Association Between Stress Hyperglycemia Ratio and Cardiovascular Outcomes in Patients Who Underwent Complex Percutaneous Coronary Intervention According to Different Glycemic Status

Abstract

Background

Complex percutaneous coronary intervention is associated with higher risk of long‐term worse cardiovascular outcomes than noncomplex procedures. Stress hyperglycemia ratio (SHR), a biomarker to reflect relative hyperglycemia, has been shown to predict risk of cardiovascular events in patients with coronary artery disease. However, whether SHR‐related cardiovascular risk could be seen in patients who underwent complex percutaneous coronary intervention who are at high risk of cardiovascular events remained undetermined.

Methods

This prospective cohort study consecutively recruited 12 220 patients who underwent complex percutaneous coronary intervention from January 2017 to December 2018 at Fuwai Hospital. Patients were categorized according to baseline SHR tertiles (tertile 1: ≤ 0.765; T2: 0.765–0.879 and T3: > 0.879). The primary end point was cardiovascular events including cardiovascular death, nonfatal myocardial infarction, and nonfatal stroke.

Results

During a median 3‐year follow‐up, 389 cardiovascular events (3.6% of the 10 772 patients who completed follow‐up) were recorded. Overall, per 1‐unit increase in SHR values was associated with 1.62‐fold (95% CI, 1.10–2.38) risk of cardiovascular events after adjusting for confounding factors. Upon stratification by glycemic status, in the fully adjust model, elevated SHR was associated with increased risk of cardiovascular events in diabetic patients (T3 vs. T1; hazard ratio [HR], 1.67 [95% CI, 1.20–2.32]).

Conclusions

We found that elevated SHR levels were associated with increased risk of cardiovascular events at long‐term follow‐up in patients who underwent complex percutaneous coronary intervention, especially in those with diabetes, suggesting that it could help in risk stratification and prognosis in this population.

Nonstandard Abbreviations and Acronyms

ABG

admission blood glucose

IDI

integrated discrimination improvement

SYNTAX

Synergy Between Percutaneous Coronary Intervention With Taxus and Cardiac Surgery

SHR

stress hyperglycemia ratio

Clinical Perspective.

What Is New?

• This study suggests that an elevated stress hyperglycemia ratio, a marker of relative hyperglycemia, may be independently associated with increased long‐term cardiovascular risk in patients undergoing complex percutaneous coronary intervention, particularly among those with diabetes.

What Are the Clinical Implications?

• These findings indicate that stress hyperglycemia ratio could potentially serve as a useful biomarker for refining long‐term cardiovascular risk assessment in high‐risk patients after complex percutaneous coronary intervention, which might inform future strategies for patient monitoring and management.

Complex percutaneous coronary intervention (PCI), accounting for almost 40% of the PCI procedures, was associated with worse clinical outcomes than noncomplex procedures. ¹ Giustino et al., ¹ for example, showed that patients who underwent complex PCI had 1.98‐fold risk of major adverse cardiac events. Thus, it is critical to identify patients undergoing complex PCI who are at a high risk of cardiovascular events so that intense strategies can be provided.

Stress hyperglycemia represents a transient physiologic response to acute diseases. ² There is a growing consensus that stress hyperglycemia is associated with adverse cardiovascular clinical outcomes. ³ , ⁴ However, the admission blood glucose (ABG) level cannot entirely reflect the acute hyperglycemic state, which could also be affected by the chronic glucose level. ³ To better evaluate the actual blood glucose status of patients, researcher has proposed to use the stress hyperglycemia ratio (SHR) to estimate relative hyperglycemia in patients with or without diabetes to identify and quantify stress hyperglycemia. ⁵ Current evidence substantiates that SHR is significantly related to worse cardiovascular outcomes and mortality in patients with coronary artery disease (CAD). ³ , ⁴ , ⁶ , ⁷

However, the relationship between SHR and cardiovascular outcomes in patients who underwent complex PCI remains unclear. Therefore, this study sought to determine the relationship between SHR and adverse cardiovascular events in patients with CAD who underwent complex PCI.

METHODS

The data that support the findings of this study are available from the corresponding author on reasonable request.

Study Design and Population

This study was a prospective, observational cohort study at Fuwai Hospital, Chinese Acedemy of Medical Sciences. The study process was in accordance with the Declaration of Helsinki and was authorized by the Fuwai Hospital Ethics Review Committee. All subjects were informed and signed an informed consent form.

From January 2017 to December 2018, a total of 12 220 patients who underwent complex PCI at Fuwai Hospital were consecutively screened. Complex PCI was defined as having at least 1 of the following features: 3 or more stents implanted, 3 or more lesions treated, bifurcation PCI, total stent length ≥60 mm, left main PCI, or heavy calcification. ⁸ Patients meeting the following criteria were excluded, including severe hepatic or kidney dysfunction, decompensated heart failure, systemic inflammatory disease, malignant tumor, acute infection, death within 7 days after PCI, missing crucial laboratory data (ABG or glycosylated hemoglobin A1c [HbA1c]), or lost to follow‐up.

