Definition of new cut‐offs of BMI and waist circumference based on body composition and insulin resistance: differences between children, adolescents and adults

OBJECTIVE: This study aims to determine associations between anthropometric traits, regional fat depots and insulin resistance in children, adolescents and adults to define new cut‐offs of body mass index (BMI) or waist circumference (WC). DESIGN: Cross‐sectional data were assessed in 433 children, adolescents and adults (aged: 6–60 years, BMI: 23.6 [21.0–27.7] kg m(−2)). Total adipose tissue (TAT), regional subcutaneous adipose tissue (SAT(total), SAT(trunk)) and visceral adipose tissue (VAT) were determined by whole‐body magnetic resonance imaging, fat mass by air‐displacement plethysmography. Insulin resistance was evaluated by homeostasis model assessment of insulin resistance (HOMA‐IR). Bivariate as well as partial correlations and regression analyses were used. Cut‐off values of BMI and WC related to regional fat depots and HOMA‐IR were analysed by receiver operating characteristics curve. RESULTS: In adults, TAT, SAT(total) and SAT(trunk) increased linearly with increasing BMI and WC, whereas they followed a cubic function in children and adolescents with a steep increase at BMI and WC ≥1 standard deviation score and VAT at WC ≥2 standard deviation score. Sex differences were apparent in adults with women having higher masses of TAT and SAT and men having higher VAT. Using established BMI or WC cut‐offs, correspondent masses of TAT, SAT(total), SAT(trunk) and VAT increased from childhood to adulthood. In all age groups, there were positive associations between BMI, WC, SAT(trunk), VAT and HOMA‐IR. When compared with normative cut‐offs of BMI or WC, HOMA‐IR‐derived cut‐offs of regional fat depots were lower in all age groups. CONCLUSIONS: Associations between BMI, WC and regional fat depots varied between children, adolescents, young and older adults. When compared with BMI‐derived and WC‐derived values, an insulin resistance‐derived cut‐off corresponded to lower masses of regional fat depots. Thus, established BMI and WC cut‐offs are not appropriate to assess metabolic disturbances associated with obesity; therefore, new cut‐offs of BMI and WC are needed for clinical practice.