Metabolic score for insulin resistance predicts major adverse cardiovascular event in premature coronary artery disease

Study design and patients

This study adhered to the ethical standards outlined in the Declaration of Helsinki and obtained approval from the Ethics Review Committee of Qilu Hospital, Shandong University. The follow-up of our retrospective study was conducted via telephone, and the ethics committee granted permission for the collection of verbal consent.

From late September 2022, we started to collect patient data from electronic medical record system of the Qilu Hospital of Shandong University. Total of 1352 patients (female<55 years, male<45 years) who were suspected of CAD underwent coronary angiography from December 2012 to July 2019 were included. 1024 patients were diagnosed with CAD, which is determined by the existence of obstructive stenosis of >50% in any of the main coronary arteries, consisting of the left main coronary artery (LM), right coronary artery (RCA), left anterior descending artery (LAD) and left circumflex coronary artery (LCX) [21]. Patients with incomplete medical records (n=64) or severe disease [including severe cardiac value disease, decompensated heart failure, non-ischemic dilated cardiomyopathy, severe hepatic disease or renal (liver function parameters>3 × upper normal value or serum creatinine>1.4 mg/dL), acute infection or inflammation, autoimmune disease or hematologic disease and malignancy, n=112] were removed from this study. 848 patients were admitted in the study and followed up from October 2022 to November 2022. 582 patients (68.6%) completed the phone follow-up after providing verbal consent (Figure 1).

Definitions and the process of data collection

Practicing clinicians gathered clinical data from medical records. The data included general conditions [age, gender, body mass index (BMI), left ventricular ejection fraction (LVEF), multivessel disease, admission for myocardial infarction(MI) and Gensini score (GS)], cardiovascular risk factors [family history of CAD (FH-CAD), current smoking, hypertension and DM], laboratory tests [total cholesterol (TC), fasting plasma glucose (FPG), low-density lipoprotein cholesterol (LDL-C), triglyceride (TG), high-density lipoprotein cholesterol (HDL-C), uric acid (UA) and serum creatinine (SCr)] and discharge medications [statins, antiplatelet drugs, angiotensin-converting enzyme inhibitors (ACEI), angiotensin receptor blockers (ARB), beta-blockers, oral hypoglycemic drugs and insulins]. To calculate BMI, the weight in kilograms is divided by the square of the height in meters. MI was diagnosed by 12-lead electrocardiography, analysis of serum myocardial enzymes and echocardiograph [22]. The severity of CAD was assessed by the results of coronary angiography and was expressed as GS [23]. Multivessel disease was defined as more than 1 major coronary artery existing ≥50% diameter stenosis. FH-CAD was defined as having a history of CAD in a first-degree relative who is younger than 65 years of age for women or younger than 55 years of age for men. The definition of DM is based on specific glucose levels: a FPG level of 7.0 mmol/L or higher, a random blood glucose level of 11.1 mmol/L or higher and a 2 h plasma glucose after oral glucose tolerance test (OGTT) of 11.1 mmol/L or higher [24] or use of oral hypoglycemic agents or insulin. The definition of hypertension was: systolic blood pressure (SBP) ≥ 140 mmHg and/or diastolic blood pressure (DBP) ≥ 90 mmHg or use of antihypertensive medication. According to the Modification of Diet in Renal Disease (MDRD) equations, the estimated glomerular filtration rate (eGFR) was determined through the application of the following formula: 175 × SCr(mg/dL)^−1.234 × age (year)^−0.179 × 0.79 (if female) [25]. The calculation of METS-IR was determined as follows: ln [(2 × FPG (mg/dL)) + fasting TG (mg/dL)] × BMI (kg/m²) ÷ Ln [HDL-C (mg/dL)] [26].

Endpoints

In the study, the primary endpoint was to determine the rate of MACE in patients with premature CAD. MACE was defined as the composite endpoint encompassing of all-cause, repeat coronary artery revascularization [coronary artery bypass graft surgery (CABG) or percutaneous coronary intervention (PCI)], non-fatal MI, and non-fatal stroke. The secondary endpoints involved all-cause death, non-fatal stroke, non-fatal MI and repeat coronary artery revascularization.