PCI Procedure and Medication Treatment

All PCI procedures and medical therapies were performed in accordance with the recommendations outlined in the guidelines and at the discretion of the cardiologist, as previously detailed. ⁹ All patients received loading doses of aspirin (300 mg), clopidogrel (600 mg), or ticagrelor (180 mg) before PCI. After the coronary intervention was completed, the characteristics of the coronary disease, including the number of stenotic vessels, unusual types of coronary stenosis, the Synergy Between Percutaneous Coronary Intervention With Taxus and Cardiac Surgery (SYNTAX) score, ¹⁰ and data related to stent implantation, were analyzed and recorded by 2 coronary intervention specialists who were blinded to the baseline data.

Data Collection and Definitions

We prospectively gathered the baseline demographic and clinical data for all patients. Collected demographic and baseline clinical characteristics included age, sex, body mass index (BMI), smoking status, and family history of CAD. We also documented the presence of concurrent comorbidities and any prior history of myocardial infarction (MI) or coronary revascularization (PCI or coronary artery bypass grafting). Clinical data consisted of the main diagnosis on admission; physical, imaging, and laboratory examination findings; and drug regimen at discharge. The ABG was assayed with a LABOSPECT 008 system (Hitachi, Tokyo, Japan), and the HbA1c value was measured with high‐performance liquid chromatography (Tosoh G8 HPLC Analyzer; Tosoh Bioscience, Tokyo, Japan). The concentrations of triglycerides, total cholesterol, low‐density lipoprotein cholesterol (LDL‐C), high‐density lipoprotein cholesterol, fasting blood glucose, and creatinine were analyzed in an enzymatic assay by automated biochemical analyzer (Hitachi 7150, Tokyo, Japan). The hsCRP (high‐sensitivity C‐reactive protein) was examined with standard biochemical techniques at the core laboratory of Fuwai Hospital. The modified biplane Simpson rule was used to assess left ventricular ejection fraction at rest. ¹¹

Estimated average chronic glycemic level was calculated with the formula [(1.59 × HbA1c %)−2.59] mmol/L, ¹² and SHR was defined as ABG (mmol/L) divided by the estimated average chronic glycemic value. ⁵ Although derived from a Western cohort, this formula used to estimate average chronic glucose has been subsequently used in studies of Asian populations investigating glycemic status and cardiovascular risk, supporting its applicability in our analysis. ¹³ , ¹⁴ Based on prior literature on stress hyperglycemia and considerations of clinical practicality, participants were stratified into 3 groups according to tertiles of the SHR (tertile 1: ≤ 0.765; T2: 0.765–0.879, and T3: > 0.879). ¹⁵ , ¹⁶ , ¹⁷

Diabetes was recorded if the patient was previously diagnosed with diabetes, received glucose‐lowering therapy or had fasting blood glucose ≥7.0 mmol/L, HbA1c ≥6.5%, or 2‐hour plasma glucose ≥11.1 mmol/L in an oral glucose tolerance test. Hypertension was defined as systolic blood pressure ≥140 mm Hg, diastolic blood pressure ≥90 mm Hg, or the use of antihypertensive therapy. ¹⁸ Stroke was defined as a previous history of cerebral bleeding, ischemic stroke, or transient ischemic attack. Moreover, we categorized the diagnosis of chronic coronary syndrome or acute coronary syndrome (ACS) on admission according to the latest relevant guidelines. ¹⁹ , ²⁰ Left main disease was determined by >50% stenosis in the left main coronary artery. Three‐vessel disease was defined as >50% stenosis in 3 vessels located in different epicardial vascular systems (eg, the left anterior descending branch, left circumflex branch, and right coronary artery).

Follow‐Up and End Point Definitions

Patients were followed up at 6‐month intervals for 3 years after discharge. Follow‐up data were collected via medical record review, clinical visits, or telephone interviews by trained investigators who were blinded to the clinical data. The primary end point was the cardiovascular event, a composite of cardiovascular death and nonfatal MI. Secondary end point was the major adverse cardiovascular event (MACE), composite of cardiovascular death, nonfatal MI, and unplanned revascularization. Death was considered cardiovascular‐caused unless unequivocal noncardiovascular cause could be established. Nonfatal MI was defined as positive cardiac troponins with typical chest pain, typical ECG serial changes, identification of an intracoronary thrombus by angiography or autopsy, or imaging evidence of new loss of viable myocardium or a new regional wall‐motion abnormality. ²¹ Unplanned revascularization was defined as revascularization for a lesion not meeting the ischemic threshold at the index procedure and not planned for staged revascularization following the index procedure.

