Association between accelerated biological aging and major adverse cardiovascular events in patients with stable coronary artery disease

Abstract

Background

Accelerated biological aging is characterized by an expedited deterioration of physiological systems. Phenotypic age (PA) derived from chronological age and 9 clinical biomarkers serves as a robust measure of this process, and phenotypic aging acceleration (PAA) was the discrepancy between phenotypic and chronological age. However, its association with major adverse cardiovascular events (MACEs) in stable coronary artery disease (SCAD) remains unclear.

Methods

This cohort study enrolled 8672 patients with SCAD who underwent initial percutaneous coronary intervention (PCI). Accelerated biological aging was assessed using phenotypic aging calculations, and MACEs were defined as death (all-cause and cardiac mortality), non-fatal myocardial infarction (MI), non-fatal stroke, and unplanned revascularization. A cox proportional hazard model was employed for analysis.

Results

Overall, the subjects were 64.7 ± 6.5 years old while the PA were 62.7 ± 12.3 years and the PAA was −2.8 (−6.4, 1.5) years. During overall 2.2 (1.4, 3.1) follow up years, 1147 MACEs were documented (165 all-cause mortality, 47 cardiac mortality, 918 unplanned revascularization, 43 non-fatal MI and 80 non-fatal strokes). Compared with phenotypically younger group (PAA < 0), phenotypically older group (PAA ≥ 0) showed higher risk for MACEs and components (all P values < 0.05). A positive dose-response association was identified between PAA and MACEs. For each 1-year PAA, there were 2.80%, 8.80%, 10.80%, 1.20%, 5.80%, 4.50% higher risk of MACEs, all-cause mortality, cardiac mortality, unplanned revascularization, non-fatal MI and non-fatal stroke, respectively (all P values < 0.05). Robust results were demonstrated in subgroup analysis.

Conclusion

In SCAD patients undergoing PCI, accelerated biological aging is associated with an increased risk of MACEs, particularly death (all-cause and cardiac mortality) and unplanned revascularization.

1. Introduction

Aging is a prominent risk factor for vascular disease and subsequent cardiovascular events, remaining a leading cause of death globally [1,2]. Genetics of aging research has uncovered a complex genetic pathway networks and aging was manifested as cellular senescence, immune-senescence and protein homeostasis degradation [3]. The burgeoning issue of an aging population coupled with the potential for anti-aging therapies has intensified the need to accurately discern the trajectory of various physiological systems and quantify aging [4]. The biological age proposed by Klemera and Doubal [5,6], an epigenetic biomarker telomere length [7,8] and phenotypic age [7,9] have been used to estimate aging. Moreover, algorithms incorporating information from standard clinical parameters have emerged as more accurate predictors of morbidity and mortality than chronological age [2,9,10]. Phenotypic age (PA) is calculated using specific algorithms composed of chronological and 9 clinical biomarkers [7]. Phenotypic aging acceleration (PAA) is defined as the discrepancy between PA and chronological age, reflecting whether biological aging is accelerated or decelerated compared to chronological age. Aging acceleration has been found to be associated with all-cause and cardiac mortality [11].

Coronary artery disease (CAD) is tightly associated with aging and was the leading cause of death around the world [3,12]. Percutaneous Coronary Intervention (PCI) is widely used in CAD patients and nearly 20% of patients with stable coronary artery disease (SCAD) patients treated with PCI [13]. Despite strict adherence to current guideline-recommended therapies, the dynamic and unpredictable SCAD progression can still lead to unexpectedly lead major adverse cardiovascular events (MACEs) [14]. Senescence cardiac cells might directly contribute to coronary vascular disease such as atherosclerosis and MI [1,15].

PAA has been identified as a mediator or predictor of cardiovascular disease, cancer, and mortality in some other study [16,17]. However, there is a scarcity of clinical research focusing on the association between PAA and the incidence of MACEs events in SCAD patients who have undergone PCI procedure Therefore, we aimed to evaluate the association of PAA with MACEs among SCAD patients after PCI procedure.

