Applying a correction procedure to the prevalence estimates of overweight and obesity in the German part of the HBSC study

BACKGROUND: Prevalence rates for overweight and obesity based on self-reported height and weight are underestimated, whereas the prevalence rate for underweight is slightly overestimated. Therefore a correction is needed. Aim of this study is to apply correction procedures to the prevalence rates developed on basis of (self-reported and measured) data from the representative German National Health Interview and Examination Survey for Children and Adolescents (KiGGS) to (self-reported) data from the German Health Behaviour in School Aged Children (HBSC) study to determine whether correction leads to higher prevalence estimates of overweight and obesity as well as lower prevalence rates for underweight. METHODS: BMI classifications based on self-reported and measured height and weight from a subsample of the KiGGS study (2,565 adolescents aged 11–15) were used to estimate two different correction formulas. The first and the second correction function are described. Furthermore, the both formulas were applied to the prevalence rates from the HBSC study (7,274 adolescents aged 11–15) which are based on self-reports collected via self-administered questionnaires. RESULTS: After applying the first correction function to self-reported data of the HBSC study, the prevalence rates of overweight and obesity increased from 5.5% to 7.8% (compared to 10.4% in the KiGGS study) and 2.7% to 3.8% (compared to 7.8% in the KiGGS study), respectively, whereas the corrected prevalence rates of underweight and severe underweight decreased from 8.0% to 6.7% (compared to 5.7% in the KiGGS study) and from 5.5% to 3.3% (compared to 2.4% in the KiGGS study), respectively. Application of the second correction function, which additionally considers body image, led to further slight corrections with an increase of the prevalence rates for overweight to 7.9% and for obese to 3.9%. CONCLUSION: Subjective BMI can be used to determine the prevalence of overweight and obesity among children and adolescents. Where there is evidence of bias, the prevalence estimates should be corrected using conditional probabilities that link measured and subjectively assessed BMI from a representative validation study. These corrections may be improved further by considering body image as an additional influential factor.