Objective: To investigate whether, diet, physical activity, sedentary behavior or television (TV) viewing predicted body mass index (BMI) among 3-7-y-old children.
Design: A triethnic cohort of 3-4-y-old children was followed for 3 y from 1986 to 1989.
Measurements: BMI was assessed at the beginning and end of each measurement year. Heart rate monitoring and observation were used to assess physical activity. Diet (calories, % calories from fat and carbohydrate), sedentary behavior and TV viewing were assessed by direct observation in each year. A repeated measures regression analysis with year as a factor and BMI at the end of each year as dependent variables was run. Nonsignificant variables were removed in a stepwise backward deletion process and significant interactions graphed.
Results: The interactions between minutes of TV viewing per hour and study year and minutes of physical activity per hour and study year were significant (P<0.05). There were also significant main effects for TV viewing, physical activity and BMI from the beginning of the study. The model accounted for 65% of the variance in BMI across the three study years. Plotting the significant interactions demonstrated that physical activity was positively associated with BMI in year 1, and negatively associated in years 2 and 3 with a stronger negative relationship in year 3 than 2. TV viewing became positively associated with BMI during the third study year.
Conclusion: Physical activity and TV viewing were the only significant predictors (other than baseline BMI) of BMI among a triethnic cohort of 3-4-y-old children followed for 3 y with both physical activity (negatively associated) and TV viewing (positively associated) becoming stronger predictors as the children aged. It appears that 6 or 7 y is a critical age when TV viewing and physical activity may affect BMI. Therefore, focusing on reducing time spent watching television and increasing time spent in physical activity may be successful means of preventing obesity among this age group.