2. Methods

2.1. Study population

The present study utilized a cohort design and strictly followed the STROBE guideline for cohort studies [18]. Initially, 14,262 eligible CAD patients who underwent PCI and followed up at Sir Run Run Shaw Hospital were identified. Demographic features and clinical data were obtained from the Hospital Information System (HIS) and telephone follow-up. According to 2020 ESC Guidelines for the management of acute coronary syndromes (ACS), cardiac troponin I (cTnI) < the 99th of normal cTnI was considered SCAD and we included 11,623 SCAD patients [19]. Exclusion criteria were listed as follows: (1) patients with missing height, weight, body mass index (BMI), blood pressure (n = 839); (2) patients with malignant cancer at baseline (n = 228); (3) patients with atrial fibrillation (n = 311); (4) patients with estimated glomerular filtration rate (eGFR) < 30 ml/(min×1.73m²) (n = 54); (5) pregnant female or age < 18 (n = 0); (6) patients with missing laboratory data in PA calculation algorithms and PAA (n = 1519). After applying all inclusion and exclusion criteria (detailed in Figure S1), a final cohort of 8,672 SCAD patients who underwent PCI were enrolled in the analysis. The studies involving human participants were reviewed and approved by the Ethics Committee of Sir Run Run Shaw Hospital. The Acceptance Number is 2020-591-03 and the Approval Number is Ethics Committee of Sir Run Run Shaw Hospital 2023 Year No.0644. All methods were performed in accordance with relevant guidelines and regulations. While informed consent was obtained from all subjects and/or their legal guardian(s). Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

2.2. Phenotypic age acceleration

PA is a new developed concept using clinical data from the third National Health and Nutrition Examination Survey (NHANES) [7,9]. By regressing the hazard of mortality on 42 clinical biomarkers along with chronological age, 9 specific clinical biomarkers in addition to chronological age were identified as the most relevant variables. These biomarkers are albumin (Alb), serum creatinine (SCr), glucose (the fasting plasma glucose, FPG), C-reactive protein (CRP), lymphocyte percentage (Lym), mean cell volume (MCV), red blood cell distribution width (RDW), alkaline phosphatase (ALP), and white blood cell count (WBC). The formula of PA is presented as follows:

$$\text{Phenotypic}\text{age}\text{= 141.50 +}\frac{\text{Ln}\text{[}-\text{0.00553 ×}\text{Ln}\text{(}{\text{e}}^{\frac{-\text{1.51714×}{\text{e}}^{\text{xb}}}{\text{0.0076927}}}\text{)}}{\text{0.09165}}$$

where xb = −19.907 − 0.0336 × Alb + 0.0095 × SCr + 0.1953 × FPG + 0.0954 × LnCRP − 0.0120 × Lym + 0.0268 × MCV + 0.3306 × RDW + 0.00188 × ALP + 0.0554 × WBC + 0.0804 × Chronological Age.

$$\text{Phenotypic Age Acceleration = Phenotypic Age - Chronological Age}$$

In the present study, participants with PAA ≥ 0 were categorized as phenotypically older, whereas those with PAA < 0 were categorized as phenotypically younger. The detailed description of the PAA has been documented elsewhere [7].

2.3. Major adverse cardiovascular events

During a median follow-up of 2.2 [1.4, 3.1] years, the occurrence of MACEs was collected through a combination of telephone interviews and a review of hospital information systems. MACEs were defined as a composite of death (both all-cause mortality and cardiac mortality), non-fatal MI, non-fatal stroke and unplanned revascularization,. As for the concept of MI, stroke and unplanned revascularization the diagnosis of nonfatal MI required appropriate symptoms, relevant biomarkers elevation, and/or electrocardiographic changes [20]. The diagnosis of stroke was conducted in accordance with current guidelines [21]. Unplanned revascularization was defined as the first occurrence of subsequent revascularization involving either the target or non-target vessel following the index procedure performed at study enrollment [22].

