We performed a retrospective cohort study involving patients diagnosed with CHD who received coronary angiography or coronary computed tomography angiography at the General Hospital of Ningxia Medical University from January 2019 to December 2021. CHD was characterized by the presence of $\ge$50% stenosis in one or more major coronary arteries. The institutional ethics review board approved the study protocol and waived the need for individual informed consent due to the retrospective design.

Inclusion criteria were: (1) patients aged 18 years or older; (2) a confirmed diagnosis of CHD (encompassing stable or unstable angina or myocardial infarction); (3) availability of baseline LDL-C and HCY measurements. Exclusion criteria included: (1) absent or incomplete clinical or laboratory data; (2) severe concomitant illnesses (such as advanced cancer or end-stage renal disease); (3) loss to follow-up within the 25-month study period.

From an initial screening of 6917 patients, 1153 were excluded due to insufficient data, and an additional 627 were lost to follow-up, yielding a final analytical cohort of 5137 individuals (Fig. 1). The study protocol was approved by the General Hospital of Ningxia Medical University ethics committee, with a waiver for patient informed consent.

Fig. 1.

Flowchart illustrating the selection process of study participants. CHD, coronary heart disease.

Participants were stratified by baseline LDL-C and HCY levels using clinically relevant cut-off points. The LDL-C cut-off of 1.8 mmol/L was selected in accordance with the recommended target for secondary prevention in current international guidelines [3, 4]. The HCY cut-off of 15 µmol/L is widely used to define hyperhomocysteinemia in clinical practice and epidemiological studies in our region, and aligns with the reference standard of our institutional laboratory.

Demographic information, lifestyle factors, clinical history (including age, sex, smoking status, alcohol consumption, hypertension, diabetes, and body mass index [BMI]), and laboratory parameters (LDL-C, HCY, total cholesterol [TC], triglycerides [TG], high-density lipoprotein [HDL-C], fasting glucose, creatinine) were obtained from electronic medical records. Blood specimens were drawn from a peripheral vein into Ethylenediaminetetraacetic acid containing tubes within 24 hours of hospital admission and analyzed using a Sysmex automated hematology analyzer (Sysmex Corporation, Kobe, Japan) and a Beckman Coulter series automated biochemistry analyzer (Beckman Coulter, Inc., Brea, CA, USA). Information on interventional procedures, including percutaneous coronary intervention (PCI), was also documented.

The follow-up period extended to 25 months after patient discharge. The primary endpoint was the occurrence of MACEs, a composite of all-cause mortality, stroke, acute myocardial infarction (AMI), or unplanned revascularization. Follow-up data were acquired through review of electronic health records or structured telephone interviews. Standardized follow-up procedures were implemented by trained staff, and any ambiguous events were reviewed by two senior clinicians.

All-cause mortality was defined as death from any cause. Stroke was identified as a disabling neurological deficit attributable to cerebral ischemia or hemorrhage. Non-fatal myocardial infarction was diagnosed based on ischemic symptoms, elevated cardiac biomarker levels, characteristic electrocardiographic changes, or imaging evidence of new myocardial necrosis, in the absence of fatal outcome. Unplanned revascularization referred to any urgent coronary procedure performed in response to an acute coronary syndrome or other acute ischemic event.

Continuous variables are summarized as mean $\pm$ standard deviation (SD) or median [interquartile range (IQR)], and categorical variables as counts and percentages. Group comparisons utilized one-way Analysis of Variance (ANOVA) for normally distributed variables, the Kruskal–Wallis test for non-normally distributed variables, and the chi-square test for categorical variables.

The continuous associations of LDL-C and HCY with MACE risk were visualized using generalized additive models with smoothing splines. Cumulative event incidence was plotted with Kaplan–Meier curves, and group differences were assessed with the log-rank test. Univariate and multivariable Cox proportional hazards models were employed to compute hazard ratios (HRs) and corresponding 95% confidence intervals (CIs). Multivariable models were adjusted for age, sex, hypertension, diabetes, smoking status, alcohol use, TC, TG, BMI, and creatinine levels.