Statistical analysis

Statistical analysis was completed through SPSS version 26.0 and R software version 4.2.1. During the follow-up period, participants were categorized based on the occurrence of MACE. The METS-IR values were divided into three groups by tertiles as follows: Tertile1(METS-IR <39.18), Tertile2(39.18 ≤ METS-IR < 45.01) and Tertile3(METS-IR≥45.01). Continuous variables were presented in terms of their mean ± standard deviation (SD) or their median values along with the 25th and 75th percentiles. For normally distributed continuous variables, the independent samples t-test or ANOVA test was employed. Conversely, for non-normally distributed continuous variables, the Mann-Whitney U-test or Kruskal-Wallis H-test was used. Categorical variables were presented as counts and percentages, and comparisons were made using the chi-square test or Fisher's exact test. Spearman or Pearson correlation analysis was used to explore the correlations between METS-IR and cardiovascular indicators, excluding those indicators used in the calculation of METS-IR, which reflected the degree of coronary stenosis, cardiac function, renal function, and metabolism. To evaluate the cumulative occurrence of MACE and secondary endpoints, Kaplan–Meier curves were utilized. The log-rank test was used to determine whether the distribution of cumulative MACE and secondary endpoints incidence differed among the groups. Variables were analyzed by univariate Cox regression analysis. Multivariate Cox proportional hazards regression was accomplished to determine whether the METS-IR can independently predict the occurrence of MACE. To assess the prediction of METS-IR with multiple confounders in occurrence of MACE, we built three Cox regression models: model 1 was adjusted for gender and age; model 2 was adjusted for variables with p<0.05 in univariate Cox analysis; and model 3 was adjusted for entire variables which involved age, LVEF, gender, admission for MI, GS, multivessel disease, FH-CAD, current smoking, hypertension, DM, TC, LDL-C, UA, eGFR, ACEI/ARB, antiplatelet drugs, statins, oral hypoglycemic drugs and insulins. The METS-IR was incorporated into the models both as continuous variables and categorical variables (the tertile of METS-IR), respectively. The standardization of METS-IR was used to determine the prediction of METS-IR per SD increase. To avoid the result deviation caused by multicollinearity, we calculated the variance inflation factor (VIF) of the variables included in the models. According to VIF<10, there is no multicollinearity in models. We further observed the relationship between METS-IR and outcome events using restricted cubic spline (RCS). Subgroups analysis which was based on gender, current smoking, FH-CAD, DM, and hypertension was completed and the p for interaction was also calculated by multiplication interaction term. The area under the curve (AUC) in time-dependent receiver operating characteristic (ROC) curves were used to explore the predictive value of METS-IR in different times. To determine whether an increased METS-IR had additional predictive value in MACE, repeat coronary artery revascularization and non-fatal MI, model 3 without and with METS-IR were compared in C-Statistic, integrated discrimination improvement (IDI) and continuous net reclassification improvement (NRI). Furthermore, model 3 without and with METS-IR were compared in Bayesian Information Criterion (BIC) and Akaike Information Criterion (AIC) by likelihood ratio test. Furthermore, we excluded endpoints occurring in the first six months of follow-up and performed sensitivity analysis. Statistical significance was determined when the p-value was less than 0.05.

Data sharing statement

The datasets used and/or analyzed in the study are available from the corresponding author upon reasonable request.

Baseline characteristics

The mean age of the patients was 44.43±6.33 years and 352 (60.5%) patients were male. According to the occurrence of MACE, the baseline characteristics were presented in Table 1. Patients with MACE were those tended to be smokers (p=0.022), or to have FH-CAD(p=0.037), DM(p<0.001), hypertension(p=0.024) or multivessel disease(p<0.001). BMI(p=0.010), GS(p=0.023), FPG(p<0.001), TG(p=0.002), HDL-C(p=0.022) and use of oral hypoglycemic drugs (p=0.013) and insulins(p<0.001) were also significantly different between the two groups. Furthermore, patients who experienced MACE had a significantly higher level of METS-IR compared to those who did not experience MACE (44.88±8.11 vs 41.68±6.87, p<0.001). As presented in Table 2, patients were divided into three groups based on the tertiles of METS-IR (Tertile1: n=194, METS-IR < 39.18; Tertile2: n=194, 39.18 ≤ METS-IR < 45.01; Tertile3: n=194, METS-IR≥45.01). Significant differences were found among three groups in terms of age, gender, BMI, GS, current smoking, FH-CAD, DM, FPG, TC, TG, HDL-C, UA, ACEI/ARB, oral hypoglycemic drugs, insulins, MACE, non-fatal MI, repeat coronary artery revascularization (Table 2). In addition, we compared the baseline data of the population finally included in the study and those who lost follow-up and found no significant difference (Supplementary Material: Supplementary Table 1).