2.4. Covariates

Smoking and drinking status were stratified into 3 levels: none (never done it), quit (quit for at least 3 months), and current [23]. BMI was calculated as weight in kilograms divided by height in meters squared and categorized into 3 groups: < 25, ≥ 25 to < 30, ≥ 30 kg/m² according to World Health Organization. Low-density lipoprotein cholesterol (LDL-C) was grouped into 3 groups: < 1.8, ≥ 1.8 to < 3.4 and ≥ 3.4 mmol/L refer to Chinese guidelines for lipid management (2023) [24]. Hypertension (HTN) was defined as a blood pressure of ≥ 140/ 90 mmHg on a time-varying basis or hypertension medication drug use. For Asian patients with prediabetes and SCAD, glycated hemoglobin A1c (HbA1c) level could be a valuable marker for the prognosis of cardiovascular events [25]. Diabetes mellitus (DM) was defined as, use of insulin or oral hypoglycemic medication, FPG ≥ 7.0 mmol/L (126 mg/dL), or HbA1c ≥ 6.5% (48 mmol/mol), or 2-h plasma glucose ≥ 200 mg/dL (11.1 mmol/L) during oral glucose tolerance test, or with classic symptoms of hyperglycemia/ hyperglycemic crisis while a random plasma glucose ≥ 200 mg/dL [26,27]. HTN treat included medication use like angiotensin converting enzyme inhibitor (ACEI)/ angiotensin receptor antagonist (ARB), calcium channel blocker (CCB) and β-blocker combined with uncontrolled blood pressure or previous HTN diagnosis. DM treat included medication use like insulin and oral hypoglycemic agents. N-terminal pro-brain natriuretic peptide (NT-proBNP) abnormal group was stratified according to age (eg. NT-proBNP level > 450 ng/L in patients under 50 years old, > 900 ng/L in patients over 50 years old, > 1800 ng/L in patients over 75 years old). Heart failure was defined as typical symptoms, abnormal NT-proBNP group, ejection fraction (EF) < 40%, EF ≥ 40% and raised plasma NT-proBNP [19,28].

2.5. Statistical analysis

Baseline characteristics of the study participants were reported using different statistical tests based on the nature of the variables. Normally distributed continuous variables were presented as mean (standard deviation) while non-normally distributed continuous variables were presented as median (interquartile range) with comparisons by the Kruskal-Wallis test, and comparisons between groups were analyzed using independent samples T-test. The categorical variable was presented as count (percentage) with comparisons by the chi-square test.

The correlation between PAA and the 10 components (age, Alb, SCr, FPG, CRP, Lym, MCV, RDW, ALP, WBC) were explored by the "corrplot" R package, using Spearman correlation analysis.

The Kaplan Meier method was used to evaluate the different probability of MACEs and secondary endpoints between phenotypically younger group and phenotypically older group.

Cox proportional hazard regression models were applied to calculate the hazard ratios (HRs) and 95% confidence intervals (CIs) for MACEs and other endpoints. Three statistical models were fitted to adjust for potential confounders of incident MACEs. Model 1 was unadjusted. Model 2 was adjusted for sex (male, female), age (years), BMI group (< 25, ≥ 25–30, or ≥ 30 kg/m²), smoking status (none, quit, or current), drinking status (none, quit, or current). Model 3 was further adjusted for LDL-C group (< 1.8, ≥ 1.8 to < 3.4, or ≥ 3.4 mmol/L), HTN, treatment for HTN (CCB, ACEI/ARB, or β-blocker), DM, treatment for DM (insulin or oral hypoglycemic agents).

A dose-response association of MACEs risk and continuous PAA was visualized by the restricted cubic spline (RCS) model with 3 knots (5th, 50th and 95th percentiles), which was adjusted in Model 3. Nonlinearity was assessed with likelihood ratio tests comparing models with and without the cubic spline term. Nonlinear P-values and the figures were got from “plotRCS” R package.

Subgroup analysis was conducted when patients were stratified by age (< 65, ≥ 65 years), sex (male, female), smoking status (none, quit, or current), drinking status (none, quit, or current), BMI (< 25, ≥ 25 to < 30, ≥ 30 kg/m²), LDL-C group (< 1.8, ≥ 1.8 to < 3.4, or ≥ 3.4 mmol/L), HTN, HTN treat, DM and DM treat and NT-proBNP abnormal.

The P values for the product terms between continuous PAA and stratification factors were used to test the significance of interactions. HR (95% CI) and P for interaction were estimated by “jstable” R package.

To test the robustness, we also conducted several sensitivity analyses. First, interaction between diabetes treatment and PAA was identified in subgroup analysis, so the DM population was singled out to analyze the association between PAA and MACEs. Second, further adjusted for heart and kidney function (EF, cTnI, NT-proBNP, and eGFR). Third, a sharp increase in endpoints due to a unified review around 1-year after PCI, landmark analysis based on 1 year as the event node was used to eliminate this effect. Forth, given that spline regression models can be sensitive to the selected knots, RCS with 4 knots at the 5th, 35th, 65th, and 95th percentiles of continuous PAA was employed. Fifth, excluded the possible heart failure patients.