All statistical analyses were conducted using R software (version 4.1.0; R Foundation for Statistical Computing, Vienna, Austria) and EmpowerStats (version 4.0; X&Y Solutions, Inc., Boston, MA, USA), a statistical platform commonly used in biomedical research. A two-sided p-value below 0.05 was deemed statistically significant.

Table 1 displays the baseline characteristics for the 5137 participants, categorized according to their LDL-C and HCY levels. Significant differences were observed across groups regarding demographic, lifestyle, and clinical variables (all p 0.05 unless specified). Subjects in the high HCY categories (low LDL-C & high HCY; high LDL-C & high HCY) were older (mean age 65.04 $\pm$ 10.59, 62.89 $\pm$ 11.50 years, respectively) and more frequently male (77.1%, 73.6%, respectively). The highest rate of smoking was found in the high LDL-C & high HCY group (51.3%). The prevalence of hypertension and diabetes also differed significantly among the subgroups. Relevant laboratory values, including fasting glucose, HCY, LDL-C, and creatinine, also exhibited significant variation (all p 0.001).

Multivariable-adjusted spline curves indicated a positive continuous relationship between LDL-C concentration and the risk of MACEs (Fig. 2A). In the unadjusted model, participants with elevated LDL-C ($\ge$1.8 mmol/L) had a 32% higher risk of MACEs relative to those with lower LDL-C (1.8 mmol/L) (HR = 1.32, 95% CI: 1.12–1.57, p = 0.001). This relationship persisted after multivariable adjustment (adjusted HR = 1.38, 95% CI: 1.09–1.73, p = 0.006; Table 2).

Fig. 2.

Multivariable adjusted spline curves for the associations of (A) LDL-C and (B) HCY with the risk of MACEs. HR, hazard ratio; MACEs, major adverse cardiovascular events.

Fig. 2B illustrates the association between HCY levels and MACEs. Participants with high HCY ($\ge$15 µmol/L) exhibited a significantly increased risk compared to those with low HCY (15 µmol/L). In the unadjusted model, high HCY was associated with a 56% elevation in risk (HR = 1.56, 95% CI: 1.29–1.88, p 0.001). Following adjustment, the association remained significant (adjusted HR = 1.47, 95% CI: 1.19–1.81, p 0.001; Table 2).

Kaplan-Meier curves displayed a graded increase in cumulative MACE incidence across the groups (Fig. 3). The lowest risk was observed in the group with low LDL-C and low HCY, followed successively by high LDL-C & low HCY, low LDL-C & high HCY, and finally the high LDL-C & high HCY group (log-rank p 0.001). In the adjusted Cox regression model, the group with both high LDL-C and high HCY had an almost twofold increased risk of MACEs compared to the reference group (adjusted HR = 1.97, 95% CI: 1.34–2.90, p = 0.001). The groups with high LDL-C & low HCY and low LDL-C & high HCY showed non-significant trends toward increased risk (adjusted HR = 1.35, 95% CI: 0.88–2.06, p = 0.169; and adjusted HR = 1.44, 95% CI: 0.97–2.14, p = 0.067, respectively; Table 2).

Fig. 3.

Kaplan–Meier survival curves for cumulative incidence of MACEs stratified by LDL-C and HCY categories. Groups include: Group1 (low LDL-C & low HCY), Group2 (high LDL-C & low HCY), Group3 (low LDL-C & high HCY), and Group4 (high LDL-C & high HCY). Log-rank test indicated significant differences among groups (p 0.001).

