Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Early-life activities mediate the association between family socioeconomic status in early childhood and physical fitness in early adolescence


The graded association between family socioeconomic status (SES) and physical fitness is evident, but little is known about the mechanism underlying this association. This study investigated the role of early-life activities as mediators of the longitudinal relationship between early-life SES and health-related physical fitness in 168 adolescents (51.2% boys; final mean age: 12.4 years old). In Wave 1 (2011–12), their parents completed questionnaires about family socioeconomic status (SES), parent–child activities, and child screen time. In Wave 2 (2014–15), participants’ physical activity levels were assessed through parent proxy-reports. In Wave 3 (2018–19), a direct assessment of handgrip strength, standing long-jump, and 6-min walk test (6MWT) performance was conducted. After controlling for demographic factors, results of mediation analyses revealed that (a) Wave 1 SES predicted Wave 3 long-jump and 6MWT performance; (b) child physical activity level in Wave 2 mediated the relation between Wave 1 SES and standing long-jump performance in Wave 3; and (c) recreational parent–child activities and child screen time in wave 1 mediated the relation between Wave 1 SES and 6MWT performance in Wave 3. Our findings suggest that the type and frequency of early-life activities play a role in the graded association between childhood SES and physical fitness in adolescence.


Family socioeconomic status (SES) is a composite measure of parental education, occupation and income, all of which are known to impact child development1. It has been reported that children from low-SES families were more difficult to reach their full potential due to the lack of nurturing care and exposure to developmentally appropriate activities, which was associated with poorer motor development2. Motor development is often assessed by levels of physical fitness3, which can be subdivided into health-related physical fitness and performance-related fitness. Physical fitness includes cardiorespiratory endurance, muscle strength, body composition and flexibility, whereas performance-related fitness reflects one’s balance, coordination, speed and agility4. For example, handgrip strength is a good indicator of the maximal isometric hand strength and upper limb muscle strength5,6; standing long-jump performance is reflective of lower limb explosive strength capacity7; and 6-min walk test (6MWT) performance is a reliable and valid measure of muscle tolerance and endurance8,9. It has been reported that children and adolescents from high-SES families were more likely to have higher mean fitness and less likely to have low/poor fitness10.

Physical activity as a potential mediator of the association between SES and physical fitness

Previous longitudinal research has demonstrated a positive reciprocal relationship between physical activity and physical fitness across childhood through early adolescence11. A systematic review and meta-analysis of 23 studies also demonstrated a positive association between vigorous-intensity physical activity and cardiorespiratory fitness among children and adolescents12. Given the health benefits of physical activities, guidelines from the World Health Organization recommend children and adolescents to engage in at least 60 min of moderate- to vigorous-intensity physical activity each day13. However, individuals from low-SES backgrounds were found to have a higher likelihood of being physically inactive14. Children from lower-SES families may also have fewer access to outdoor play equipment such as bicycles and jump ropes and green areas15,16. The lack of these outdoor play opportunities in turn could have negative effects on motor development17,18.

Recreational activities as a potential mediator of the association between SES and physical fitness

Although the link between physical activity and physical fitness is evident, much less is known about the potential effect of other recreational activities on physical fitness. Notably, in early childhood, physical activities are usually unstructured and performed through engagement in other types of recreational activities such as singing and storytelling. A previous study showed that participation in a developmentally appropriate music and movement program was associated with improvement in jumping and dynamic balance among preschoolers19. On the other hand, while drawing activities may not have direct effects on gross motor skill development, previous studies found that drawing skills were associated with executive functions such as inhibition and fine motor control20. These cognitive processes are also involved in exercise training and may relate to exercise motivation through self-regulation behavior such as the ability to start and stick to a workout routine amid other distractions21. For example, electronic devices are increasingly ubiquitous in today’s society and frequently used for online learning and entertainment purposes such that time for movement and play could be limited even in the preschool years22. Previous studies have demonstrated that frequent use of electronic devices is associated with a reduction in parent–child interaction time23. While health authorities recommend children and adolescents to limit their screen time to two hours or less per day24, evidence suggests that many Chinese children cannot meet this recommendation25,26. The risk of overuse of electronic devices is higher among children from disadvantaged families27. Electronic devices are also found more often in the bedroom of children from low-SES families28. Given that physical fitness involves multiple components, it is unclear whether different types of early-life activities would have similar or differential effects on these physical fitness components.