Two-sided P values < 0.05 were considered statistical significance. Data were analyzed by R software (R version 4.3.1).

3. Results

3.1. Baseline characteristics of the participants

The characteristics of 8,672 patients with SCAD underwent PCI were summarized in Table 1 (male, 69.80%; age, 64.7 ± 10.3 years; PA, 62.7 ± 12.3 years). The median (interquartile range) concentration of PAA was −2.8 (−6.4, 1.5) years. Compared with phenotypically younger group (PAA < 0), phenotypically older group (PAA ≥ 0) was prone to be older (age, 64.2 ± 9.9 vs. 65.8 ± 10.3; PA, 58.8 ± 10.5 vs. 70.6 ± 11.8). Phenotypically older group had higher BMI, 24.2 (22.4, 26.3) vs.24.6 (22.6, 26.9) and was more likely to be male (66.50% vs. 76.60%), smoker (never smoke, 64.60% vs. 57.60%) and drinker (never drink, 73.40% vs. 71.10%). Also participants in the phenotypically older group were more susceptible to diabetes (59.30% vs. 71.00%) and hypertension (64.20% vs. 71.00%). All above P values <0.001*.

Table 1.

Baseline characteristics.

Variables	Overall	Phenotypically younger	Phenotypically older	P value
Follow-up, y	2.2 [1.4, 3.1]	2.2 [1.4, 3.1]	2.5 [1.6, 3.3]	<0.001*
Age, y	64.7 (10.3)	64.2 (9.9)	65.8 (10.8)	<0.001*
Phenotypic age, y	62.7 (12.3)	58.8 (10.5)	70.6 (11.8)	<0.001*
PAA, y	−2.8 [−6.4, 1.5]	−5.1 [−7.7, −2.7]	3.7 [1.6, 7.2]	<0.001*
Male, n (%)	6053 (69.8)	3878 (66.5)	2175 (76.6)	<0.001*
BMI, kg/m2	24.4 [22.4, 26.5]	24.2 [22.4, 26.3]	24.6 [22.6, 26.9]	<0.001*
Smoking status, n (%)				<0.001*
Current	1767 (20.4)	1118 (19.2)	649 (22.9)
None	5406 (62.3)	3770 (64.6)	1636 (57.6)
Quit	1499 (17.3)	944 (16.2)	555 (19.5)
Drinking status, n (%)				0.001*
Current,	1570 (18.1)	1057 (18.1)	513 (18.1)
None	6298 (72.6)	4280 (73.4)	2018 (71.1)
Quit	804 (9.3)	495 (8.5)	309 (10.9)
PA algorithm's biomarkers
Alb, g/L	42 (4)	43 (4)	41 (4)	<0.001*
FPG, mmol/L	6.2 (1.8)	5.6 (1.0)	7.5 (2.3)	<0.001*
Lym, %	26 (8)	27 (7)	23 (7)	<0.001*
SCr, μmol/L	77 (21)	73 (15)	87 (27)	<0.001*
CRP, mg/dl	0.1 [0.1, 0.3]	0.1 [0.1, 0.2]	0.2 [0.1, 0.5]	<0.001*
RDW, %	13.2 (0.8)	13.0 (0.7)	13.7 (0.9)	<0.001*
WBC, 1000 cells/μL	6.2 (1.6)	6.0 (1.5)	6.8 (1.7)	<0.001*
ALP, U/L	84 (23)	83 (23)	87 (24)	<0.001*
MCV, fl	92 (4)	92 (4)	92 (4)	0.931
Other biomarkers
Hb, g/L	134 (15)	136 (14)	132 (16)	<0.001*
LDL-C, mmol/L	2.1 [1.6, 2.8]	2.1 [1.6, 2.8]	2.0 [1.5, 2.7]	<0.001*
EF, %	67 (15)	68 (17)	64 (10)	<0.001*
NT-proBNP, ng/L	98 [44, 299]	80 [40, 213]	171 [58, 619]	<0.001*
cTnI, ng/L	0.01 [0.01, 0.01]	0.01 [0.01, 0.01]	0.01 [0.01, 0.01]	<0.001*
HbA1c, %	6.3 (1.1)	6.0 (0.9)	6.7 (1.4)	<0.001*
eGFR, ml/min×1.73 m2	87 (19)	90 (17)	80 (21)	<0.001*
Disease and treatment, n (%)
DM,	5474 (63.1)	3457 (59.3)	2017 (71.0)	<0.001*
DM treat	2547 (29.9)	1396 (24.4)	1151 (41.2)	<0.001*
Insulin	1316 (15.2)	732 (12.6)	584 (20.6)	<0.001*
OralDM	2498 (28.8)	1369 (23.5)	1129 (39.8)	<0.001*
HTN	5662 (66.5)	3677 (64.2)	1985 (71.0)	<0.001*
HTN treat	3947 (45.5)	2536 (43.5)	1411 (49.7)	<0.001*
β-blocker	1250 (14.7)	782 (13.7)	468 (16.8)	<0.001*
ACEI/ARB	1702 (20.0)	1055 (18.4)	647 (23.2)	<0.001*
CCB	2445 (28.7)	1627 (28.4)	818 (29.3)	0.091

