Validity of overweight and obesity in a nation based on self-report versus measurement device data

Objective: To analyze overweight and obesity in a nation by self-report (SR) data and by data about the target person provided via other household members compared to measurement devices ( MD). The magnitude of hypothesized under-reporting by the SR data should be estimated with adjustment for age, gender, and education. Design: Two cross-sectional studies, nationally representative health examination surveys ( response rates: 61.4 and 73.2%, respectively). Setting: Adult general population of Germany aged 20 - 79 years. Subjects: Sample 1 included 6806 residents. Samples 2 (n = 98 673) and 3 ( n 34 960) included residents in the second survey. Measurements: In sample 1 MD data were collected, in sample 2 SR data were collected, and in sample 3 body weight and height information was provided from another household member living together with the target person. Results: MD data revealed higher proportions of overweight and obesity compared to SR. Among women with body mass index (BMI) 35.00 or higher, the odds ratio ( OR) was 3.9 (95% confidence interval, CI, 3.2 - 4.7), and among men 2.8 ( CI, 2.2 - 3.6) for MD versus SR. Data from other household members also revealed higher proportions of overweight and obesity than SR ( OR for BMI 35.00 or higher 2.1, CI, 1.7 - 2.5, for data from other household members versus SR in women and OR 1.3, CI, 1.1 - 1.5, in men). Conclusions: MD data should be used when providing proportions of overweight and obesity in a nation such as Germany.