Statistical Analysis

Continuous variables are expressed as mean±SD. Categorical variables are presented as number (percentage). The Kolmogorov–Smirnov test was used to test the distribution pattern. The differences of baseline characteristics between groups were analyzed with the Student's t test, Mann–Whitney U test, Kruskal–Wallis H test, or chi‐square test where appropriate. To assess potential attrition bias, baseline characteristics were compared between patients who completed the 3‐year follow‐up and those who were lost to follow‐up. The cumulative incidence of cardiovascular events according to SHR tertiles was estimated using the Kaplan–Meier method, and differences between groups were compared using the log‐rank test. Curves are presented with 95% confidence bands to illustrate the precision of the estimates. Univariable and multivariable Cox regression models were used to calculate the hazard ratios (HRs) and 95% CIs. Covariates for the multivariable Cox regression models were selected based on clinical relevance and a review of the literature on prognostic factors in patients undergoing complex PCI. ²² , ²³ These potential confounders included age, male, BMI, hypertension, diabetes, ACS, previous MI, previous revascularization, current smoker, previous stroke, left ventricular ejection fraction, total cholesterol, LDL‐C, hsCRP, serum creatinine, SYNTAX score, stent number, aspirin use, clopidogrel use, and statins use. In Cox regression models, SHR tertiles were entered nominally with T1 as the reference category, and HRs for T2 and T3 were estimated separately. The proportional hazards assumption for all covariates included in the Cox regression models was assessed using Schoenfeld residuals. The individual P values >0.05 for each covariate indicate no significant departure from the proportional hazards assumption (Table S1). Restricted cubic spline plots were created to assess linearity assumptions of the relationship between SHR and clinical end points. The overall significance of the association and potential nonlinearity were evaluated. In the restricted cubic spline model, we also adjusted for age and sex. The association between SHR tertiles and the presence of high coronary anatomical complexity, defined as a baseline SYNTAX score >33, ²⁴ was examined using univariable and multivariable logistic regression. Odds ratios and 95% CIs were calculated with the lowest SHR tertile (T1) as the reference. Improvements in risk discrimination were assessed using Harrell's C‐statistic, and the integrated discrimination improvement (IDI) modified for survival analyses. ²⁵ To estimate the precision of the change in C‐statistic (ΔC) between nested models, we performed bootstrap resampling with 500 replicates to derive a 95% CI for ΔC. To explore a clinically actionable threshold for risk stratification, a receiver operating characteristic curve analysis was performed. SHR was entered as a continuous variable to predict the binary outcome of whether a cardiovascular event occurred during the 3‐year follow‐up. The optimal cutoff value was determined by maximizing the Youden index. Subgroup analyses were also performed to assess the influence of SHR on cardiovascular events in different subgroups. To assess the potential impact of survival bias, a sensitivity analysis was performed by including patients who died within 7 days after the index PCI procedure, who were excluded from the primary cohort. To assess the potential confounding effect of antidiabetic treatment, we conducted a sensitivity analysis in which the use of insulin (yes/no) and oral hypoglycemic medication (yes/no) were added as covariates to the primary multivariable Cox regression model. Two‐tailed P values <0.05 were considered as statistically significant. All statistical analyses were performed using R version 4.0.2 (The R Foundation).

RESULTS

Baseline Characteristics

A total of 10 772 patients underwent complex PCI and received follow‐up were enrolled in our study (Figure 1). The average age was 60.23±10.12 years, 8447 (78.4%) patients were men, 6651 (61.7%) patients were diagnosed with ACS, 7018 (65.2%) patients suffered with hypertension, 5065 (40.7%) subjects were diagnosed with diabetes, and 3338 (30.1%) of patients were current smokers (Table 1). Baseline characteristics of patients retained in the study versus those lost to follow‐up (n=237, 2.2%) are compared. Although minor differences were noted in a few parameters, all were adjusted for in the multivariable analysis (Table S2).

Figure 1. Study flow chart.

CAD indicates coronary artery disease; CVE, cardiovascular events; and PCI, percutaneous coronary intervention.

Table 1.