Overall, n = 8672; Phenotypically younger group, n = 5832; Phenotypically older group, n = 2840.

Categorical data are presented as n (%) and continuous data are expressed as mean (standard deviation) or median [interquartile range].

Y, years; PA, phenotypic age; PAA, phenotypic aging acceleration; BMI, body mass index; Alb, Albumin; FPG, the fasting plasma glucose; Lym, Lymphocyte percentage; SCr, serum creatinine; CRP, C-reactive protein; RDW, Red Cell Distribution Width; WBC, White Blood Cell Count; ALP, Alkaline Phosphatase; MCV, Mean Corpuscular Volume; Hb, Hemoglobin; LDL-C, Low-Density Lipoprotein Cholesterol; EF, ejection fraction; NT-proBNP, N-terminal of the prohormone brain natriuretic peptide; cTnI, Cardiac Troponin I; HbA1c, Glycated Hemoglobin A1c; eGFR, estimated glomerular filtration rate; DM, diabetes mellitus; OralDM, oral diabetes mellitus drugs; HTN, hypertension; ACEI, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor antagonist; CCB, calcium channel blocker.

^*P value <0.05.

3.2. PAA and components

For PAA and the components in PAA calculation algorithms (age, Alb, SCr, FPG, CRP, Lym, MCV, RDW, ALP, WBC), no correlation existed except for Lym and PAA (a weak negative correlation, coefficient = −0.32; P = 0.030*) (Fig. 1) [29].

Fig. 1.

Correlation analysis about PAA and the biomarkers of algorithm.

Lymphocyte percentage showed weak negative association about PAA with a Spearman correlation coefficient of −0.32, *P value < 0.05.

PAA, phenotypic aging acceleration; CA, chronological age; Alb, Albumin; FPG, the fasting plasma glucose; Lym, Lymphocyte percentage; SCr, serum creatinine; CRP, C-reactive protein; RDW, Red Cell Distribution Width; WBC, White Blood Cell Count; ALP, Alkaline Phosphatase; MCV, Mean Corpuscular Volume

3.3. Phenotypically younger/older group and MACEs

In SCAD patients, the phenotypically older group was more likely to have MACEs (19.33% vs. 13.35%), death (all-cause mortality, 3.61% vs. 1.14%; cardiac mortality, 1.11% vs. 0.28%), revascularization (13.55% vs. 11.02%), non-fatal MI (0.64 % vs. 0.43%) and non-fatal stroke (1.28% vs. 0.76%) compared to phenotypically younger group (Table S1). Kaplan–Meier curves showed higher probability of MACEs, death, revascularization and non-fatal stroke (All log rank P values < 0.05) in phenotypically older group (Fig. 2).

Fig. 2.

Kaplan–Meier survival curve and risk table for phenotypically younger group and phenotypically older group.

MACEs, (B) revascularization, (C) all-cause mortality, (D) cardiac mortality, (E) non-fatal MI and (F) non-fatal stroke.

Multivariate adjustment cox proportional hazard models were conducted to better support the association of MACEs and phenotypically younger/older group (Table S3). Phenotypically older group was significantly associated with higher possibility of MACEs. Model 3 demonstrated that phenotypically older group had higher risk of MACEs (HR: 1.354; 95% CI: 1.198,1.538), all-cause mortality (HR: 2.467; 95% CI: 1.793, 3.419), cardiac mortality (HR: 3.187; 95% CI: 1.708, 5.946), unplanned revascularization (HR: 1.191; 95% CI: 1.037, 1.368) in reference to the phenotypically younger group (All P values < 0.05).

