The present work firstly assessed eGDR’s association with ISR following PCI in CAD. The results revealed eGDR was independently negatively associated with the increased risk of ISR following PCI in NSTE-ACS; furthermore, eGDR improved the predictive ability of routine cardiovascular risk factors for ISR; moreover, the predictive value of eGDR for ISR was mainly reflected in patients without T2DM.
IR is the most important pathogenetic mechanism of T2DM and metabolic syndrome, with the main features including the following two aspects: decreased ability of insulin to induce glucose uptake and use; body compensation by secreting more insulin for inducing hyperinsulinemia to stabilize blood sugar. Insulin resistance causes endothelial dysfunction, oxidative stress, and the activation of inflammatory responses, ultimately leading to the formation of atherosclerotic plaques . Currently, assessment techniques for insulin resistance mostly encompass two categories: direct assessment methods and simple surrogate assessment indicators. Applying the hyperinsulinemic-euglycemic clamp and the insulin suppression test are both direct assessment methods for insulin resistance. By applying the hyperinsulinemic-euglycemic clamp, researchers confirmed that IR is tightly associated with coronary atherosclerotic heart disease, with a predictive role independent of other risk factors [30–32]. For simple surrogate assessment indicators of IR, many clinical studies have used homoeostasis model assessment of insulin resistance (HOMA-IR) as an assessment method to explore the relationship between IR and cardiovascular disease (CVD), with consistent results. Indeed, IR is highly associated with atherosclerosis  and predicts CVD onset and poor prognosis in non-diabetic individuals [34–36]. However, in clinical practice, fasting insulin levels are not routinely measured even in diabetics, let alone in individuals without diabetes. In addition, insulin measurement methods do not yield consistent data across laboratories, especially in case of low insulin levels. Therefore, researchers have proposed a variety of simpler alternative assessment indicators of insulin resistance, including triglyceride-glucose (TyG) index, triglyceride/high-density lipoprotein cholesterol (TG/HDL-C), visceral adiposity index (VAI) and lipid accumulation product (LAP), which are highly correlated with the incidence and prognosis of ASCVD [37–40]. eGDR is also a simple surrogate measure of this type of IR.
It has long been confirmed that diabetes could predict the occurrence of ISR [41, 42], and a study suggested that diabetes is the most effective predictor of ISR . In addition, a meta-analysis showed ISR incidence is markedly elevated in diabetic patients in comparison with non-diabetics . Therefore, diabetes can almost be considered the clearest risk factor for ISR. A long time ago, it was shown that IR is a common feature of CVD patients undergoing stent surgery, and an important marker of restenosis after PCI, with a deterioration process related to endothelial dysfunction, nitric oxide production disorders and activity defects . In recent years, studies applying HOMA-IR have confirmed that insulin resistance is highly correlated with ISR occurrence after PCI, representing an independent predictor of ISR [12, 14]. In addition, a study using TyG as an evaluation index of IR found that TyG is independently and positively correlated with ISR risk following DES implantation in ACS patients .
As for eGDR, its associations with stroke incidence and mortality in T2DM patients have been demonstrated . In addition, eGDR was also shown to be closely related to elevated risk of all-cause mortality after CABG in T2DM patients, independent of other cardiovascular and metabolic risk factors . The above findings suggest that eGDR has great potential in predicting ASCVD prognosis and ISR events after PCI. This study clarified the predictive potential of eGDR for ISR occurrence post-PCI in NSTE-ACS cases, which is consistent with previous findings. The present work not only confirmed IR could predict ISR occurrence upon PCI in NSTE-ACS cases, but also revealed a new and effective indicator applicable for ISR prediction. In data analysis, we attempted to include diabetes and related variates in the multivariate analysis, but after final adjustment, eGDR lost statistical significance in ISR prediction. Therefore, a subgroup analysis was carried out based on the diabetes status. The results revealed eGDR only had a predictive value in ISR for the non-diabetic subgroup. Furthermore, incremental effect analysis in the diabetes and non-diabetes groups was also consistent with the above subgroup analysis. This could explain the lack of significance for eGDR in models incorporating diabetes and associated variates. This finding suggests that eGDR might be important in routine assessment of CVD cases, which requires further investigation in large prospective trials. Finally, whether eGDR can really be used clinically as a powerful predictor of ISR after PCI needs to be assessed via comparison with other IR evaluation indicators.
There were still some limitations of the present study that need to be further confirmed by more rationally designed studies. First, this was a single-center observational study of Chinese individuals. Therefore, multi-center trials or even randomized controlled studies with larger samples and greater racial diversity are warranted to further clarify the current results. Additionally, UA cases in this study cohort constituted the greatest part of all cases, and the current findings might not reflect the prognostic value of eGDR for ISR in NSTEMI. Furthermore, regarding repeat coronary angiography after discharge, ISR detection was not based on intracoronary imaging, and its accuracy was insufficient. Moreover, this work did not clarify the specific time when ISR occurred within 48 months after discharge and lacked short-term and long-term ISR analyses.