Background: Body mass index (BMI) is commonly used to assess nutritional status; however, it cannot distinguish between fat and lean tissue. In young adults, this limitation may mask excess adiposity and distort diet–adiposity associations. Bioelectrical impedance analysis (BIA) provides more detailed measures, including percent of body fat (PBF), skeletal muscle mass (SMM), and the visceral fat level. Objectives: To examine how combining BMI with BIA-based classifications of adiposity influences the assessment of diet–body composition associations in young adults. Methods: This cross-sectional study of 285 young adults (median age 18 years, IQR: 18–20) used InBody BIA to classify participants by BMI and PBF. Dietary habits were assessed via food frequency questionnaire covering eight food groups. Group comparisons used Mann–Whitney U tests with Cohen’s d effect sizes; correlations used Spearman’s rank correlation. Results: Thirty-five participants (12.3%) were BMI-Normal but PBF-High (normal BMI with elevated body fat), a phenotype missed by BMI screening; overall BMI-PBF agreement was 75.4%. Physical activity (IPAQ) correlated significantly with body composition markers, PBF (rho = −0.177, p = 0.003) and SMM (rho = +0.186, p = 0.002), but not with BMI (rho = +0.060, p = 0.310). BMI showed an inverse association with self-reported sweets consumption (rho = −0.138, p = 0.020), likely reflecting a reporting bias rather than true intake, as this pattern disappeared when examining actual adiposity (PBF: rho = +0.032, p = 0.591). Conclusions: Combining BIA with BMI may improve the detection of elevated body fat (12.3% prevalence of normal BMI with elevated body fat); BMI-based screening may not identify all individuals with elevated body fat. Physical activity associations support the complementary value of BIA alongside BMI. Apparent diet–BMI associations may be confounded by adiposity misclassification and reporting bias.
Nagy et al. (2026) studied this question.