Do self- or parent-reported dietary, physical activity, and sedentary behaviors predict worsening obesity in children?

Abstract

Objective: To determine whether information gathered during routine healthcare visits regarding obesity related risk factors and risk behaviors predicts increases in BMI z-score over time among overweight and obese children.

Study design: Medical records from 168 overweight and 441 obese patients seen for repeated visits between September 2003 and April 2006 were examined for reported dietary, physical activity, and sedentary behaviors, family history of obesity and diabetes mellitus, documented Acanthosis nigricans, and BMI values. Random-effects regression analysis was done to determine whether demographic, familial, or behavioral data predicted changes in BMI z-score over time.

Results: The presence of A nigricans and a family history of obesity were associated with an increase in BMI z-score (beta=0.56, SE=0.09, P<.001 and beta=0.31, SE=0.13, P=.021). These risk factors explained 8% and 7% of the variation in BMI z-score respectively. Self- or parent-reported dietary and physical activity behaviors did not predict change in BMI z-score.

Conclusions: Our findings suggest that the risk factors and self- or parent-reported risk behaviors routinely assessed by pediatric clinicians have limited ability to predict future growth trends, demonstrating the difficulty in determining which patients have the greatest risk of progression of obesity.

Figure. Proportion of Overweight and Obese Patients with a Decrease, no Change, or Increase in BMI z-Score at 1, 2 and 3 Years and the Proportion with a Change In Diagnostic Category at the End of Follow-Up

The percent of overweight and obese patients represented by the grey bars was determined using absolute change in BMI z-score of 0.1(i.e. Decreased; BMI z-score <−0.1 {n=14, 10, and 1 overweight patients and n=50, 29 and 16 obese patients at 1, 2, and 3 years respectively}, Unchanged; −0.1 ≤ BMI z-score ≤ 0.1, Increased; BMI z-score > 0.1 {n=22, 9, and 2 overweight patients and n=38, 19, and 10 obese patients at 1, 2, and 3 years}.