Original article
Validity of self-reported height and weight and predictors of bias in adolescents

https://doi.org/10.1016/j.jadohealth.2004.07.014Get rights and content

Abstract

Purpose

The purpose of this study was to assess the validity of self-reported height and weight, the accuracy of screening for overweight and obesity using these data, and the predictors of bias in self-reported height and weight in adolescents.

Method

The Health Behavior in School-Aged Children (HBSC) survey was used to assess self-reported and measured height and weight in a community sample of 418 students in year 11 from 21 secondary schools in Wales. Participants also provided self-report data on dieting and body perception.

Results

Self-reported and measured height and weight were highly correlated but a bias of underreporting of body weight by an average of .52 kg contributed to underestimation of the prevalence of overweight and obesity. Based on self-report data, 13.9% of the sample was identified as overweight and 2.8% was identified as obese, but measured data showed rates of 18.7% and 4.4%, respectively. Overweight and obese participants showed greater bias and variability in self-reported weight than normal/underweight participants. Body mass index (BMI) and body dissatisfaction predicted bias in self-reported weight.

Conclusion

Self-report bias had significant consequences for the accuracy of a screen for overweight and obesity. Actual and perceived body size each contribute to underreporting body weight. Self-reports will remain an important health surveillance tool but should not be relied on exclusively to detect weight problems. Accuracy checks of self-report data may improve surveys of weight problems in young people.

Section snippets

Participants

Data were used from the 1998 Welsh sample of the Health Behavior School-Aged Children (HBSC) study [14], [15]. Before data collection, the study procedures were approved by the Health Promotion Division of the Welsh Assembly Government and the World Health Organization for formal endorsement of its conformity with the Organization’s standards for ethical research practices.

A cluster sample of 51 schools represented a national distribution of schools by size, geographic location (stratified by

Results

Descriptive statistics on measured and self-reported weight, height, and BMI are shown in Table 1. Self-reported weight was less than measured weight in girls, t(159) = 4.45, p < .001, but not in boys. There were no differences found between self-reported height and measured height in boys or in girls. BMIs based on self-report data were lower than BMIs based on measured data in both boys (t(140) = 2.18, p = .03) and girls (t(149) = 2.25, p = .03). The degree of self-report bias in weight,

Discussion

Self-reported height and weight can be used to calculate BMI and provide a categorical measure of overweight and obesity using cut-off points in BMI. The goal of this study was to assess the validity of these categoric data as a screen for overweight and obesity and to explore predictors of self-report bias. Despite highly correlated self-reports and body measurements, the screen missed nearly half of all cases of overweight and obesity. The high specificity and low sensitivity of the screen

References (18)

There are more references available in the full text version of this article.

Cited by (0)

View full text