Age Moderates The Relationship Between Physical Activity And Bmi Percentile In Young Children

Purpose: Identify age categories that maximize the correlation between MVPA and BMI percentile (BMI%) in child aged 2-5.

Methods: MVPA was estimated from accelerometer data and summarized into minutes using cut-points suggest by Pate et al. The 228 children wore the monitors for 7 days. BMI% was computed using standard CDC criteria. A SAS macro was used to estimate and summarize the partial correlations (controlling for Race, Sex, Income) between BMI% and MVPA across all possible 3 and 4 category age groupings. The best groupings (average correlation or largest correlations) were examined. Groups defined by the age cut-offs needed at least 25 children to be considered.

Results: The average child was 3.5 years, had a BMI% of 60.8, and participated in 59.4 minutes of MVPA per day. Controlling for age, sex, race, and income the correlation between BMI% and MVPA was 0.09 (p=0.19). The results suggested to create three groups split at 2.9 and 4.4 yrs. The correlations between BMI% and MVPA were -0.03 (<2.9), 0.23(2.9-4.4), and -0.34 (4.4+). In addition, the age by BMi% interaction was significant (p = 0.012). Mean physical activity for the three age groups across 3 BMi groups are plotted in figure 1.

Conclusion: Age moderates the relationship between MVPA and BMI in young children. The upper cut off 4.4 years matches very closely with the age at which the BMI curve for the 60th% starts to plateau. Preliminary results show that this upper age cut may be different for girls and boys, and may match the plateau point in their BMI growth curves.