The role of triglycerides-glucose index to predict 10-year first and recurrent cardiovascular disease events: a sex-based analysis from ATTICA and GREECS prospective studies

Abstract

Abstract Background/Introduction Triglycerides-glucose index (TyG) has been used in apparently healthy individuals to define insulin resistance and liver steatosis. Additionally, findings from very recent studies challenge the use of this index as a predictor of cardiovascular disease (CVD) onset as well as a marker of atherosclerosis in patients with established CVD. Purpose To evaluate the association between TyG and 10-year first or recurrent fatal/non fatal CVD event. Methods The samples from two prospective epidemiological studies implemented in Greece were used. In particular, in ATTICA study, in 2001–02, 1,514 men and 1,528 women (>18 years) free of CVD, at baseline, living in greater Athens area, Greece, were enrolled and 10-year follow up was performed (2011–12) in 2,020 participants (n=317 cases). In GREECS study, in 2003–04 almost all consecutive 2,172 acute coronary syndrome (ACS) patients of 6 Greek hospitals were enrolled. In 2013–14, 10-year follow-up was performed in 1,918 participants. TyG was assessed at baseline using a standard formula. Results In ATTICA, ranking from 1st (i.e. <8.0) to 3rd TyG tertile (i.e. >8.6), 10-year first CVD incidence was 5.6%, 14.2% and 24.1% (p<0.001); the respective man-to-woman incidence ratio was 1.86, 1.17 and 1.19. In GREECS, ranking from 1st (i.e. <8.7) to 3rd TyG tertile (i.e. >9.3), 10-year recurrent CVD incidence was 35.3%, 43.2% and 35.9% (p=0.11); the respective man-to-woman incidence ratio was 1.08, 0.99 and 1.23. Multi-adjusted Cox regression analysis in ATTICA study revealed that participants assigned in the 3rd TyG tertile had about 77% higher CVD risk compared with their 1st tertile counterparts [Hazard Ratio (HR)=1.77, 95% Confidence Interval (95% CI) (1.06, 2.96), p=0.02]; sex-based stratified analysis revealed that this association remained significant only in women [HR=2.29 95% CI (1.20, 4.38), p=0.01] while in case of men this association was borderline significant [HR=1.70 95% CI (0.95, 3.35), p=0.10]. The total correct classification rate was around 83–85% in all models and similar with models adjusted separately for triglycerides or glucose levels. Using the area under the Receiver Operation Characteristic ROC) curve (AUC) analysis TyG had the best discriminative ability in both sexes with a small advantage in favor of women and minor discrepancies with the commonly used –triglycerides and glucose– biomarkers (Women: AUCTyG=0.694, AUCtriglycerides=0.678, AUCglucose=0.601 / Men: AUCTyG=0.662, AUCtriglycerides=0.634, AUCglucose=0.623). In case of GREECS study, no significant trends were observed. Conclusions The findings here suggest a predictive role of TyG against long-term CVD onset which comes in line with recent works; yet its added value against conventional markers such as glucose and triglycerides was not confirmed. Additionally, no significant prognostic effect of TyG against CVD recurrence was observed challenging its clinical use in secondary prevention spectrum. Funding Acknowledgement Type of funding source: Other. Main funding source(s): This work was supported by a research grant from Hellenic Atherosclerosis Society. The ATTICA study is supported by research grants from the Hellenic Cardiology Society [HCS2002] and the Hellenic Atherosclerosis Society [HAS2003].