Baseline Characteristics According to Cardiovascular Events^*

	Overall (N=10 772)	Survivors (N=10 383)	Events (N=389)	P value
Age, y	60.23±10.12	60.11±10.06	63.58±11.03	<0.001
Male sex	8447 (78.4)	8154 (78.5)	293 (75.3)	0.147
Body mass index, kg/m2	25.97±3.21	25.99±3.19	25.42±3.68	0.001
Hypertension	7018 (65.2)	6740 (64.9)	278 (71.5)	0.009
Diabetes	5065 (47.0)	4850 (46.7)	215 (55.3)	0.001
Clinical presentation				0.001
Chronic coronary syndrome	4121 (38.3)	4005 (38.6)	116 (29.8)
Acute coronary syndrome	6651 (61.7)	6378 (61.4)	273 (70.2)
Previous myocardial infarction	2855 (26.5)	2711 (26.1)	144 (37.0)	<0.001
Family history of coronary artery disease	1250 (11.6)	1203 (11.6)	47 (12.1)	0.826
Previous revascularization†	2727 (25.3)	2600 (25.0)	127 (32.6)	0.001
Current smoker	3338 (31.0)	3208 (30.9)	130 (33.4)	0.317
Previous stroke	1470 (13.6)	1401 (13.5)	69 (17.7)	0.02
Peripheral artery disease	741 (6.9)	704 (6.8)	37 (9.5)	0.047
Left ventricular ejection fraction, %	61.56±6.89	61.67±6.72	58.51±9.87	<0.001
Admission blood glucose, mmol/L	6.68±2.56	6.65±2.52	7.50±3.25	<0.001
Glycosylated hemoglobin A1c, %	6.54±1.27	6.53±1.27	6.82±1.45	<0.001
Total cholesterol, mmol/L	4.05±1.07	4.05±1.07	4.06±1.11	0.881
Triglycerides, mmol/L	1.74±1.12	1.74±1.12	1.73±1.15	0.879
Low‐density lipoprotein cholesterol, mmol/L	2.45±0.92	2.45±0.92	2.46±0.97	0.805
High‐density lipoprotein cholesterol, mmol/L	1.09±0.29	1.09±0.29	1.09±0.29	0.661
High‐sensitivy C‐reactive protein, mg/L	2.77±3.20	2.75±3.18	3.35±3.46	<0.001
Serum creatinine, μmol/L	83.23±17.85	83.02±17.36	88.81±27.45	<0.001
Angiographic and procedural data
Synergy Between Percutaneous Coronary Intervention With Taxus and Cardiac Surgeryscore	17.17±9.17	17.10±9.16	18.86±9.20	0.021
Severe calcification	811 (7.5)	761 (7.3)	50 (12.9)	<0.001
Left main disease	2076 (19.3)	1985 (19.1)	91 (23.4)	0.042
Three‐vessel disease	5961 (55.3)	5731 (55.2)	230 (59.1)	0.139
Chronic total occlusion lesion	1424 (13.2)	1373 (13.2)	51 (13.1)	1
Ostial lesion	1694 (15.7)	1625 (15.7)	69 (17.7)	0.299
Thrombotic lesion	225 (2.1)	209 (2.0)	16 (4.1)	0.008
Type B2/C lesion	8499 (78.9)	8186 (78.8)	313 (80.5)	0.48
Stent number	4.35±2.90	4.36±2.92	3.88±2.43	0.001
Total stent length	63.89±33.78	64.04±33.81	60.06±32.82	0.023
Treated lesion number	1.91±0.82	1.91±0.81	1.86±0.85	0.268
Medications
Aspirin	7750 (71.9)	7486 (72.1)	264 (67.9)	0.077
Clopidogrel	8978 (83.3)	8650 (83.3)	328 (84.3)	0.649
Statins	10 455 (97.1)	10 076 (97.0)	379 (97.4)	0.772
Angiotensin‐converting enzyme inhibitor/angiotensin II receptor blocker	2750 (25.5)	2658 (25.6)	92 (23.7)	0.420
β‐blocker	9630 (89.4)	9280 (89.4)	350 (90.0)	0.770
Calcium channel blocker	3747 (34.8)	3608 (34.7)	139 (35.7)	0.730

^*Values are expressed as mean±SD and count (percentage).

^† Revascularization included percutaneous coronary intervention and coronary artery bypass grafting.

Baseline characteristics of participants stratified by SHR tertiles are presented in Table 2. Patients in the highest SHR tertile (T3) were more likely to present with ACS and to be current smokers. They also had a higher prevalence of family history of CAD. Biochemically, this group exhibited higher levels of fasting blood glucose, triglycerides, and hsCRP, alongside lower levels of HbA1c. Regarding coronary anatomy, T3 patients had higher SYNTAX scores and a greater prevalence of left main disease, 3‐vessel disease, and thrombotic lesions. Furthermore, continuous SHR was significantly correlated with traditional cardiovascular risk factors, including age, BMI, left ventricular ejection fraction, ABG, HbA1c, total cholesterol, triglycerides, LDL‐C, high‐density lipoprotein cholesterol, hsCRP, and serum creatinine (Table S3). Distribution of SHR could be obtained in Figure S1.

Table 2.

