The non-linear association between the estimated small dense LDL cholesterol-glucose index and incident Sarcopenia: a prospective cohort study

Sarcopenia is an age-related disorder marked by progressive declines in skeletal muscle mass and function and is closely linked to metabolic abnormalities. The estimated small dense low-density lipoprotein cholesterol–glucose (sdLG) index integrates lipid and glycemic information, but its prospective association with incident sarcopenia remains unclear. We prospectively followed a cohort of middle-aged and older Chinese adults without sarcopenia at baseline. Small dense low-density lipoprotein cholesterol (sdLDL-C) levels was estimated from routinely measured lipid parameters using a previously established equation and was combined with fasting glucose to calculate the sdLG index. We used Cox proportional hazards regression to calculate hazard ratios (HRs) and 95% confidence intervals (CIs) for incident sarcopenia. Potential nonlinearity was examined using restricted cubic splines, followed by an exploratory two-piecewise Cox analysis. Subgroup and sensitivity analyses were evaluated to detect the robustness of the findings. Among 6,065 participants, 611 developed incident sarcopenias during follow-up. In the multivariable-adjusted Cox model, each 1-unit increase in sdLG was associated with a lower hazard of incident sarcopenia (HR, 0.672; 95% CI, 0.550–0.821; p < 0.001). Restricted cubic spline analysis indicated a nonlinear association (P for nonlinearity = 0.040). Exploratory two-piecewise regression analysis identified an inflection point at 7.896. Below this value, each 1-unit increase in sdLG was associated with a lower hazard of incident sarcopenia (HR, 0.564; 95% CI, 0.443–0.718; p < 0.001). At or above 7.896, the point estimate changed direction but was imprecise and not statistically significant (HR, 1.364; 95% CI, 0.773–2.408; p = 0.285). The overall pattern was generally consistent across subgroup and sensitivity analyses, although the association was attenuated when BMI category was used to account for adiposity Among middle-aged and older Chinese adults, baseline sdLG was nonlinearly associated with incident sarcopenia, with an inverse association primarily observed below the data-derived inflection point. The association at higher sdLG values was imprecise and not statistically significant, and adjustment for BMI attenuated the overall association. These findings should therefore be interpreted cautiously and do not establish a protective effect of higher sdLG or a clinically meaningful cutoff. Independent prospective studies incorporating direct body-composition measurements are needed to confirm the observed exposure–response pattern.