Central adiposity and sex are associated with the integrative architecture of physiological networks in healthy youth: a cross-sectional study in a cohort of Chinese college students

Figures

• 

  Figure 1. Multi-system physiological network architecture in Chinese healthy youth. (A) Full-sample network structure showing three distinct functional modules (Body composition, Autonomic function, ECG parameters) with inter-module connections. Only edges with absolute partial correlation > |0.01| are displayed. Edge width is proportional to connection strength, with red indicating positive partial correlations and blue indicating negative partial correlations. Dashed lines between modules represent inter-modular connectivity, with width scaled to connection strength. (B) Node strength stability with 95% CI from bootstrap analysis (CSC = 0.750), demonstrating high network reliability. (C) Node expected influence centrality ranking, showing WC’s superiority in centrality while BFP demonstrates higher connectivity. (D) Key intra-module connections showing the strongest partial correlations within each functional module. WC, Waist Circumference; BMI, Body Mass Index; BFP, Body Fat Percentage; ABSI, A Body Shape Index; ECG, Electrocardiographic; HR, Heart Rate; SDNN, Standard Deviation of NN intervals; rMSSD, Root Mean Square of Successive Differences; LF, Low Frequency power; HF, High…

• 

  Figure 2. Sex dimorphism in physiological network organization. (A) Male-specific network topology with enhanced ECG module connectivity. Only edges with absolute partial correlation > 0.01 are displayed. Edge visualization follows the same conventions as Figure 1A. (B) Female-specific network exhibiting distinct autonomic-ECG module integration. Network visualization conventions identical to Figure 1A. (C) Centrality differences (Male - Female) for key nodes, with HR and QTc showing higher centrality in males. (D) Butterfly plot comparing node expected influence between sexes, revealing sex-specific hub patterns. WC, Waist Circumference; BMI, Body Mass Index; BFP, Body Fat Percentage; ABSI, A Body Shape Index; ECG, Electrocardiographic; HR, Heart Rate; SDNN, Standard Deviation of NN intervals; rMSSD, Root Mean Square of Successive Differences; LF, Low Frequency power; HF, High Frequency power; QTd, QT dispersion; QTc, Corrected QT interval.

• 

  Figure 3. Obesity effects and sex-specific network alterations. (A) Forest plot of obesity effects on global network strength across three obesity indicators (BMI, WC, BFP) and sex strata. Negative values indicate network strength reduction in obesity. Error bars represent 95% CI from bootstrap analysis.For detailed sample sizes of participants with obesity in each subgroup, see Table 3. (B) Radar chart comparing multi-dimensional characteristics of obesity indicators. Dimensions: Centrality (expected influence, reflecting hub status), Stability (bootstrap stability coefficient, CSC = 0.750), Specificity (uniqueness relative to other obesity indicators), Connectivity (number of significant network connections). ABSI is included for network topology comparison but excluded from obesity grouping analysis due to lack of established thresholds. (C) Sex × obesity interaction effects on global network strength, suggesting potential gender-specific patterns (uncorrected p = 0.007 for BMI interaction, FDR-corrected q = 0.065). (D) Methodological sensitivity analysis confirming robustness of findings to QTc formula, covariate adjustment, and sparsity parameter choices. WC, Waist Circumference; BMI,…