3.4. Continuous PAA and MACEs

During a median follow-up of 2.23 years, 1147 MACEs were documented, including 165 all-cause mortality (47 cardiac mortality), 918 unplanned revascularization, 43 non-fatal MI and 80 non-fatal stroke. In multivariate adjustment cox proportional hazard models, PAA was significantly associated with higher possibility of MACEs and secondary endpoints (Table 2). In Model 3, for each 1-year PAA faster, there were 2.80%, 8.80%, 10.8%, 1.20%, 5.80%, 4.50% higher risk for MACEs, all-cause mortality, cardiac mortality, revascularization, non-fatal MI and non-fatal stroke, respectively (All P values < 0.05).

Table 2.

HRs (95% CIs) for MACEs and continuous PAA (n = 8672).

Characteristic	n (%)	Model 1			Model 2			Model 3
		HR	95%CI	P	HR	95%CI	P	HR	95%CI	P
MACEs	1147 (15.20)	1.027	(1.018,1.037)	<0.001*	1.025	(1.015,1.035)	<0.001*	1.028	(1.018,1.038)	<0.001*
All-cause mortality	165 (1.90)	1.098	(1.074,1.124)	<0.001*	1.080	(1.055,1.106)	<0.001*	1.089	(1.063,1.115)	<0.001*
Cardiac mortality	47 (0.50)	1.121	(1.074,1.169)	<0.001*	1.107	(1.061,1.156)	<0.001*	1.109	(1.060,1.159)	<0.001*
Revascularization	918 (11.80)	1.013	(1.002,1.023)	0.022*	1.011	(1.000,1.023)	0.046*	1.012	(1.001,1.024)	0.038*
Non-fatal MI	43 (0.50)	1.048	(1.000,1.099)	0.049*	1.047	(0.998,1.099)	0.063	1.057	(1.005,1.113)	0.032*
Non-fatal stroke	80 (0.90)	1.043	(1.007,1.080)	0.018*	1.037	(1.002,1.074)	0.041*	1.045	(1.008,1.084)	0.017*

Cox proportional hazards models were used to estimate the HRs (95% CIs) for MACEs and continuous PAA.

Model 1 was unadjusted;

Model 2 was adjusted for sex (male, female), age (years), BMI group (< 25, ≥ 25–30, or ≥ 30), smoking status (none, quit, or current), drinking status (none, quit or current);

Model 3 was additionally adjusted model, further adjusted for LDL group (< 1.8, ≥1.8 to < 3.4 or ≥ 3.4), hypertension, treatment for hypertension (CCB, ACEI/ARB, or β-blocker), diabetes, treatment for diabetes (insulin or oral drugs).

MACEs, major adverse cardiovascular events; MI, myocardial infarction.

^*P value < 0.05.

A dose–response association was observed between continuous PAA and MACEs, unplanned revascularization, all-cause mortality, cardiac mortality, non-fatal MI after multivariate adjustment (P for non-linearity of non-fatal stroke = 0.049*, other P values for non-linearity > 0.05) (Fig. 3).

Fig. 3.

Dose-response associations between continuous PAA and MACEs.

Restricted cubic spline regression model adjusted in Model 3 with 3 knots (5th, 50th, and 95th percentiles) was employed to estimate the dose–response relation between PAA and MACEs. (A) MACEs, (B) revascularization, (C) all-cause mortality, (D) cardiac mortality, (E) non-fatal MI and (F) non-fatal stroke.

3.5. Subgroup analysis

Consistent results were observed when stratified by age (< 65, ≥ 65 years), sex (male, female), smoking status (none, quit, or current), drinking status (none, quit, or current), BMI (< 25, ≥ 25 to < 30, ≥ 30 kg/m²), LDL-C group (< 1.8, ≥ 1.8 to < 3.4, or ≥ 3.4 mmol/L), HTN, HTN treat, DM and DM treat and NT-proBNP abnormal (Figs. 4 and 5; Figure S2-S6). Except for DM treat, no significant interactions were detected between PAA and strata variables (all interact P values > 0.05). Thus, SCAD patients with DM were extracted for further explored in sensitivity analysis.

Fig. 4.

Subgroup analysis of continuous PAA and MACEs.