LDL-C is universally acknowledged as a primary pathogenic agent in coronary atherosclerosis, a view substantiated by extensive evidence. For instance, investigations in renal transplant recipients have indicated that a high LDL-C/HDL-C ratio markedly increases cardiovascular morbidity and mortality, emphasizing its utility as a risk indicator [16]. Likewise, in individuals with high-risk hypercholesterolemia, LDL-C levels exhibit a linear correlation with cardiovascular disease incidence, affirming that reducing LDL-C decreases risk [17]. Furthermore, in patients presenting with acute coronary syndrome (ACS), reduced HDL-C levels are independently associated with increased cardiovascular event rates, underscoring the critical influence of lipid metabolism on clinical outcomes [18]. Despite effective LDL-C-lowering therapies, residual risk remains, implicating other factors such as triglycerides [19] and inflammation [20]. Our results reinforce LDL-C as a crucial predictor of cardiovascular risk in CHD patients and further identify elevated HCY as a factor contributing to residual risk.

Elevated plasma HCY is associated with pro-atherogenic and pro-thrombotic mechanisms [21, 22]. A meta-analysis found that each 25% increase in plasma HCY concentration corresponds to a 10% greater risk of cardiovascular disease and a 20% increased risk of stroke [20]. Similarly, an increment of 5 µmol/L in HCY was associated with a 52% elevated risk of new-onset heart disease and a 32% higher mortality risk [23]. Clinically, high levels of HCY levels predict short-term adverse events in acute myocardial infarction patients [24] and are an independent predictor of MACEs in patients with ACS [25]. Moreover, the co-existence of hypertension and hyperhomocysteinemia has been strongly linked to carotid plaque development [26], amplifying overall cardiovascular risk. Although randomized controlled trials on HCY-lowering treatments have not uniformly shown clinical benefits [27], our findings suggest that elevated HCY retains clinical relevance.

The most compelling finding of this study was the combined effect of LDL-C and HCY on MACE risk. Patients with high concentrations of both indices the highest risk of MACEs. LDL-C primarily facilitates lipid accumulation within the arterial wall, advancing atherosclerosis and stenosis [28], whereas HCY aggravates vascular damage through endothelial dysfunction, oxidative stress, and inflammatory processes [26]. This combined effect stems from interconnected pathways where HCY-induced endothelial dysfunction and oxidative stress promote the retention and modification of LDL-C—a pivotal atherogenic step. Conversely, the inflammatory milieu from oxidized LDL exacerbates HCY-mediated metabolic disturbances, creating a vicious cycle that amplifies atherosclerosis.

The inconsistent outcomes of large-scale HCY-lowering trials with B vitamins [12], despite effectively reducing plasma HCY levels [29], may be explained by several factors: intervention timing may be too late in established CHD, HCY may be a marker of underlying pathology rather than a modifiable target, and benefits may be restricted to unselected genetic or nutritional subgroups. Our findings—that risk is greatest with concurrent hyperlipidemia—suggest future trials must account for this synergy. It is critical to emphasize that our results do not contradict the current clinical guidelines; they do not support the routine use of B-vitamin supplementation for CHD patients.

Current clinical guidelines prioritize LDL-C reduction as the foundation of secondary prevention [3, 4], yet residual risk remains substantial. Our study suggests that HCY assessment may help identify high-risk individuals. For patients with both high LDL-C and high HCY, aggressive lipid-lowering (e.g., with Proprotein Convertase Subtilisin/Kexin type 9 inhibitors) combined with HCY-lowering interventions might improve outcomes [30]. A dual-biomarker approach could enable more precise and personalized cardiovascular risk management.

Our findings advocate for a dual-biomarker strategy to identify high-risk CHD patients who may benefit from intensified management. However, optimizing secondary prevention extends beyond risk identification to ensuring long-term adherence to prescribed therapies. The challenge now lies in converting risk identification into sustained adherence. Digital strategies provide a solution: Digital Health Interventions, such as remote monitoring and messaging, boost medication adherence, as evidenced in post-ACS care [31]. Simultaneously, clinicians must steer patients toward trustworthy online health information [32]. Embedding these tools into care pathways is vital for ensuring long-term risk reduction.