The adjusted model accounted for 10.4% of the variance in parent-rated competence, and parental support received the largest share of the model’s explained variance among the focal predictors.

The researchers also used a cross-sectional statistical decomposition to ask how much of the support association could be accounted for by four measured parental behaviors: home sports equipment, caregiver co-activity, maternal exercise and paternal exercise. For physical activity, the total association per one-point increase in support was 13.8 minutes per day. Behaviors statistically accounted for 5.5 minutes, or 40%, of that association: 3.11 minutes through home equipment, 1.71 through co-activity and 0.86 through maternal exercise. The paternal-exercise component was close to zero. For motor competence, the behaviors accounted for 0.15 of 0.88 points, or 17%, leaving 0.73 points not statistically accounted for by those measures. The reduced-overlap support score produced similar proportions of 32% and 20%, respectively. These are partitions of statistical associations, not tests of mediation or evidence of a causal pathway.

Exploratory moderation analysis produced one notable result. The association between parental support and physical activity was stronger as parental education increased, by 2.2 minutes per day per standard deviation of support for each standard deviation of education. The interaction remained significant after Benjamini–Hochberg correction, with an adjusted p value of 0.013. Urban and higher-income interactions were significant only at the nominal level and did not survive correction. None of six interactions for motor competence was significant. Because these tests were not preregistered, the researchers classified them as exploratory and called for prospective replication. Nonlinear analyses also indicated that the relationships between support and parent-rated motor competence, and between co-activity and physical activity, were mildly curved but monotonic within the observed ranges. Adjusted motor-competence estimates rose from 10.80 points at the 10th percentile of support to 12.24 at the 90th percentile, while adjusted physical activity rose from 105.2 minutes per day at the 10th percentile of co-activity to 130.5 minutes at the 90th.

Several limitations constrain interpretation. The design was cross-sectional, so it could not establish temporal order or causal direction. Caregivers supplied information about their own support, the family environment, their child’s activity and motor competence, creating a substantial possibility of shared-rater and common-method bias. A caregiver who perceived a child as active might also describe the child as more sociable or rate the family as more supportive. Missing responses removed some participants from the main models, and children excluded from those analyses tended to come from less advantaged and less supportive households. Multiple imputation and inverse-probability weighting reproduced the principal estimates, but neither can correct selection on unmeasured factors. Activity was reconstructed from broad categories, parental obesity was a self-reported yes-or-no measure, and motor competence was never assessed objectively. The pandemic conditions under which the survey was fielded had also reshaped household routines, although that context was shared across the sample.

Some counterintuitive findings also require caution. Longer parental screen time was associated with more child physical activity, with adjusted estimates of 3.8 minutes per day for fathers and 2.8 for mothers, and the associations persisted in sensitivity analyses. The study provides no evidence that screen exposure benefits activity; unmeasured household characteristics or reporting patterns could explain the result. Reported parental obesity was not associated with either primary outcome. The weighted sample was 54.2% boys and 61.4% urban, with a mean age of 65.8 months. Overall, the findings show that activity volume and perceived motor skill are related to overlapping but distinguishable family characteristics. The authors conclude that longitudinal studies with independently measured activity and motor skills are needed to determine whether parental support, equipment, co-activity and child disposition precede changes in movement behavior and competence.

Subject of Research: Cross-sectional associations between family factors, preschool physical activity and parent-rated motor competence in China

Article Title: Family correlates of physical activity and parent-rated motor competence in preschool children from a socioecological perspective: a cross-sectional analysis of 10,967 children in China

Article References: Zhang, Y., Pan, X., Gao, Y., Li, M., Jiang, L., Zhao, D., Chen, J., Wang, C., Zhao, X., & Hu, H. (2026). Family correlates of physical activity and parent-rated motor competence in preschool children from a socioecological perspective: a cross-sectional analysis of 10,967 children in China. Frontiers in Psychology, 17, Article 1898292. https://doi.org/10.3389/fpsyg.2026.1898292

Image Credits: AI Generated

DOI: 10.3389/fpsyg.2026.1898292

Keywords: preschool children, physical activity, motor competence, parental support, family environment, home sports equipment, co-activity, child sociability, China, cross-sectional study, Family, correlates