This study was a retrospective analysis of a prospective cohort done at Fuwai Hospital, Chinese Academy of Medical Sciences. The study was approved by the Institutional Review Board of Fuwai Hospital and followed the guidelines set forth in the Declaration of Helsinki. All participants provided written informed consent. This research was reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting criteria [13].

Between January 2017 and December 2018, a total of 13,506 individuals diagnosed with diabetes and angiography-confirmed CAD had regular post-discharge monitoring. Inclusion criteria were: 1) age ≥ 18 years; 2) participants had to have angiography results confirming the presence of CAD, with at least one coronary artery showing a stenosis of 50% or more; 3) participants had to have a confirmed diagnosis of diabetes. Diabetes was diagnosed if the patient had a previous diagnosis of diabetes, was on treatment to decrease glucose levels, or had a fasting blood glucose (FBG) level of 7.0 mmol/L or higher, HbA1c level of 6.5% or higher, or a 2-hour plasma glucose level of 11.1 mmol/L or higher during an oral glucose tolerance test [3]. Exclusion criteria were: 1) absence of important laboratory data; 2) severe liver or renal malfunction; 3) decompensated heart failure; 4) systemic inflammatory illness; 5) malignant tumor; 6) anemia; 7) hematological disorder; and 8) patient who lost to follow-up. Finally, 11921 individuals diagnosed with both diabetes and CAD were included in this study (Fig. S1).

The calculation of HGI was performed using the methods proposed by Hempe et al. [8]. Briefly, the baseline FBG and HbA1c data of all individuals were used to assess the linear relationship between FBG and HbA1c in the study group. (Fig. S2). Subsequently, a predicted HbA1c value was calculated by combining the baseline FBG with the subsample linear regression equation (predicted HbA1c = 0.013 × FBG [mmol/L] + 5.455). The baseline HGI was calculated by subtracting the expected HbA1c from the actual HbA1c. All participants would be classified based on their baseline HGI quintiles (quintile 1 [Q1]: ≤ −0.840; Q2: −0.840 to −0.322; Q3: −0.322 to 0.075; Q4: 0.075 to 0.790; Q5: ≥ 0.790).

Demographic and clinical information for all patients was obtained prospectively. The demographic data collected consisted of age, sex, body mass index (BMI), presence of concomitant disorders, smoking status, family history of CAD, history of prior myocardial infarction (MI), and history of percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG). The clinical data included the primary diagnosis at admission, the results of physical, radiological, and laboratory examinations, and the prescribed drug regimen upon discharge.

Upon admission, we obtained laboratory samples from each participant by drawing blood from the cubital vein after a minimum of 12 h of fasting. Our center’s clinical chemistry branch conducted all the exams. An enzymatic test was used to assess the amounts of triglycerides (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), FBG, and creatinine. The analysis was performed using an automated biochemical analyzer (Hitachi 7150, Tokyo, Japan). The calculation of low-density lipoprotein cholesterol (LDL-C) was performed using the Friedewald technique [14]. HbA1c was determined using high-performance liquid chromatography (Tosoh G8 HPLC Analyzer; Tosoh Bioscience, Tokyo, Japan). The high sensitivity C reactive protein (hsCRP) was analyzed using conventional biochemical methods at the central laboratory of Fuwai Hospital. The estimated glomerular filtration rate (eGFR) was determined using the Chinese-modified Modification of Diet in Renal Disease equation [15]. The modified biplane Simpson rule was used to evaluate the left ventricular ejection fraction (LVEF) at a state of rest [16].

The procedure of coronary angiography was carried out using standard techniques by experienced interventional cardiologists. Two experienced interventional cardiologists, working separately, examined the angiographic data obtained from the catheter laboratory at Fuwai Hospital. They documented the specific features of CAD, including unique types of narrowing in the coronary arteries, as well as the SYNergy between percutaneous coronary intervention with TAXus and cardiac surgery (SYNTAX) score.

Hypertension was characterized as having a systolic blood pressure (SBP) equal to or more than 140 mmHg, a diastolic blood pressure (DBP) equal to or greater than 90 mmHg, or the use of antihypertensive treatment [17]. Chronic kidney disease (CKD) was defined as the eGFR < 60 mL/min/1.73 m2 persisting for a duration of at least 3 months [18].

Participants were monitored at 6-month intervals until December 31, 2021, after their discharge. The data for endpoints were collected from medical records, clinical visits, and/or telephone interviews by experienced investigators who were unaware of the study design. The primary endpoint was the major adverse cardiac event (MACE), which included CV death and nonfatal myocardial infarction (MI). The secondary endpoints included all-cause death, CV death, nonfatal MI, and unplanned revascularization. Death was classified as CV-related unless a clear non-CV cause could be determined. Nonfatal MI was defined as the presence of positive cardiac troponins together with typical chest pain, characteristic electrocardiogram serial alterations, identification of an intracoronary thrombus by angiography or autopsy, or imaging data indicating fresh loss of viable myocardium or a new regional wall-motion abnormality [19]. Unplanned revascularization refers to the need for a treatment to restore blood flow to a lesion that did not satisfy the threshold for ischemia during the first operation and was not intended to be treated with a subsequent planned revascularization. Two separate physicians meticulously assessed all events.

Continuous variables were presented as mean ± standard deviation (SD) or median (interquartile range, IQR), and were compared using either the Student t-test or the Mann-Whitney U test. Categorical variables were represented as numerical values and percentages, and were compared using either the Fisher’s exact test or the chi-square test. The Kaplan-Meier curves were used to demonstrate the cumulative incidence of clinical endpoints across different groups, and the log-rank test was then used to compare these incidences. After adjusting for age and sex, we used restricted cubic spline (RCS) models to examine the presence of nonlinearity between continuous HGI and the likelihood of experiencing MACE during a 3-year period. The hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated using both univariable and multivariable Cox regression models. The multivariable Cox regression model used age, sex, BMI, duration of diabetes, acute coronary syndrome (ACS) presentation, family history of CAD, MI histories, previous revascularization, hypertension, previous stroke, PAD, current smoker, LVEF, serum creatinine, TG, LDL-C, HDL-C, TC, hsCRP, LM/three-vessel disease, chronic total occlusion (CTO) lesion, ostial lesion, type B2/C lesion, severe calcification, aspirin use, statins use, and insulin use as covariates. A two-tailed P value less than 0.05 was considered to be statistically significant. The analyses were performed using R version 4.0.3 software (R Foundation, Vienna, Austria).