Inter-P, P for interaction, *P value < 0.05.

Fig. 5.

Subgroup analysis of continuous PAA and revascularization.

Inter-P, P for interaction, *P value < 0.05.

3.6. Sensitivity analysis

Sensitivity analysis confirmed the robustness of the results and revealed several new findings. First, stronger associations with MACEs and other endpoints were found in DM population (Table S2, S4, S5). In Model 3, for each 1-year PAA faster, there were 3.10%, 9.00%, 11.00%, 1.50%, 11.20%, 5.80% higher risk of MACEs, all-cause mortality, cardiac mortality, unplanned revascularization, non-fatal MI and non-fatal stroke, respectively. Second, Model 4 which further adjusted for EF, cTnI, NT-proBNP, and eGFR did not change the association between PAA, phenotypically younger/older group and MACEs (Table S6, S7). Third, in landmark analysis, phenotypically older group raised the probability of MACEs, revascularization, all-cause mortality and cardiac mortality significantly compared with phenotypically younger during 1-year longer follow-up period (Figure S7). Forth, RCS with 4 knots also showed dose–response relations between continuous PAA with MACEs, revascularization, all-cause mortality and cardiac mortality (all P for non-linearity > 0.05) (Figure S8). Fifth, exclusion of the possible heart failure patients remained the similar results (Table S8, S9).

4. Discussion

PA is a newly developed concept incorporating 9 blood biomarkers in conjunction with chronological age. As a reflection of biological aging, it provides a more accurate estimation of aging related diseases, such as atherosclerosis, myocardial infarction, cardiac fibrosis and mortality [10,15]. A study estimated the association of PAA and morbidity/mortality risk in NHANES IV showed that each 1-year increase in PAA was associated with a 9% higher risk of mortality. Additionally, PAA exhibited predictive value for disease-specific mortality, including heart disease, cancer, and diabetes, surpassing the predictive power of other models [9]. In a prospective cohort study in the UK Biobank, accelerated biological aging elevated the risk of depression/anxiety, with an approximately 11.3% higher risk per standard deviation of PAA [30,31]. Studies had indicated that unhealthy lifestyles, childhood adversities and life course traumas could accelerate PAA, which mediated traumas and higher risk of cardiovascular diseases [32,33].

However, few studies have specifically evaluated the association between PAA and MACEs in SCAD patients. This cohort study found that phenotypically older group was more likely to experience MACEs after PCI, identifying positive dose–response association between PAA and MACEs. Specifically, each1-year acceleration in PAA was associated with an increased risk of 2.80%, 8.80%, 10.80%, 1.20%, 5.80%, 4.50% higher possibility of MACEs, death, unplanned revascularization, non-fatal MI and non-fatal stroke. Exploratory analysis also showed similar results in subgroups and sensitivity analysis.

Aging of the cardiovascular system involves various interconnected factors. It encompasses not only a decline in the compliance of the heart and large blood vessels, fibrosis and calcification of valvular rings, and reduced responsiveness of cardiomyocytes to stimulation, but also the premature biological aging observed in senescent cells within atherosclerotic plaques [15,34]. The explanation linking accelerated biological aging to a higher risk of MACEs may lie in inflammation and immunity. First of all, biological age is composed of several biomarkers of inflammation, such as CRP, WBC, lymphocyte percentage and RDW. Secondly, research has found that the aging of cardiovascular system is associated with excessive stress and chronic low-grade inflammation superimposed on limited cardiac regeneration capacity [35,36]. Numerous molecules and pathways have been implicated in promoting cellular senescence, thereby contributing to the development of cardiovascular diseases such as hypertension, diabetes, and atherosclerosis. These include sirtuins, Klotho, the renin-angiotensin-aldosterone system (RAAS), insulin-like growth factor binding proteins (IGFBPs), nuclear factor erythroid 2-related factor 2 (NRF2), and the mechanistic target of rapamycin (mTOR) signaling pathway [37].The concept of “inflammaging” highlights the pivotal roles played by aging and inflammation in the development of cardiovascular disease [[38], [39], [40]]. Inflammaging not only interacts with cardiovascular risk factors like obesity and diabetes but also exacerbates their detrimental effects on the cardiovascular system. Moreover, cardiovascular conditions themselves can further intensify inflammation, establishing a vicious cycle between inflammatory aging and cardiovascular disease [41]. Therefore, understanding inflammaging and its underlying contributors or molecular mediators could offer potential targets for innovative therapeutic strategies aimed at promoting healthy aging and conserving healthcare resources worldwide [42].

