BMI screening in schools: Does an active consent process yield a representative sample of students?

Background: School-based BMI assessment tracks childhood obesity trends and monitors intervention outcomes at a population level, and identifies at-risk students at an individual level. Although school-based BMI assessments have been recommended by the Institute of Medicine, the practice is controversial and some parents are uncomfortable with the school's involvement in this process. To address parental concerns, some schools utilize active parental consent procedures, which may result in a biased sample and inaccurate overweight and obesity prevalence estimates.

Purpose: to determine if the active consent sample obtained in West Virginia schools provided BMI data that were representative of the state.

Significance: determining whether active parental consent provides an unbiased BMI sample addresses the utility of the procedure for surveillance.

Methodology: we measured BMI in a cluster sample of students and compared those data with the BMI results obtained by active consent.

Findings/Results: we compared the proportion of students who were underweight, healthy weight, overweight, and obese in the cluster sample of 1640 students from 34 schools in 16 counties with the active consent sample of 7046 students in 332 schools in 55 counties obtained through the CARDIAC project, which is charged with conducting BMI assessment in WV schools. Results of a 2 x 4 chi-square and post-hoc analyses at the state level indicated a significant difference (p=.008) between the samples, which was attributable to differences in underweight students only (p=.002).

Conclusions/Recommendations: the active consent process provides a representative estimate of the prevalence of overweight and obese students in WV.