The present study

Although there is evidence documenting the association between family SES and physical fitness, its underlying mechanism remains elusive with limited evidence on the predictive value of early-life activities for physical fitness in adolescence. It seems plausible that the impact of early-life SES on motor development may vary by the frequency and type of childhood activities. Hence, this study aimed to fill in this knowledge gap by examining the role of childhood activities in explaining the relationship between early-life SES and adolescent physical fitness.


Descriptive statistics

Descriptive statistics for the study sample are presented in Table 1. They were 12.39 years old on average with 51.2% as boys and 48.8% as girls. A total of 65 (38.7%) had doctor-diagnosed physical diseases. In Wave 1, the average monthly household income of the study sample was USD 6282; 95% of the fathers and 60% of the mothers reported having a job. Regarding parental education level, 78% of the fathers and 79% of the mothers completed secondary education or above. In Wave 3, the mean (and SD) of handgrip strength was 35.59 ± 9.22 kg, whereas the mean (and SD) of 6MWT and standing long-jump performance were 600.58 ± 83.05 m and 130.29 ± 22.63 cm, respectively.

Table 1 Subject characteristics (n = 168).

Associations between Wave 1 family SES and Wave 3 fitness test performance

Results showed that higher Wave 1 family SES predicted better 6MWT performance in Wave 3 in both unadjusted (β(95%CI) = 13.62(0.74, 26.50), p = 0.038) and adjusted (β(95%CI) = 14.54(1.60, 27.48), p = 0.028) models. After adjusting for child age, gender, and history of physical disease, higher Wave 1 family SES predicted better standing long-jump performance in Wave 3 (β(95%CI) = 3.97(0.39, 7.56), p = 0.030). The association between Wave 1 family SES and Wave 3 handgrip strength performance was not significant before and after adjusting for the demographic confounders.

Bivariate correlations

In addition to the significant correlations with Wave 1 family SES (Table 2), when examining associations between early-life activities and fitness levels, better 6MWT performance was associated with higher CPCIS—recreational activities score (r = 0.19, p < 0.05) and less screen time (r = -0.20, p < 0.05) in Wave 1. Better standing long-jump performance correlated with higher PARCY level in Wave 2 (r = 0.35, p < 0.001), whereas handgrip strength performance did not significantly correlate with any of the early-life activities.

Table 2 Correlation matrix for study variables.

Pathways from Wave 1 family SES to Wave 3 standing long-jump performance

After adjusting for confounders, higher Wave 1 family SES predicted better long-jump performance in Wave 3 (β(95%CI) = 4.29(0.68, 7.89), p = 0.020). Figure 1 illustrates the path model from Wave 1 family SES to Wave 3 long-jump performance, and Table 3 provides the effect estimates for each path. We found that after adjusting for Wave 2 PARCY level, the direct effect of Wave 1 family SES on Wave 3 long-jump performance was not significant (β(95%CI) = 2.52 (-1.17, 6.17), p = 0.181). On the other hand, the indirect effect of Wave 1 family SES on Wave 3 long-jump performance through Wave 2 PARCY level was significant (β(95%CI) = 1.77 (0.58, 3.63), p = 0.021). Specifically, higher Wave 1 family SES predicted an increase in Wave 2 physical activity level which in turn led to better Wave 3 long-jump performance.

Figure 1
figure 1

Path model linking family SES at Wave 1 to standing long-jump test performance at Wave 3.

Table 3 Path estimates for the model linking family SES at Wave 1 and standing long-jump test performance at Wave 3.

Pathways from Wave 1 family SES to Wave 3 6MWT performance

After adjusting for confounders, higher Wave 1 family SES predicted better 6MWT performance in Wave 3 (β(95%CI) = 14.54(1.75, 27.63), p = 0.028). Figure 2 illustrates the path model from Wave 1 family SES to Wave 3 6MWT performance, and Table 4 provides the effect estimates for each path. The model involving two mediators showed significant indirect effects (β(95%CI): 5.24(1.80, 10.32), p = 0.014), but the direct effect of Wave 1 family SES on 6MWT performance in Wave 3 was not significant (β (95%CI) = 9.30(-3.55, 22.70), p = 0.170). In the fitness-enhancing pathway, higher family SES was associated with higher frequency of recreational parent–child activities in Wave 1 which in turn predicted better 6MWT performance in Wave 3. In the fitness-depressing pathway, lower family SES was associated with more screen time in Wave 1 which in turn predicted worse 6MWT performance in Wave 3.

Figure 2
figure 2

Path model linking family SES at Wave 1 to 6-min walk test performance at Wave 3.

Table 4 Path estimates for the model linking family SES at Wave 1 and 6-min walk test performance test performance at Wave 3.


Using a prospective longitudinal design, this research is unique in examining the influence of early-life SES on health-related physical fitness and the role of childhood activities as potential mediators of this relationship. We found that a higher level of family SES in early childhood predicted a higher level of lower limb explosive strength (as reflected by better standing long-jump performance) and a higher level of lower limb muscle endurance (as reflected by better 6MWT performance) in early adolescence. While the effect of early-life SES on lower limb explosive strength was mediated by physical activity level in middle childhood, its effect on lower limb muscle endurance occurred through recreational parent–child activities and screen time in early childhood. By contrast, the effect of early-life SES on upper limb muscle strength (as reflected by handgrip strength performance) was small and not significant. Our results indicate that children from lower-SES families could have a higher likelihood of engaging in sedentary activities such as screen use over recreational and physical activities, which in turn can lead to poor physical fitness in adolescence.

While larger studies are needed to confirm the present results, the finding of early-life activities as mediators of the association between early-life SES and standing long-jump and 6MWT performance suggests that non-screen recreational activities are conducive to the training of lower limb explosive strength and muscle endurance. However, these recreational opportunities could be more limited among children from lower-SES families15,16. Effective home- and school-based lower limb muscle training programs would likely benefit underprivileged children. By contrast, the finding of no SES differences in handgrip strength performance suggests that improvement in upper limb muscle strength can be achieved irrespective of SES background. Future studies may consider exploring the effect of routine muscle-strengthening activities29 such as moving, lifting or carrying heavy things on upper limb muscle strength.

Previous research has reported that youth from higher-SES backgrounds were more likely to maintain regular sport participation30. Consistent with this finding, we observed higher physical activity level in middle childhood among those students from higher-SES families in early childhood. Also consistent with the previous finding of insufficient parental involvement in low-SES families31, adolescents from higher-SES families in this study were found to perform more recreational activities with parents in early childhood. Although this study did not measure the frequency of sport and physical activities in early childhood, the finding of a positive correlation between recreational parent–child activities in preschool years and physical activity level in primary school years adds to the literature that physical activity habits can be nurtured through engagement in general play-based parent–child activities at early ages. These activities can increase children’s awareness and understanding of the extent to which their parents are able and willing to support their hobbies and interests which is a key factor in promoting children’s physical activity level32.

In addition, it is generally agreed that activities in the early life stages have lasting effects on activity behaviors and developmental outcomes33. The associations of sedentary screen time with reduced muscle endurance and increased risk of health problems are evident34. Previous research also found an association between daily screen time at age 1 year and screen time later in life35. Our study provides additional evidence that early-life play-based activities can contribute to later gains in lower limb muscle endurance. This echoes the findings of a previous study reporting that participation in a range of developmentally appropriate, enjoyable and safe play-based activities through the early years is associated with better cardiorespiratory and musculoskeletal fitness and motor development36. It should be noted that many sports and fitness programs are effective in improving jumping performance and lower limb explosive strength37. We also found that physical activity level in middle childhood was predictive of lower limb explosive strength, but children from lower-SES families could have lower physical activity levels. School-based health education programs should be implemented to help children from disadvantaged backgrounds to gain access to physical activity38.

A major strength of our study was the use of longitudinal data across three time points from early childhood through adolescence which allowed for tracking the effect of early activities on long-term health-related outcomes. In addition, direct assessments were administered to measure three major health-related physical fitness components (upper limb muscle strength, lower limb muscle endurance, and lower limb explosive strength). This provides a comprehensive set of objective fitness data to increase the robustness of the present results. However, since the outcome, exposure, and mediator measures were collected in different waves, this study cannot infer causality, but temporality. Furthermore, we used parent-proxy reports to measure childhood activities, which could be subject to recall bias. Future research should use multiple-informants or direct observation methods to collect data on childhood activities. Another limitation is that we did not measure physical activity level in Wave 1, and thus we cannot draw any conclusion on the importance of recreational physical activity in early childhood as a potential predictor of physical fitness in adolescence. Furthermore, our study sample was not representative of Hong Kong population, as they came from only two local districts. Nonetheless, despite small sample size, significant relationships were still observed, which could give a helpful future direction for further research.

In conclusion, we found significant associations between family SES in early childhood and different components of health-related physical fitness in early adolescence. These associations were mediated by different types of childhood activities. Children from low-SES backgrounds, compared to the more affluent groups, could have weaker lower limb muscles due to insufficient activity levels. School-based programs incorporating lower limb muscle training should be encouraged. More studies are also needed to explore other mechanisms underlying the socioeconomic gradient in physical fitness using robust strategies such as large samples, objective activity measures, and multiple informants.



This study was part of a longitudinal cohort project aiming to examine the impact of early-life family SES on long-term health and development39. During the first year of project (2011), 22 kindergartens were randomly selected from two districts of Hong Kong (Hong Kong Island (an affluent district with about 41 kindergartens in total) and Yuen Long (a less wealthy district with about 71 kindergartens in total)), and 20 kindergartens (9 from Hong Kong Island and 11 from Yuen Long) agreed to participate. All students in the final year (K3) of these selected schools were eligible and invited to take part in the project. Three waves of data collection have been conducted. In Wave 1 when the child was in K3 (2011–12), 492 parents provided data on family demographics, screen time, and frequency of parent–child activities. In Wave 2 when the child was in Grade 3 (2014–15), 386 reported children’s physical activity level. In Wave 3 when the child was in Grade 7 (2018–19), 168 children performed the physical fitness tests. In Wave 1, the cohort was aged 5 to 6 years and 12 to 13 years in Wave 3. Only data of children who completed both Wave 1 and Wave 3 assessments were analyzed in this study. The flow chart is shown in Fig. 3. Written informed consents from parents/guardians were obtained prior to recruitment and data collection. Wave 3 completers and dropouts were similar in terms of children’s age and gender and parental education level and occupation and monthly household income in Wave 1.

Figure 3
figure 3

Study flow chart.


Family socioeconomic status (Wave 1)

Family SES was analyzed as a family SES index z-score39. In Wave 1, information on family SES was collected via sociodemographic questionnaire and included items pertaining to four key SES factors: (1) maternal and paternal educational attainment; (2) maternal and paternal occupation; (3) family monthly income adjusted for household size; and (4) family asset score. A principal component analysis with varimax rotation was employed to combine these four SES-factors into a one-factor variable40. This SES index calculation method has been used in other local community and population studies39,41,42.


Health-related physical fitness (Wave 3)

Three fitness tests were carried out during a health check-up: (i) handgrip test, (ii) 6MWT, and (iii) standing long-jump test. A trained research assistant first guided the adolescent participants how to perform each test, and the other research assistant recorded their performance. The participants performed each test separately from each other according to the given instructions as described below:

Upper limb muscle strength was measured in kilograms using a handgrip dynamometer (TKK 5101 Grip D; Takey, Tokyo, Japan; precision 0.1 kg). Each student squeezed gradually and continuously for at least 2 s while standing with arms straight down the side and without the dynamometer touching any part of the body except the hand being measured. The test was performed on each hand twice in a random order. The best measurement for each hand was recorded and the sum of the best measurement scores achieved by both hands was used in the analysis43. This is a reliable and valid procedure for both healthy and clinical populations with an intraclass correlation coefficient ranging from 0.92 to 0.9644.

Standing long-jump test with two-foot taking-off and landing was carried out to measure the explosive strength of lower limb with an acceptable reliability (intra-difference close to 0)45. Participants were asked to stand with feet slightly apart behind a line marked on the floor and to jump forward as far as possible. Three attempts were allowed. The best distance between the take-off line and the point where the back of heel is landing on was used in the analysis46.

Lower limb muscle endurance was assessed by the 6MWT. Participants were asked to walk as far as possible along a 30-m office corridor for 6 min. The 6-min walk distance was measured once and recorded in meters47. This test is considered reliable and valid for assessing exercise tolerance and endurance with an intraclass correlation coefficient of 0.9448.


Recreational parent–child activities (Wave 1)

The Chinese Parent–Child Interaction Scale (CPCIS)—Recreational Activities subscale, a validated parent–child interaction measure in Hong Kong with good internal consistency (Cronbach’s alpha = 0.71)49, was completed by the parent in Wave 1 to assess the weekly frequency of the following parent–child activities: reading, drawing, singing, storytelling, and discussing news and current affairs39. Items were rated on a 4-point scale ranging from 0 = none to 3 = 4 times or above per week and averaged to generate an overall mean score (M = 1.82, SD = 0.62), with higher scores representing more frequent recreational parent–child activities.

