Muscle Strength as a Key Independent Predictor of Arterial Stiffness and Metabolic Syndrome in Aging Mexicans: Unveiling the Sarcopenic Obesity Paradox

Figures

• 

  Figure 1. (A) Prevalence of the four phenotypes (n = 2087). (B) Heatmap of cardiometabolic risk by phenotype (red = higher risk; protective variables were inverted). Because multiple biomarkers and functional indicators are compared across the four phenotypes, p-values were corrected for multiple testing using the Benjamini–Hochberg false discovery rate (FDR); after correction, BMI, waist circumference, pulse pressure, handgrip strength, gait speed and balance remained significant (q < 0.001), with the total-cholesterol/HDL-C ratio and HDL-C at q < 0.05, whereas C-reactive protein (q = 0.79) and 25(OH)D (q = 0.24) did not; so, the heatmap is intended primarily for descriptive visualization. Abbreviations: PP, pulse pressure; CRP, C-reactive protein; HDL-C, high-density lipoprotein cholesterol; HbA1c, glycated hemoglobin; 25(OH)D, 25-hydroxyvitamin D; WHtR, waist-to-height ratio. Statistics: Kruskal–Wallis test (all variables non-normally distributed, Shapiro–Wilk p < 0.001).

• 

  Figure 2. Obese–sarcopenic paradox by phenotype. (A) Pulse pressure. (B) Gait speed. (C) Handgrip strength. (D) Serum 25(OH)D. Bars represent means; error bars indicate SEM; dashed lines denote clinical thresholds. *** p < 0.001; ns = not significant (Kruskal–Wallis).

• 

  Figure 3. Strength gradient across handgrip strength tertiles. (A) Pulse pressure. (B) Metabolic syndrome prevalence. (C) Vitamin D deficiency prevalence. Of note, BMI remained similar across all tertiles (non-significant), indicating that the observed gradient is attributable to muscle function rather than adiposity. (D) Restricted cubic spline of continuous handgrip strength versus pulse pressure, adjusted for age and sex (natural cubic spline, 3 degrees of freedom; shaded band = 95% CI). The overall association was significant (p = 0.032), with no evidence of departure from linearity (test for non-linearity p = 0.60), supporting an approximately linear dose–response relationship.

• 

  Figure 4. (A) Sex-adjusted OLS model corrected for multicollinearity (n = 1757; R2 = 0.207). BMI was removed from the set of continuous predictors due to high collinearity with WHtR (r = 0.85; original VIF > 100); all VIF values in the final specification were < 5. HC3 robust standard errors were used to account for heteroscedasticity (Breusch–Pagan p < 0.001). Significant predictors: age, metabolic syndrome, WHtR, and handgrip strength. (B) Sex-stratified model—men only, corrected for multicollinearity (n = 706; R2 = 0.162). After removing BMI, metabolic syndrome (β = +6.06 mmHg; p < 0.001) and age (β = +0.59 per year; p < 0.001) were the only significant predictors in men. Handgrip strength was not significant (β = −0.05; p = 0.488). (C) Sex-stratified model—women only, corrected for multicollinearity (n = 1051; R2 = 0.248). After removing BMI, handgrip strength (β = −0.13 per kg; p = 0.010) and gait speed (β = −6.38 m/s; p = 0.048) emerged as independent predictors exclusively in women; WHtR reached marginal significance (β = −13.86; p = 0.057). Metabolic syndrome (β = +8.06; p < 0.001) and age (β = +0.64 per year; p < 0.001) remained the dominant predictors.

• 

  Figure 5. Age-stratified trajectories. (A) Obese–sarcopenic phenotype and vitamin D deficiency. (B) Slow gait and metabolic syndrome. (C) Mean pulse pressure (60 mmHg threshold indicated). (D) Variance inflation factor (VIF) collinearity table: green, VIF < 5; yellow, VIF 5–10; red, VIF > 10.