We found an interaction between DM treatment and PAA leading us to stratify SCAD patients by DM status. In this specific population, PAA exhibited a more significant association with MACEs and secondary endpoints. Several studies have demonstrated that that coronary artery disease in patients with diabetes is characterized by a higher risk of atherosclerosis, macrophage infiltration, and plaque thrombosis, indicating that diabetes itself is a potent risk factor for atherosclerosis [43,44]. In the correlation analysis of PAA and its constituent indicators (Fig. 1), the Spearman correlation coefficient of PAA and FPG was 0.50, P = 0.09. This suggest a moderately correlated trend, although it did not reached statistical significance [29].

Our study possesses several strengths. First, our analysis extends the epidemiological evidence regarding the association between PAA and MACEs in SCAD patients over a long follow-up period. Second, this study benefited from a relatively large cohort size complete follow-up of MACEs events, as well as comprehensive laboratory and demographic data, which collectively allowed us to accurately estimate the association between PAA and MACEs. Third, a multitude of potential confounders were carefully adjusted, including sex, age, smoking status, drinking status, BMI, LDL-C levels, HTN, DM and medication use. Fourth, we stratified a specific population where the impact of PAA on the increased risk of MACEs was more pronounced among SCAD patients with diabetes.

Several limitations should also be considered. First, we did not calculate PAA values in different periods, which may not reflect the long-term status. Second, PAA interacted with DM treatment in this study and therefore, we analyzed patients with diabetes as a distinct population in our sensitivity analysis. Third, our study population consisted of SCAD patients who underwent PCI, the baseline data revealed that there were more individuals with phenotypically younger compared to those with phenotypically older. Therefore, our results may not be generalizable to all populations. Fourth, our study was observational in nature and could only demonstrate association rather than establish causation. Fifth, despite adjustments for major potential known confounding variables in the multivariable Cox regression analysis, we cannot exclude a possible residual bias due to the post-hoc nature of this study, nor can we assess all factors and parameters related to heart diseases. Sixth, as a multicenter study, there might be variations in testing modalities and the technical proficiency of operators, potentially introducing bias or measurement error.

Finally, it is crucial to note that the population of this study primarily consisted of patients with cardiovascular disease who required PCI intervention. Thus, they exhibiteddistinct demographic characteristics. To enhance the generalizability of the results, further studies involving a more diverse populations are warranted.

5. Conclusion

PAA increased the risk of MACEs in SCAD patients, especially all-cause mortality, cardiac mortality and unplanned revascularization. Further studies are warranted to confirm these findings.

CRediT authorship contribution statement

Dan'an Wang: Conceptualization, Methodology, formal analysis, literature screening, Writing- Original draft preparation. Zakareya M. Alsalman: Conceptualization, visualization, Writing- Reviewing and Editing, formal analysis. Yuan Fang: formal analysis, Writing- Reviewing and Editing. Zijie Wang: Formal analysis, Writing- Reviewing and Editing. Duanbin Li: Formal analysis, Writing- Reviewing and Editing. Wenbin Zhang: Methodology, visualization, literature screening, supervision.

Ethics statement

The studies involving human participants were reviewed and approved by Ethics Committee of Sir Run Run Shaw Hospital. The Acceptance Number is 2020-591-03 and the Approval Number is Ethics Committee of Sir Run Run Shaw Hospital 2023 Year No.0644. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

Declaration of Generative AI and AI-assisted technologies in the writing process

Not applicable.

Funding

Wenbin Zhang reports grants from Noncommunicable Chronic Diseases-National Science and Technology Major Project [grant number, 2023ZD0503900, 2023ZD0503904], the National Natural Science Foundation of China [grant number 82070408], the Medical Health Science and Technology Project of Zhejiang Provincial Health Commission [grant number 2021RC014], and the Traditional Chinese Medicine Science and Technology Project of Zhejiang Province [grant number 2021ZB172].

Data statement

The original data that support the findings of this study are available from the corresponding author 3313011@zju.edu.cn upon reasonable request.

Declaration of competing interest

The authors declare that they have no competing interests.

Appendix A. Supplementary data

The following are Supplementary data to this article: