Repository logo
 

Evaluating the relationship between physical activity, gross motor skills and healthy growth in preschoolers using structural equation modeling

Date

2017

Authors

Courtney, Jimikaye B, author
Grimm, Kevin, author
Boles, Richard E, author
Johnson, Susan L, author
Bellows, Laura L, author

Journal Title

Journal ISSN

Volume Title

Abstract

Purpose: The purpose of this study was to use structural equation modeling (SEM) to understand the directionality of the relationship between physical activity (PA) and gross motor skills (GMS) in predicting healthy growth (body mass index, (BMI)) in preschoolers. Procedure: Baseline data from preschoolers (4-5 years old) enrolled in the Colorado Longitudinal Eating And Physical activity (LEAP) study were used. Preschoolers were assessed on GMS (balance, locomotor and ball skills) and fitness (Bruininks-Oseretsky Test of Motor Proficiency, Second Edition), PA (Actical accelerometers), and perceived physical competence (The Pictorial Scale of Perceived Competence and Social Acceptance for Young Children, (PPC)). Measured height and weight were used to calculate preschoolers’ BMI. SEM was performed using Mplus software to test two models. Both models used the same combination of manifest variables to define six latent variables: three GMS-balance skills, locomotor skills, ball skills, PA, PPC, and fitness. Model 1 tested PA predicting the three GMS (balance, locomotor, and ball skills) and GMS predicting BMI. Model 2 tested the three GMS predicting PA and PA predicting BMI. Paths for PPC and fitness were tested in both models. All latent variables and BMI were regressed on preschooler ethnicity, age, and sex. Model fit was assessed using Chi-square and root mean square error of approximation (RMSEA), with p >.05 and p <.05 as indicators of close fit, respectively. Results: Model 1, testing PA predicting the three GMS (balance, locomotor, and ball skills) and GMS predicting BMI, demonstrated acceptable fit (Chi-square(556)=805, p>.05; RMSEA=.044). PA significantly, positively predicted locomotor and ball skills. Locomotor skills significantly, positively predicted fitness. Fitness did not significantly predict BMI; however, the relationship was in the expected direction (b=-0.176, p=.09). Model 2, testing three GMS predicting PA and PA predicting BMI, demonstrated acceptable fit (Chi-square(558)=824, p>.05; RMSEA=.045). Locomotor skills significantly, positively predicted PA. PA significantly, positively predicted fitness. Fitness did not significantly predict BMI; however, the relationship was in the expected direction (b=-0.132, p=.18). Implications/Future Directions: Both models showed significant pathways from locomotor skills to PA, and vice-versa, suggesting the need for additional research to examine the potential for reciprocity between PA and locomotor skills. Ball skills were not predictive of PA, likely due to preschoolers’ relatively low ball skill proficiency. Additional analyses will test individual latent variables in each model as mediators and will test direct paths from physical activity to fitness (model 1) and from physical activity to BMI (both models). Additional research is required to determine whether fitness or PA is a more appropriate predictor of health risk (BMI) in preschoolers. Longitudinal data are necessary to determine how the directionality of these relationships changes throughout child development, a next step for the Colorado LEAP study data set.

Description

Rights Access

Subject

preschool children
gross motor skills
physical activity
body mass index
structural equation modeling

Citation

Associated Publications

Collections