Baseline Characteristics According to SHR Tertiles^*

Characteristics	SHR T1, ≤0.765 (N=3602)	SHR T2, 0.765–0.879 (N=3579)	SHR T3, >0.879 (N=3591)	P value
Age, y	60.52±10.28	60.13±10.07	60.05±10.01	0.114
Male sex	2796 (77.6)	2793 (78.0)	2858 (79.6)	0.103
Body mass index, kg/m2	25.96±3.29	26.00±3.18	25.96±3.17	0.859
Hypertension	2343 (65.0)	2289 (64.0)	2386 (66.4)	0.086
Diabetes	1624 (45.1)	1161 (32.4)	2280 (63.5)	<0.001
Clinical presentation				<0.001
Chronic coronary syndrome	1409 (39.1)	1447 (40.4)	1265 (35.2)
Acute coronary syndrome	2193 (60.9)	2132 (59.6)	2326 (64.8)
Previous myocardial infarction	970 (26.9)	888 (24.8)	997 (27.8)	0.014
Family history of coronary artery disease	381 (10.6)	411 (11.5)	458 (12.8)	0.015
Previous revascularization†	920 (25.5)	889 (24.8)	918 (25.6)	0.725
Current smoker	1194 (33.1)	1087 (30.4)	1057 (29.4)	0.002
Previous stroke	514 (14.3)	467 (13.0)	489 (13.6)	0.320
Peripheral artery disease	251 (7.0)	237 (6.6)	253 (7.0)	0.752
Left ventricular ejection fraction, %	61.51±6.78	62.00±6.52	61.15±7.31	<0.001
Laboratory tests
Admission blood glucose, mmol/L	5.45±1.25	6.03±1.37	8.55±3.28	<0.001
HbA1c, %	6.72±1.30	6.26±1.03	6.65±1.41	<0.001
Total cholesterol, mmol/L	3.98±1.05	4.12±1.07	4.06±1.08	<0.001
Triglycerides, mmol/L	1.68±1.01	1.73±1.11	1.80±1.23	<0.001
Low‐density lipoprotein cholesterol, mmol/L	2.39±0.89	2.50±0.93	2.46±0.94	<0.001
High‐density lipoprotein cholesterol, mmol/L	1.09±0.29	1.11±0.29	1.08±0.28	<0.001
High‐sensitivity C‐reactive protein, mg/L	2.84±3.18	2.63±3.06	2.84±3.36	0.025
Serum creatinine, μmol/L	83.08±17.16	83.03±17.00	83.59±19.30	0.344
Angiographic and procedural data
Synergy Between Percutaneous Coronary Intervention With Taxus and Cardiac Surgery score	16.24±4.93	16.32±5.32	16.59±5.60	0.015
Severe calcification	264 (7.3)	263 (7.3)	284 (7.9)	0.572
Left main disease	693 (19.2)	646 (18.0)	737 (20.5)	0.029
Three‐vessel disease	1996 (55.4)	1919 (53.6)	2046 (57.0)	0.017
Chronic total occlusion lesion	484 (13.4)	506 (14.1)	434 (12.1)	0.033
Ostial lesion	571 (15.9)	542 (15.1)	581 (16.2)	0.469
Thrombotic lesion	42 (1.2)	42 (1.2)	141 (3.9)	<0.001
Type B2/C lesion	2798 (77.7)	2853 (79.7)	2848 (79.3)	0.081
Treated lesion number	1.92±0.84	1.88±0.79	1.92±0.82	0.050
Stent number	4.45±3.24	4.25±2.66	4.34±2.78	0.020
Total stent length	64.30±34.39	62.75±32.19	64.62±34.68	0.044
Medications
Aspirin	2647 (73.5)	2605 (72.8)	2498 (69.6)	<0.001
Clopidogrel	3043 (84.5)	3016 (84.3)	2919 (81.3)	<0.001
Statins	3487 (96.8)	3488 (97.5)	3480 (96.9)	0.215
Angiotensin‐converting enzyme inhibitor/angiotensin receptor blocker	957 (26.6)	882 (24.6)	911 (25.4)	0.168
β‐blocker	3192 (88.6)	3173 (88.7)	3265 (90.9)	0.001
Calcium channel blocker	1282 (35.6)	1228 (34.3)	1237 (34.4)	0.457

SHR indicates stress hyperglycemia ratio.

^*Values are expressed as mean ± SD and count (percentage).

^† Revascularization included percutaneous coronary intervention and coronary artery bypass grafting.

Clinical Outcomes for Adverse Cardiovascular Events

A total of 389 cardiovascular events and 328 MACEs were recorded (Table 3). Compared with nonevent participations, patients suffered with cardiovascular events were more likely to be older, suffering with hypertension and diabetes, presented with ACS, involved with left main disease, had higher levels of ABG, HbA1c, serum creatinine, and SYNTAX score (Table 1).

Table 3.

SHR in Relation to cardiovascular Events

	Events (%)	Univariable analysis		Multivariable analysis*
		HR (95% CI)	P value	HR (95% CI)	P value
Cardiovascular events†
Continuous	389 (3.6)	2.44 (1.67–3.58)	<0.001	1.62 (1.10–2.38)	0.015
SHR tertiles
T1	112 (3.1)	Reference	…	Reference	…
T2	109 (3.0)	0.98 (0.75–1.27)	0.857	1.03 (0.79–1.35)	0.808
T3	168 (4.7)	1.52 (1.19–1.92)	<0.001	1.42 (1.11–1.81)	0.005
MACE‡
Continuous	328 (3.0)	2.61 (1.74–3.91)	<0.001	1.59 (1.05–2.41)	0.029
SHR tertiles
T1	93 (2.6)	Reference	…	Reference	…
T2	90 (2.5)	0.97 (0.73–1.30)	0.840	1.04 (0.78–1.39)	0.784
T3	145 (4.0)	1.58 (1.21–2.04)	<0.001	1.44 (1.10–1.87)	0.008

HR indicates hazard ratio; and SHR, stress hyperglycemia ratio.

^*Adjusted for age, male, body mass index, hypertension, diabetes, acute coronary syndrome, previous myocardial infarction, previous revascularization, current smoker, previous stroke, left ventricular ejection fraction, total cholesterol, low‐density lipoprotein cholesterol, high‐sensitivity C‐reactive protein, serum creatinine, Synergy Between Percutaneous Coronary Intervention With Taxus and Cardiac Surgeryscore, stent number, aspirin use, clopidogrel use, and statins use.

^† Cardiovascular events were defined as a composite of cardiovascular death, nonfatal myocardial infarction and nonfatal stroke.

^‡ MACEs were defined as the composite of cardiovascular death and nonfatal myocardial infarction.

The prevalence of cardiovascular events in the SHR T1, T2, and T3 groups were 112 (3.1%), 109 (3.0%), and 168 (4.7%) respectively, and the prevalence of MACEs in the SHR T1, T2, and T3 groups were 93 (2.6%), 90 (2.5%), and 145 (4.0%) respectively (Table 3). Kaplan–Meier survival analyses showed a significant difference in the incidence of cardiovascular events and MACEs among the 3 groups at the 3‐year follow‐up, with the highest cardiovascular events and MACEs rate in SHR T3 (all P values <0.001). Details of the Kaplan–Meier survival analyses are presented in Figure 2.

Figure 2. Kaplan–Meier curves for (A) cardiovascular events and (B) MACEs among SHR tertiles.

Shaded areas represent 95% confidence bands. P value was calculated by the log‐rank test. These curves represent unadjusted cumulative event rates. MACE indicates major adverse cardiovascular event; and SHR, stress hyperglycemia ratio.

Results of the restricted cubic spline analyses indicated significant overall associations of SHR with both cardiovascular events and MACE rates at 3‐year follow‐ups after adjustment for age and sex (all overall P<0.001). All P values for non‐linearity were >0.05, confirming that linear relationships adequately describe these associations (Figure S2). The multivariable Cox regression analyses results showed that in comparisons with SHR T1 subjects, the multivariable‐adjusted HR for cardiovascular events and MACEs at the 3‐year follow‐up were 1.52 (95% CI, 1.19–1.92) and 1.58 (95% CI, 1.21–2.04) for SHR T3 subjects respectively. No difference could be seen for the risk of cardiovascular events or MACEs between SHR T1 and T2 subjects. Moreover, the HR per unit increase of SHR in predicted cardiovascular event was 1.62 (95% CI, 1.10–2.38), and in predicted MACE was 1.59 (95% CI, 1.05–2.41) (Table 3).

SHR and Cardiovascular Events According to Different Glucose Metabolism Status

Table S4 and Figure 3 show the results of a stratified multivariable Cox regression analysis of SHR and cardiovascular events according to different glucose metabolism status. For patients with diabetes, SHRs T3 were significantly associated with cardiovascular events (HR, 1.67 [95% CI, 1.20–2.32]) and MACEs (HR, 1.72 [95% CI, 1.20–2.47]). For those without diabetes, no significant difference of risk of cardiovascular events and MACEs was observed among SHR tertile groups (all P>0.05).

Figure 3. SHR in relation to (A) cardiovascular events and (B) MACEs according to different glucose metabolism status.

HR indicates hazard ratio; MACE, major adverse cardiovascular event; and SHR, stress hyperglycemia ratio.

Incremental Effect of SHR on Risk Discrimination With Different Glucose Metabolism Status

For patients with diabetes, the C‐statistic value of the original model with traditional cardiovascular risk factors (including age, male sex, BMI, hypertension, ACS, previous MI, previous revascularization, current smoker, previous stroke, left ventricular ejection fraction, total cholesterol, LDL‐C, hsCRP, serum creatinine, SYNTAX score, stent number, aspirin use, clopidogrel use, and statins use) was 0.674 (95% CI, 0.653–0.725). Adding SHR to this model increased the C‐statistic to 0.686 (95% CI, 0.668–0.735). The bootstrap 95% CI for ΔC was 0.002 to 0.030, and the IDI was 0.10% (P<0.001), indicating a significant improvement in risk discrimination (Table S5). In contrast, for patients without diabetes, adding SHR did not significantly improve the model (ΔC 95% CI, −0.003 to 0.010; IDI P=0.333) (Table S5).

Optimal SHR Cutoff Value and Associated Cardiovascular Risk

The receiver operating characteristic analysis identified an optimal SHR cutoff value of 0.978 for predicting cardiovascular events. When patients were dichotomized using this threshold, those with SHR >0.978 had a significantly increased risk of cardiovascular events (HR, 1.66 [95% CI, 1.30–2.11]) and MACEs (HR, 1.62 [95% CI, 1.25–2.10]) compared with those with SHR ≤0.978 in the fully adjusted model (Table S6).

Association Between SHR and High Anatomical Complexity

The higher SHR was associated with increased odds of having a SYNTAX score >33. Specifically, patients in the highest SHR tertile (T3) had a greater risk (adjusted odds ratio, 1.51; 95% CI, 1.07–2.12, P=0.019) of a SYNTAX score >33 compared with those in the lowest tertile (T1) (Table S7).

Incremental Effect of SHR on SYNYAX Score for Predicting the Adverse Cardiovascular Events Risk

Adding SHR to a baseline model containing only the SYNTAX score improved the C‐statistic from 0.524 (95% CI, 0.502–0.549) to 0.567 (95% CI, 0.537–0.597). The ΔC was 0.044 (95% CI, 0.016–0.080), and the IDI was 0.17% (P<0.001). Similar results were observed for the prediction of MACEs (ΔC: 0.055 [95% CI, 0.023–0.090]; IDI 0.17%, P<0.001) (Table S8).

Subgroup Analysis for the Association Between Continuous SHR and Study End Points

Subgroup analyses were performed to evaluate the association of continuous SHR with cardiovascular events across different populations. No significant interaction was observed for any of the tested subgroups (all P for interaction >0.05). However, point estimates for the association between SHR and cardiovascular events risk were nonsignificant with wide CIs in several subgroups, including those with BMI <25 kg/m², chronic coronary syndrome, absence of diabetes, absence of hypertension, LDL‐C <1.8 mmol/L, and hsCRP ≥2 mg/L (Figure 4, Table S9).

Figure 4. Subgroup analysis for the association between continuous SHR and study end points.

ACS indicates acute coronary syndrome; BMI, body mass index; CAD, coronary artery disease; CCS, chronic coronary syndrome; HR, hazard ratio; hsCRP, high‐sensitivity C‐reactive protein; LDL‐C, low‐density lipoprotein cholesterol; and SHR, stress hyperglycemia ratio

Sensitivity Analysis

First, we included patients who died within 7 days post PCI. Similar to the primary analysis, SHR T3 was significantly associated with cardiovascular events (HR, 1.40 [95% CI, 1.10–1.78]) and MACEs (HR, 1.42 [95% CI, 1.09–1.84]) after including patients who died within 7 days post PCI (Table S10). Then, we additionally adjusted for the use of insulin and oral hypoglycemic medication, which yielded results consistent with the primary analysis, indicating SHR T3 was significantly associated with cardiovascular events (HR, 1.44 [95% CI 1.13–1.84]) and MACEs (HR, 1.46 [95% CI, 1.12–1.91]) (Table S11).

DISCUSSION

In the current study, the association between SHR and adverse cardiovascular events in patients who underwent complex PCI was evaluated, revealing the following 2 findings: (1) SHR was independently associated with cardiovascular events and MACEs in patients who underwent complex PCI; (2) when considering glucose metabolism status, SHR was significantly associated with cardiovascular events and MACEs in patients with diabetes, associations that could not been seen in patients without diabetes; and (3) moreover, adding SHR to the original models with traditional cardiovascular risk factors improved the risk prediction for cardiovascular events in the group with diabetes but not in the group without diabetes. Our study provided important information supporting that elevated SHR was associated with higher risk of cardiovascular events in patients underwent complex PCI and demonstrated, for the first time, that such association was more pronounced in patients with diabetes, suggesting SHR could help in risk stratification and prognosis in this population.

As a result of an ageing population, the volume of complex procedures has steadily increased. ²⁶ Complex CAD often goes along with extensive comorbidities, especially abnormal glucose metabolism status, chronic renal disease, and heart failure. ²⁷ Complex PCI may have less favorable outcomes than simpler procedures. In the e‐Ultimaster Registry, 9793 patients who underwent complex PCI had more target lesion failure (HR, 1.41), cardiac death (HR, 1.28), and target vessel myocardial infarction (HR, 1.48) than 26 056 patients undergoing noncomplex PCI. ²⁸ In the observational PROMETHEUS study, complex PCI was associated with higher risks of death, myocardial infarction, and unplanned revascularization at both 90 days and 1 year. ²⁹ However, the biomarkers for postcomplex PCI monitoring had not been evaluated systematically. It was still lack of evidence in the field of prognosis factor for patients who underwent PCI.

Stress hyperglycemia and SHR has been proven to be a strong predictor of a poor prognosis in patients with CAD who underwent PCI. ³ , ⁶ , ³⁰ For acute disease, Yang et al. ³ found there were U‐shaped associations of SHR with MACE rate at 2‐year follow‐ups and J‐shaped associations of SHR with in‐hospital cardiac death and MI and that at 2‐year follow‐up in patients with ACS who underwent PCI. For chronic diseases, Xu et al. ⁶ focused on patients with chronic coronary syndrome and reported a significant association of a high SHR with elevated in‐hospital mortality, further indicating the potential prognostic value of the SHR in patients without stress conditions. However, for patients underwent complex PCI, there was no research elucidating the association of the SHR with long‐term prognosis thus far. Herein, we focused on this point for the first time and revealed that a high SHR was associated with an increase in long‐term cardiovascular risk in patients underwent complex PCI.

Interestingly, in our study, we found the association with the SHR seems to be different in patients with coronary heart disease with differing glycemic status (with or without diabetes). Previous studies did not come to a consensus in the application of SHR to patients with or without diabetes. Kojima et al. ³¹ found that high SHR was significantly associated with a poorer long‐term prognosis in individuals without diabetes. Cui et al. ⁴ also demonstrated a strong positive correlation between SHR and long‐term mortality in patients with acute MI with or without diabetes. However, Schmitz et al. ³² advocated that SHR was significantly associated with higher short‐term mortality in patients with AMI and was found to be associated with higher long‐term mortality only in patients with diabetes. Those studies did not take the complexity of CAD into the consideration, which might result in the heterogeneity of SHR in their study population. Zhang et al. ⁷ investigated the association between the SHR and severity of coronary artery disease under different glucose metabolic states, which found high SHR was closely related to multivessel CAD, especially in the population with diabetes. Current, no study had not demonstrated the association of SHR and adverse cardiovascular events risk in patients with complex CAD underlying different glucose metabolic status. Our stratification analysis based on the patients with or with diabetes in baseline, revealing that an elevated SHR only served as a prognostic indicator in patients with diabetes. This finding suggests that the predictive value of the SHR for cardiovascular events might be influenced by the abnormal glucose metabolic status.

Beyond its prognostic role, we observed that a higher SHR was associated with greater coronary anatomical complexity (SYNTAX score >33), a threshold at which current guidelines favor coronary artery bypass grafting over PCI. ²⁴ Although our data are observational and cannot inform treatment decisions, this association raises the hypothesis that SHR might help identify a phenotype of patients with diabetes who are more likely to have anatomy suitable for guideline‐directed coronary artery bypass grafting. This possibility warrants investigation in future studies designed to evaluate the interplay between metabolic stress, coronary complexity, and long‐term outcomes of different revascularization strategies.

Our observed association between elevated SHR and adverse cardiovascular outcomes may be partially mediated by its influence on the procedural success and durability of PCI. Acute hyperglycemia, reflected by a high SHR, is associated with endothelial dysfunction, heightened oxidative stress, and a prothrombotic state. ³³ This biological milieu may compromise microvascular perfusion, increase the risk of periprocedural complications (eg, no‐reflow, distal embolization), and potentially hinder the achievement of optimal or complete revascularization. ³⁴ Furthermore, the metabolic stress indicated by SHR could be a marker of more diffuse, unstable, or rapidly progressive coronary disease, which poses inherent technical challenges for complete anatomical revascularization via PCI. ³⁵ Consequently, a higher SHR might identify patients in whom the metabolic environment and coronary plaque characteristics interact to limit the efficacy and completeness of PCI, thereby contributing to their elevated long‐term risk. ³⁵ Future studies integrating detailed procedural data are warranted to directly examine this pathway.

Although no statistically significant interaction was found, the association between SHR and cardiovascular events was not statistically significant in certain subgroups, such as patients with chronic coronary syndrome, patients without diabetes, and those without hypertension. The wide CIs in these analyses likely reflect limited statistical power due to smaller sample sizes or lower event rates within these specific strata. Consequently, the absence of a statistically significant association in these subgroups should not be interpreted as evidence of a null effect but rather underscores the uncertainty of the estimate within these smaller cohorts. The overall lack of significant interaction suggests that the direction of the association is largely consistent, but its magnitude and precision may vary across patient characteristics. These findings highlight that the prognostic value of SHR is most clearly established in the broader, higher‐risk population with diabetes undergoing complex PCI, and further large‐scale studies would be needed to definitively characterize its role in specific, lower‐risk subgroups.

This study, using a large‐scale clinical cohort, validates the prognostic value of the SHR in patients undergoing complex PCI and provides a clinically actionable threshold (0.978) for risk stratification. However, this study still has several limitations. First, the study design employed in this research was prospective and observational, as opposed to a randomized controlled trial. Consequently, it was not feasible to establish a definitive causal relationship between the SHR in conjunction with glucose metabolic status and the incidence of cardiovascular events. Second, dynamic changes in the SHR and glucose metabolic status during follow‐up were not present in our study. It was still unknown about the association between changes of SHR and glucose metabolic status and prognosis for CAD population who underwent complex PCI. Third, the study could not comprehensively evaluate all glucose metabolic factors and parameters, duo to the limitation of data collocation. Fourth, despite controlling for potential confounders as covariates in multivariable regression models, it is important to acknowledge that the impact of uncollected confounders cannot be completely disregarded. Consequently, further larger‐scale studies are required to validate the findings of this study.

CONCLUSIONS

To our knowledge, this is the first study to demonstrate that elevated SHR levels were associated with increased risk of cardiovascular events at long‐term follow‐up in patients underwent complex PCI, especially in those with diabetes, suggesting that it could help in risk stratification and prognosis in this population.

Sources of Funding

This study is supported by the Noncommunicable Chronic Diseases‐National Science and Technology Major Project (2025ZD0548200).

Disclosures

None.

Supporting information

Tables S1–S11

Figures S1–S2

STROBE Checklist