A Correlational Study between Anthropometric Measurements and Posture Indicators in an AI-Based Body Assessment System: Focusing on Gender Differences among Young Adults in Their 20s
This study examined the correlational relationships between anthropometric measurements and composite posture assessment scores using data from an AI-based fitness classification system (FitTrix) among 20s adults residing in Seoul and Gyeonggi Province. A total of 150 adults in their 20s (50 males, 100 females) participated in the study. Descriptive statistics, Pearson correlation analyses, independent samples t-tests, multiple linear regression analyses, and chi-square tests were performed. Results indicated that waist circumference (r = 0.51), shoulder width (r = 0.42), and pelvic width (r = 0.38) showed significant positive correlations with composite scores (all p < 0.001). Statistically significant differences were found between genders in all anthropometric measurements (p < 0.001), with height (t = 13.24, p < 0.001, Cohen’s d = 2.10) and shoulder width (t = 8.64, p < 0.001, Cohen’s d = 1.37) demonstrating very large effect sizes. Multiple linear regression analysis revealed that waist circumference, shoulder width, and pelvic width significantly explained variance in composite scores (R² = 0.57, F = 30.12, p < 0.001), with the female model showing higher explanatory power (R² = 0.61) than the male model (R² = 0.50). Posture classification indicators showed significant gender differences (χ² = 23.15 ~ 41.23, p < 0.001). These findings provide important scientific evidence for early posture management in 20s adults, developing individualized fitness prescriptions, and enhancing the reliability of AI-based body assessment systems.