Self-reported versus measured anthropometry yields clinically relevant body mass index underestimation

Abstract Objective This study aims to quantify bias in self-reported height and weight, assess its impact on body mass index (BMI)-based obesity classification, and identify clinical and demographic predictors of misreporting in a specialist outpatient population. Design Cross-sectional observational study comparing self-reported and measured anthropometric data, with multivariable regression analyses to investigate factors associated with discrepancies in height, weight, and BMI. Methods A total of 408 adult outpatients were enrolled in a tertiary-level outpatient endocrinology and obesity clinic in Italy. Self-reported height and weight were recorded immediately before standardized measurements. Body mass index was calculated from both data sources. Differences were expressed as Δ height, Δ weight, and Δ BMI. Agreement in WHO BMI categories was evaluated, and sensitivity/specificity for detecting obesity (BMI ≥30 kg/m²) was calculated using measured BMI as reference. Multivariable linear models included sex, age, menopausal status (in women), measured BMI, and waist circumference. Results Mean (95% confidence intervals) misreporting was +3.13 cm (+2.90 to +3.36) for height, −0.97 kg (−1.29 to −0.65) for weight, and −1.63 kg/m² (−1.80 to −1.47) for BMI. WHO BMI-category misclassification occurred in 20.0% of participants, primarily due to under-classification (19.4%). Sensitivity and specificity for detecting obesity based on self-report were 0.72 and 1.00, respectively. Age predicted height over-reporting; age and measured BMI predicted both weight under-reporting and BMI underestimation. Conclusions Self-reported anthropometric data systematically underestimate BMI and result in substantial misclassification of obesity. Given the clinical and policy implications of BMI-based criteria for obesity treatment access, objective measurement should be prioritized in all relevant settings.