Screen time (Wave 1)

Children’s screen time was reported by a parent in Wave 1, with a screen time questionnaire designed to assess time spent watching television, video gaming on handheld and/or other types of game consoles, and using computers, tablet computers and smartphones for studying, gaming and/or web browsing. The child’s average daily screen time was calculated by averaging the amount of time reported for weekends and weekdays using the weighted average formula ([2 × weekend + 5 × weekday]/7) and reported in hours (M = 2.25, SD = 1.84). This calculation method has been used in previous community studies50,51.

Physical activity level (Wave 2)

In Wave 2, a parent completed the Physical Activity Rating Questionnaire for Children and Youth (PARCY) to report the child’s physical activity level on a 11-point scale ranging from 0 = no exercise at all in the last year to 10 = doing vigorous exercise almost every day in the last year (M = 5.34, SD = 2.19). The questionnaire consists of one assessment item that takes into consideration the physical activity frequency, duration, and intensity in the past year. It has been used in previous community and clinical epidemiological studies, showing good content validity index (90%) and test–retest reliability (0.86)41,52.


Children’s gender and date of birth were collected using the parent-completed sociodemographic questionnaire in Wave 1. During the health check-up in Wave 3, parents reported whether their children had any physician-diagnosed physical diseases. Children’s age at assessment was computed by subtracting date of birth from date of assessment. Informed by previous research53, age, gender, and history of physical diseases were associated with family SES and fitness levels and thus included as potential confounders in the mediation models.

Data analysis

Descriptive statistics were first generated in SPSS version 25 (IBM corp, Armonk, NY, USA) to describe sample characteristics. In our analytic sample, missing data ranged from 1.19 to 11.31%. Assuming data were missing at random, we imputed missing data (5 imputed datasets) using multiple imputation by chained equations (MICE) R package54. All variables selected for analysis models were used to predict missing data. In addition to checking for missing values, continuous data were also checked for outliers and deviations from normality. Linear regression analyses were then performed to examine the unadjusted and adjusted relations between Wave 1 family SES and Wave 3 fitness test performance. Pearson’s correlations were also computed to assess the relationships between early-life family SES and activities and adolescent fitness outcomes. In order to test the hypothesis that the relations between early-life family SES and adolescent fitness outcomes were mediated by childhood activities, mediation models were performed, using Lavaan 0.6–555 in R 3.6.3, on the basis of the regression and correlation results. Mediators in the model were Wave 1 screen time, Wave 1 CPCIS—recreational activities and Wave 2 PARCY level. All paths were adjusted for the potential influence of covariates in the model. In addition, we controlled for Wave 1 screen time, Wave 1 CPCIS—recreational activities or Wave 2 PARCY level depending on the model tested. The indirect effect is estimated by the product of the effect of Wave 1 family SES on the mediator and the effect of the mediator on the fitness outcome, given the direct effect of Wave 1 family SES and other covariates (Figs. 1 and 3). The indirect and direct effects and their 95% CI were estimated with 5000 bootstrap samples using the bias-corrected percentile method in Lavaan55. All the regression estimates (β) were unstandardized, with a p-value < 0.05 denoting statistical significance.

Sample size estimation

Power analysis was conducted assuming two potential mediators. A path model with two potential mediators and small- to medium-sized path estimates (r = 0.25) would require a sample size of 146 to detect the two indirect effects at 0.05 significance level and 80% statistical power56.

Ethical approval

The study protocol was approved by the Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster (UW 18-057). All research activities were performed in accordance with the Declaration of Helsinki. Informed consent was obtained from the parents or legal guardians of all participants.



Socioeconomic status


6-Minute walk test


Chinese Parent–Child Interaction Scale


Physical Activity Rating Questionnaire for Children and Youth


  1. Bradley, R. H. & Corwyn, R. F. Socioeconomic status and child development. Annu. Rev. Psychol. 53, 371–399 (2002).

    Article  Google Scholar 

  2. McPhillips, M. & Jordan-Black, J.-A. The effect of social disadvantage on motor development in young children: A comparative study. J. Child Psychol. Psychiatry 48, 1214–1222. (2007).

    Article  PubMed  Google Scholar 

  3. Goodway, J. D., Ozmun, J. C. & Gallahue, D. L. Understanding Motor Development: Infants, Children, Adolescents, Adults. (Jones & Bartlett Learning, 2019).

  4. Caspersen, C. J., Powell, K. E. & Christenson, G. M. Physical activity, exercise, and physical fitness: Definitions and distinctions for health-related research. Public Health Rep. 100, 126–131 (1985).

    CAS  PubMed  PubMed Central  Google Scholar 

  5. Gómez-Campos, R. et al. Normative data for handgrip strength in children and adolescents in the Maule Region, Chile: Evaluation based on chronological and biological age. PloS one 13 (2018).

