Predictive Value of Serum Gamma-glutamyltranspeptidase for Future Cardiometabolic Dysregulation in Adolescents- a 10-year longitudinal study

Abstract

Serum gamma-glutamyltransferase (γ-GT) is implicated in the pathogenesis of atherosclerosis and metabolic syndrome (MetS) in adults. The relationships between γ-GT and cardiometabolic dysregulation remains unclear in adolescents. We enrolled 7,072 Taiwanese adolescents and followed them for a median of 6.8 years. The optimal cut-off values (CoVs) of baseline γ-GT to predict future MetS, hypertension (HTN), and type 2 diabetes (T2DM) were determined by receiving operating characteristic (ROC) curve. Using these CoVs, the participants were divided into normal- and high-level groups. Cox proportional hazard analysis was used to calculate hazard ratios (HRs) for the subjects with a high level of γ-GT for the risk of future cardiometabolic dysregulation. Serum γ-GT was significantly higher in the subjects with MetS than in those without MetS at baseline (p < 0.001). The optimal CoVs of γ-GT were 12 U/L for boys and 11 U/L for girls. In multivariate Cox regression analysis, a higher serum γ-GT level increased the risk of future MetS (HRs 1.98 and 2.85 for boys and girls, respectively, both p < 0.001), but not new onset HTN and T2DM. In conclusion, serum γ-GT levels not only demonstrated an excellent correlation with the presence of MetS and also in predicting future MetS in adolescents.

Figure 1

Enrollment flow diagram. A total of 11,370 participants aged from 10 to 15 years who underwent regular health examinations from 1999 to 2008 at MJ Health Screening Centers were enrolled. Among them, the subjects with only one visit (n = 3,545), missing data of MetS components or γ-GT (n = 512), and a history of alcohol consumption, HTN, type 1 diabetes and those taking medications known to affect MetS components or serum γ-GT levels (n = 241) were excluded. The remaining 7,072 subjects were enrolled as the study cohort. In stage 1, the optimal CoVs of baseline γ-GT to differentiate the subjects with and without MetS were identified by ROC curve. Using these CoVs, the aim of second stage was to validate its predictive role on future MetS, HTN and T2DM.

Figure 2

Receiver operating characteristic curves for serum γ-GT in both genders. Receiver operating characteristic curves and optimal cut-off values for serum γ-GT for differentiating between MetS and non-MetS in (a) male and (b) female adolescents.

Figure 3

Kaplan-Meier plot of developing future MetS, HTN and T2DM by different γ-GT levels. Kaplan-Meier curves estimate with log rank test was applied for the event-free probability between the subjects with normal γ-GT levels (≤12 U/L) and high γ-GT levels (>12 U/L).