Validation of a Weight History Questionnaire to Identify Adolescent Obesity

• Published in: Obesity surgery 2013-09, Vol.23 (9), p.1404-1412
• Main Authors: Jenkins, Todd M., Buncher, C. Ralph, Akers, Rachel, Daniels, Stephen R., Lawson, M. Louise, Khoury, Philip R., Wilson, Tawny P., Inge, Thomas H.
• Format: Article
• Language: English
• Subjects: Adolescent; Adolescent Behavior; Adult; Age Factors; Age of Onset; Body Height; Body Mass Index; Body Weight; Clinical Research; Female; Humans; Medicine; Medicine & Public Health; Mental Recall; Obesity; Ohio - epidemiology; Pediatric Obesity - epidemiology; Questionnaires; Recall; Reproducibility of Results; Retrospective Studies; Surgery; Surveys and Questionnaires - standards; Validity
• ISSN: 0960-8923; 1708-0428
• DOI: 10.1007/s11695-013-0901-7

Background Past body weight may be a more informative factor than current weight for risk of chronic disease development. Often, investigators must rely on subject recall to gauge past body weights. The Cincinnati Weight History Questionnaire (CWHQ) was developed to aid in the retrospective identification of adults who were obese during adolescence. Methods To assess validity, the CWHQ was administered to a subset of National Heart, Lung, and Blood Growth and Health Study (NGHS) participants, a group of young adult females for whom historical measured anthropometrics were available. One hundred ninety-eight NGHS participants were contacted, of whom 191 (97 %) responded (age 26–29). Participants were asked to recall height and weight from ages 13 and 18, which were compared to previously measured values. Multiple indices of validity (Bland–Altman plots, sensitivity, and specificity) were calculated. Results The CWHQ was moderately sensitive (range, 19–66 %), but highly specific (range, 89–100 %). Recalled height and weight values used to determine body mass index (BMI) underestimated BMI based on recorded height and weight at ages 13 and 18. Differences in calculated BMI based on recalled and measured height and weight were found to increase with BMI calculated using measured values. Conclusions The CWHQ proved to be a moderately sensitive, but highly specific instrument for detecting adolescent obesity in a cohort of young adult females. Epidemiologic research seeking to discriminate between adults with adult-onset vs. adolescent-onset obesity may find the CWHQ useful.