  6. Yin, L., Tang, C. & Tao, X. Criterion-related validity of a simple muscle strength test to assess whole body muscle strength in Chinese children aged 10 to 12 years. Biomed. Res. Int. (2018).

    Article  PubMed  PubMed Central  Google Scholar 

  7. Chung, L. M. Y., Chow, L. P. Y. & Chung, J. W. Y. Normative reference of standing long jump indicates gender difference in lower muscular strength of pubertal growth. Health 5, 6–11 (2013).

    Article  Google Scholar 

  8. Lammers, A. E., Hislop, A. A., Flynn, Y. & Haworth, S. G. The 6-minute walk test: Normal values for children of 4–11 years of age. Arch. Dis. Child. 93, 464–468 (2008).

    CAS  Article  Google Scholar 

  9. Li, A. M. et al. Standard reference for the six-minute-walk test in healthy children aged 7 to 16 years. Am. J. Respir. Crit. Care Med. 176, 174–180 (2007).

    Article  Google Scholar 

  10. Wolfe, A. M., Lee, J. A. & Laurson, K. R. Socioeconomic status and physical fitness in youth: Findings from the NHANES National Youth Fitness Survey. J. Sports Sci. 38, 534–541. (2020).

    Article  PubMed  Google Scholar 

  11. Lima, R. A. et al. Physical activity and motor competence present a positive reciprocal longitudinal relationship across childhood and early adolescence. J. Phys. Activity Health 14, 440–447 (2017).

    Article  Google Scholar 

  12. García-Hermoso, A., Ezzatvar, Y., Ramírez-Vélez, R., Olloquequi, J. & Izquierdo, M. Is device-measured vigorous physical activity associated with health-related outcomes in children and adolescents? A systematic review and meta-analysis. J. Sport Health Sci. 10, 296–307. (2021).

    Article  PubMed  Google Scholar 

  13. World Health Organisation. Physical Activity and Young People.

  14. Lee, P. H., Yu, Y.-Y., McDowell, I., Leung, G. M. & Lam, T. H. A cluster analysis of patterns of objectively measured physical activity in Hong Kong. Public Health Nutr. 16, 1436–1444. (2013).

    CAS  Article  PubMed  Google Scholar 

  15. Rydenstam, T., Fell, T., Buli, B. G., King, A. C. & Bälter, K. Using citizen science to understand the prerequisites for physical activity among adolescents in low socioeconomic status neighborhoods—The NESLA study. Health Place 65, 102387. (2020).

    Article  PubMed  Google Scholar 

  16. Kantomaa, M. T., Tammelin, T. H., Näyhä, S. & Taanila, A. M. Adolescents’ physical activity in relation to family income and parents’ education. Prev. Med. 44, 410–415 (2007).

    Article  Google Scholar 

  17. Becker, D. R., Grist, C. L., Caudle, L. A. & Watson, M. K. Complex physical activities, outdoor play, and school readiness among preschoolers. Global Educ. Rev. 5, 110–122 (2018).

    Google Scholar 

  18. Ahmetoglu, E. The contributions of familial and environmental factors to children’s connection with nature and outdoor activities. Early Child Dev. Care 189, 233–243. (2019).

    Article  Google Scholar 

  19. Zachopoulou, E., Tsapakidou, A. & Derri, V. The effects of a developmentally appropriate music and movement program on motor performance. Early Childh. Res. Q. 19, 631–642. (2004).

    Article  Google Scholar 

  20. Simpson, A. et al. Fine motor control underlies the association between response inhibition and drawing skill in early development. Child Dev. 90, 911–923. (2019).

    Article  PubMed  Google Scholar 

  21. Hagger, M. S., Wood, C. W., Stiff, C. & Chatzisarantis, N. L. D. Self-regulation and self-control in exercise: the strength-energy model. Int. Rev. Sport Exerc. Psychol. 3, 62–86. (2010).

    Article  Google Scholar 

  22. Woo, E. H., White, P. & Lai, C. W. Musculoskeletal impact of the use of various types of electronic devices on university students in Hong Kong: An evaluation by means of self-reported questionnaire. Man. Ther. 26, 47–53 (2016).

    Article  Google Scholar 

  23. Bozzola, E. et al. Media devices in pre-school children: The recommendations of the Italian pediatric society. Ital. J. Pediatr. 44, 69 (2018).

