Obesity and shorter QT intervals: sex- and adiposity distribution-specific patterns of ventricular repolarization in healthy Asian youth

Figures

• 

  Fig. 1. Sex-specific correlation network of obesity, HRV, HR and QTcF. a, Male participants (n = 777). b, Female participants (n = 2,379). Chord diagrams visualize sex-stratified Spearman’s rank correlation coefficients (ρ). Only correlations with FDR p < 0.05 are displayed. Chord width and color intensity represent the absolute strength of the correlation (thicker/darker indicates stronger). Red chords indicate positive correlations; blue chords indicate negative correlations. Abbreviations: BMI, body mass index; BFP, body fat percentage; WC, waist circumference; WHR, waist-to-hip ratio; HR, heart rate; QTcF, Fridericia-corrected QT interval; SDNN, standard deviation of all normal-to-normal intervals; rMSSD, root mean square of successive differences; LF/HF, ratio of low-frequency to high-frequency power; FDR, false discovery rate

• 

  Fig. 2. Sex-stratified comparison of physiological phenotypes between obesity and non-obesity groups. a, QTcF. b, SDNN. c, rMSSD. d, LF/HF. This composite figure illustrates differences between individuals with and without obesity across four physiological indicators, according to four distinct criteria (2024 Chinese Guidelines for the Clinical Management of Obesity): WC, BMI, BFP, and WHR. Within each panel, bars are grouped by obesity criterion (WC, BMI, BFP, WHR, from left to right). The sample size for the between-group comparison within each sex and criterion is indicated beneath the x-axis label in the format “[n with obesity] vs [n without obesity]”. Each group contains four bars representing: men with obesity (dark blue), men without obesity (light blue), women with obesity (dark red), and women without obesity (light red). Bar height represents the group median, and error bars indicate the 95% CI. The Cliff’s delta (δ) value annotated above each bar indicates the result of the within-sex, within-criterion comparison (Mann-Whitney U test, * FDR p < 0.05). Abbreviations: QTcF, Fridericia-corrected QT interval; SDNN, standard deviation of all normal-to-normal intervals; rMSSD, root…

• 

  Fig. 3. Sex-specific patterns of obesity-related electrophysiological alterations. a, Forest plots of Cliff’s delta (δ) effect sizes : comparisons between men with obesity and women with obesity across physiological indicators, stratified by the same obesity definition. δ < 0 indicates the value is lower in men than in women. Error bars represent 95% CI. b, Forest plots quantifying QTcF differences based on clinical obesity thresholds. Solid circles represent unadjusted mean differences (obesity minus non-obesity). Hollow circles represent adjusted regression coefficients (β) from Huber-robust regression models (ε = 1.35), accounting for HR, age, physical activity level, and measurement time. Error bars represent 95% CI. In both panels, asterisks denote statistical significance after FDR adjustment (* FDR p < 0.05). Obesity was defined according to the 2024 Chinese Guidelines for the Clinical Management of Obesity. Sample sizes for each obesity type are provided in Supplementary Table S3.Abbreviations: BMI, body mass index; BFP, body fat percentage; WC, waist circumference; WHR, waist-to-hip ratio; QTcF, Fridericia-corrected QT interval; SDNN, standard deviation of all normal-to-normal…

• 

  Fig. 4. Mediation analysis exploring the role of autonomic function in the obesity-QTcF association. Path diagrams illustrate the standardized direct and indirect effects of obesity metrics (BMI and BFP) on QTcF through HRV mediators, with separate paths visualized for male (blue, n = 777) and female (red, n = 2,379) participants. Standardized path coefficients (β) with 95% CI (in parentheses) were derived from bootstrap analysis (1,000 repetitions). Solid arrows indicate statistically significant paths (FDR p < 0.05); dashed arrows indicate non-significant mediation paths. Path labels: a (obesity metric → HRV mediator), b (HRV mediator → QTcF), c (direct effect: obesity metric → QTcF). Pm represents the mediation proportion (indirect effect/total effect × 100%). All models were adjusted for HR, age, physical activity level, and measurement time. Parallel multiple mediation analysis was performed using bootstrap indirect effect testing. Full results are in Supplementary Table S4. Abbreviations: BMI, body mass index; BFP, body fat percentage; QTcF, Fridericia-corrected QT interval; HRV, heart rate variability; SDNN, standard deviation of all normal-to-normal intervals; rMSSD, root mean…