Tengfei Gu, Guanghui Li, Gen Li, Pan Gao, Yang Su, Xiaoyang Li, Ziniu Zhao
In this cross-sectional secondary analysis, a higher exploratory GGT-eGFR-derived index remained associated with higher baPWV after multivariable adjustment. Direct comparison with ln(GGT) and eGFR showed only modest differences in model fit, and prospective external validation is required before clinical application can be considered.
BACKGROUND: Elevated gamma-glutamyltransferase (GGT), a marker of oxidative stress and cardiometabolic burden, and reduced estimated glomerular filtration rate (eGFR) are each linked to arterial stiffness, but they are usually examined separately. This study evaluated whether a single index integrating both-the GGT-eGFR composite index-is associated with brachial-ankle pulse wave velocity (baPWV).
METHODS: This was a cross-sectional analysis of 893 adults from a Japanese health check-up program. The composite index was calculated as ln(GGT) × 100/eGFR. Its association with baPWV (continuous) and with elevated baPWV (≥1400 and ≥1600 cm/s) was assessed using multivariable linear and logistic regression, restricted cubic splines, joint GGT×eGFR stratification, subgroup analyses, and sensitivity analyses. Models containing the composite index were also compared with models containing ln(GGT), eGFR, or both components.
RESULTS: baPWV increased across index quartiles (1295 to 1531 cm/s). In the fully adjusted model, each 1-SD increment in the index was associated with a 20.46 cm/s higher baPWV (95% CI 5.64-35.28; P = 0.007); the highest versus lowest quartile difference was 74.77 cm/s (95% CI 33.96-115.57; P<0.001). The index was also associated with baPWV ≥1400 cm/s (OR 1.32, 95% CI 1.06-1.64) and ≥1600 cm/s (OR 1.36, 95% CI 1.04-1.79). Direct model comparisons showed a modest improvement in model fit for continuous baPWV after inclusion of the composite index, while categorical model performance was broadly comparable across models. Restricted cubic spline analysis showed an overall association with baPWV (P for overall = 0.011), with no statistically significant evidence of departure from linearity (P for non-linearity = 0.192). The association was consistent across sensitivity analyses.
CONCLUSION: In this cross-sectional secondary analysis, a higher exploratory GGT-eGFR-derived index remained associated with higher baPWV after multivariable adjustment. Direct comparison with ln(GGT) and eGFR showed only modest differences in model fit, and prospective external validation is required before clinical application can be considered.