    Article  Google Scholar 

  24. Huang, W. Y. et al. Results from Hong Kong’s 2018 report card on physical activity for children and youth. J. Phys. Activity Health 15, S370–S372 (2018).

    Article  Google Scholar 

  25. Wu, X. et al. The relationship between screen time, nighttime sleep duration, and behavioural problems in preschool children in China. Eur. Child Adolesc. Psychiatry 26, 541–548 (2017).

    Article  Google Scholar 

  26. Lam, J. W. K., Sit, C. H. P. & Cerin, E. Physical activity and sedentary behaviours in Hong Kong primary school children: Prevalence and gender differences. Prev. Med. 51, 96–97. (2010).

    Article  PubMed  Google Scholar 

  27. Tandon, P. S. et al. Home environment relationships with children’s physical activity, sedentary time, and screen time by socioeconomic status. Int. J. Behav. Nutr. Phys. Act. 9, 88 (2012).

    Article  Google Scholar 

  28. Fu, K.-W. et al. Parental restriction reduces the harmful effects of in-bedroom electronic devices. Arch. Dis. Child. 102, 1125–1131 (2017).

    Article  Google Scholar 

  29. Freeston, J. et al. Associations between multiple indicators of socio-economic status and muscle-strengthening activity participation in a nationally representative population sample of Australian adults. Prev. Med. 102, 44–48. (2017).

    Article  PubMed  Google Scholar 

  30. Kwon, S., Janz, K. F., Letuchy, E. M., Burns, T. L. & Levy, S. M. Parental characteristic patterns associated with maintaining healthy physical activity behavior during childhood and adolescence. Int. J. Behav. Nutr. Phys. Act. 13, 58. (2016).

    Article  PubMed  PubMed Central  Google Scholar 

  31. Li, X., Yang, H., Wang, H. & Jia, J. Family socioeconomic status and home-based parental involvement: A mediation analysis of parental attitudes and expectations. Child Youth Serv. Rev. 116, 105111. (2020).

    Article  Google Scholar 

  32. Gustafson, S. L. & Rhodes, R. E. Parental correlates of physical activity in children and early adolescents. Sports Med. 36, 79–97 (2006).

    Article  Google Scholar 

  33. Figueroa, R. & An, R. Motor skill competence and physical activity in preschoolers: A review. Matern. Child Health J. 21, 136–146. (2017).

    Article  PubMed  Google Scholar 

  34. Domingues-Montanari, S. Clinical and psychological effects of excessive screen time on children. J. Paediatr. Child Health 53, 333–338. (2017).

    Article  PubMed  Google Scholar 

  35. Xu, H., Wen, L. M., Hardy, L. L. & Rissel, C. A 5-year longitudinal analysis of modifiable predictors for outdoor play and screen-time of 2- to 5-year-olds. Int. J. Behav. Nutr. Phys. Act. 13, 96. (2016).

    Article  PubMed  PubMed Central  Google Scholar 

  36. Okely, A. D. et al. A collaborative approach to adopting/adapting guidelines—The Australian 24-Hour Movement Guidelines for the early years (Birth to 5 years): An integration of physical activity, sedentary behavior, and sleep. BMC Public Health 17, 869. (2017).

    Article  PubMed  PubMed Central  Google Scholar 

  37. Sedano, S., Matheu, A., Redondo, J. C. & Cuadrado, G. Effects of plyometric training on explosive strength, acceleration capacity and kicking speed in young elite soccer players. J. Sports Med. Phys. Fitness 51, 50–58 (2011).

    CAS  PubMed  Google Scholar 

  38. Dai, C.-L. School health program: Impacting physical activity behaviors among disadvantaged students. J. Sch. Health 89, 468–475. (2019).

    Article  PubMed  Google Scholar 

  39. Ip, P. et al. Socioeconomic gradients in school readiness of Chinese preschool children: The mediating role of family processes and kindergarten quality. Early Childh. Res. Q. 35, 111–123. (2016).

    Article  Google Scholar 

  40. Vyas, S. & Kumaranayake, L. Constructing socio-economic status indices: How to use principal components analysis. Health Policy Plan. 21, 459–468 (2006).

    Article  Google Scholar 

  41. Ho, F. K. W., Louie, L. H. T., Chow, C. B., Wong, W. H. S. & Ip, P. Physical activity improves mental health through resilience in Hong Kong Chinese adolescents. BMC Pediatr. 15, 48. (2015).

    Article  PubMed  PubMed Central  Google Scholar 

  42. Lo, C. K. M. et al. Prevalence of child maltreatment and its association with parenting style: A population study in Hong Kong. Int. J. Environ. Res. Public Health 16, 1130 (2019).