Correlations between METS-IR and traditional cardiovascular indicators

The correlation between METS-IR and traditional cardiovascular indicators was tested using Pearson or Spearman correlation analysis. Table 3 showed that METS-IR was positively associated with GS and UA, and negatively associated with age (p<0.05). There was no significant correlation between METS-IR and LVEF, LDL-C, and eGFR.

METS-IR and MACE

During the 63 months (interquartile range, 44-81 months) follow-up, 118 (20.27%) MACEs were recorded, including 8 (1.37%) all-cause death, 38 (6.53%) non-fatal MI, 69 (11.86%) repeat coronary artery revascularization and 4 (0.69%) non-fatal strokes. Kaplan–Meier survival plots (Figure 2) demonstrated that the higher METS-IR group had a significantly elevated cumulative incidence of MACE, non-fatal MI and repeat coronary artery revascularization (MACE: log-rank test, p < 0.001; non-fatal MI: log-rank test, p = 0.011; repeat coronary artery revascularization; log-rank test, p < 0.001). As shown in Table 4, BMI, multivessel disease, current smoking, FH-CAD, DM, hypertension, FPG, TG and METS-IR were related with the risk of MACE, while HDL-C was a protective factor for MACE, use of statins can reduce risk of MACE. In addition, patients who need to take oral hypoglycemic drugs or insulins had a higher risk of MACE. The adjusted HR (95% CI) for risk of MACE with per SD increase in METS-IR was 1.59 (1.32–1.87) (Table 4). Whether regarded as continuous or categorical variable, the METS-IR kept significant after adjusting for confounders. In model 3, per SD increase in METS-IR, a 41.1% (HR=1.41; 95%CI 1.16-1.72) increase was found in the risk of developing MACE. Compared with patients in the lowest tertile, the adjusted HR (model 3) for MACE was 1.58(0.93-2.70) and 2.48(1.45-4.26) in the middle and highest tertile. Significantly increased risk of MACE from tertile1 to tertile3 of METS-IR (p for trend<0.001) was found (Table 5). The associations between METS-IR and non-fatal MI and repeat coronary artery revascularization were also examined (Table 6). RCS was used to further explore the association between METS-IR and MACE, repeat coronary artery revascularization and non-fatal MI (Supplementary Material: Supplementary Figures 1–3).

Subgroup analysis

Subgroup analysis, based on the gender, current smoking, FH-CAD, DM, and hypertension, was used to examine the association between METS-IR and MACE. There was non-significant interaction found in this subgroup analysis. Significance was observed in every subgroup. The evidence suggested that METS-IR was a significant risk factor for incidence of MACE in general premature CAD patients (Figure 3).

Assessment of the prognostic capability of METS-IR for MACE

The AUC was 0.74 at 2 years, 0.69 at 4 years, and 0.63 at 6 years (Figure 4). The incremental predictive value of METS-IR for MACE, repeat coronary artery revascularization and non-fatal MI were presented in Table 7. According to C-Statistic, risk prediction for MACE was significantly improved after adding METS-IR to existing risk prediction model (C-Statistic increased from 0.71 to 0.72, p<0.001). Additionally, the improvement in risk prediction of METS-IR for MACE was also reflected in NRI and IDI [Continuous NRI (95%CI) = 0.19 (-0.01-0.39), p=0.061; IDI (95%CI) = 0.02(0.00-0.03), p=0.007]. Significant improvement was also discovered in risk prediction for non-fatal MI and repeat coronary artery revascularization. Compared with model 3 without METS-IR, the AIC and BIC were significantly improved in predicting MACE and non-fatal MI and significant improvement of AIC was only observed in predicting repeat coronary artery revascularization (Table 8).

Sensitivity analysis

To control the effect of PCI surgical quality and anticoagulant administration in occurrence of study endpoints, we excluded endpoints occurring in first six months of follow-up and performed sensitivity analysis. The results of the sensitivity analysis were consistent with the results of the initial analysis (Supplementary Material: Supplementary Table 2).