    Article  Google Scholar 

  43. Roberts, H. C. et al. A review of the measurement of grip strength in clinical and epidemiological studies: Towards a standardised approach. Age Ageing 40, 423–429 (2011).

    Article  Google Scholar 

  44. Bobos, P., Nazari, G., Lu, Z. & MacDermid, J. C. Measurement properties of the hand grip strength assessment: A systematic review with meta-analysis. Arch. Phys. Med. Rehabil. 101, 553–565. (2020).

    Article  PubMed  Google Scholar 

  45. Ortega, F. B. et al. Reliability of health-related physical fitness tests in European adolescents. The HELENA Study. Int. J. Obes. 32, S49–S57. (2008).

    Article  Google Scholar 

  46. Castro-Piñero, J. et al. Percentile values for muscular strength field tests in children aged 6 to 17 years: Influence of weight status. J. Strength Cond. Res. 23, 2295–2310 (2009).

    Article  Google Scholar 

  47. American Thoracic Society. ATS statement. Am. J. Respir. Crit. Care Med. 166, 111–117. (2002).

    Article  Google Scholar 

  48. Li, A. M. et al. The six-minute walk test in healthy children: Reliability and validity. Eur. Respir. J. 25, 1057–1060. (2005).

    CAS  Article  PubMed  Google Scholar 

  49. Ip, P. et al. Rasch validation of the Chinese parent–child interaction scale (CPCIS). World J. Pediatr. 14, 238–246 (2018).

    Article  Google Scholar 

  50. Wong, R. S. et al. A longitudinal study of the relation between childhood activities and psychosocial adjustment in early adolescence. Int. J. Environ. Res. Public Health 18, 5299 (2021).

    Article  Google Scholar 

  51. Wong, R. S. et al. Parent technology use, parent-child interaction, child screen time, and child psychosocial problems among disadvantaged families. J. Pediatr. 226, 258–265. (2020).

    Article  PubMed  Google Scholar 

  52. Chung, O. K. J., Li, H. C. W., Chiu, S. Y., Ho, K. Y. E. & Lopez, V. The impact of cancer and its treatment on physical activity levels and behavior in Hong Kong Chinese childhood cancer survivors. Cancer Nurs. 37 (2014).

  53. Lindgren, M. et al. Physical activity pattern, cardiorespiratory fitness, and socioeconomic status in the SCAPIS pilot trial — A cross-sectional study. Prevent. Med. Rep. 4, 44–49. (2016).

    Article  Google Scholar 

  54. Buuren, S. V. & Groothuis-Oudshoorn, K. mice: Multivariate imputation by chained equations in R. J. Stat. Softw. 1–68 (2010).

  55. Rosseel, Y. Lavaan: An R package for structural equation modeling and more. Version 0.5–12 (BETA). J. Stat. Softw. 48, 1–36 (2012).

    Article  Google Scholar 

  56. Schoemann, A. M., Boulton, A. J. & Short, S. D. Determining power and sample size for simple and complex mediation models. Soc. Psychol. Pers. Sci. 8, 379–386. (2017).

    Article  Google Scholar 

Download references


We thank the schools, children, and parents for participation in this study. The cohort study was supported by three research grants from the Research Grants Council of the Hong Kong Special Administrative Region, China (Project Nos. 743413, 746111 and 17606318).


The cohort study was supported by three research grants from the Research Grants Council of the Hong Kong Special Administrative Region, China (Project Nos. 743413, 746111 and 17606318).

Author information




R.S.W. and K.T.S.T. contributed equally to data interpretation, drafted the initial manuscript and performed major parts of the statistical analysis. B.N.K.C. interpreted the results of the data and drafted the initial manuscript. F.K.W.H. and W.H.S.W. contributed to overall study coordination and data interpretation and critically revised the manuscript. N.R., K.L.C., J.S., H.K.S., W.W.Y.T., J.C.S.Y. and I.C.K.W. contributed to data interpretation and critically revised the manuscript. P.I. contributed to the overall study design and data interpretation and critically revised the manuscript. All authors have approved the final manuscript as submitted.

Corresponding author

Correspondence to Patrick Ip.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher's note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Wong, R.S., Tung, K.T.S., Chan, B.N.K. et al. Early-life activities mediate the association between family socioeconomic status in early childhood and physical fitness in early adolescence. Sci Rep 12, 81 (2022).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI:


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing