This systematic review was conducted and reported according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Guideline) statement (15). The protocol was registered in PROSPERO (https://www.crd.york.ac.uk/PROSPERO), and the registration number is CRD42023425791.

Published articles searches were performed by two independent authors (LX and JL) using the following databases: PubMed, Web of science, Cochrane library, China National Knowledge Infrastructure (CNKI), Wanfang Database, and Chinese Scientific Journals Database (VIP), up to 18 December 2022. Search terms such as “coronary heart disease”, “myocardial infarction”, “traditional Chinese medicine”, “lipid”, “LDL-C”, “HDL-C”, “cholesterol”, “triglyceride”, “randomized controlled trial”, and their synonyms were applied in the search strategy. Detailed search strategies can be found in the Supplementary Material S1. The frequency statistics of single CHM was conducted to determine the commonly used drugs.

The main inclusion criteria were as follows: (1) All participants were diagnosed with CHD; (2) Randomized controlled clinical trials included at least 30 patients in each group; (3) Treatments including conventional Western medicine (i.e., antiplatelet, stable plaque, improve ventricular remodeling, etc.) were used in both the control and intervention group, and oral Chinese medicine was administered in the intervention group; and (4) Outcomes of interest involved blood lipid parameters and cardiovascular events.

The exclusion criteria were as follows: (1) Oral Chinese medicine was used in the control group or non-oral Chinese medicine was used in the CHM group; (2) Non-clinical research, review, meta-analysis, and non-core journals; (3) Studies with incomplete information.

The retrieved studies were extracted by two authors (LX and XW) independently following the inclusion and exclusion criteria and included the following details: title, first author, document source, year of publication, sample size, age, gender, disease, interventions, course of treatment, follow-up time, and outcome indicators. Any disagreements were settled through consultation with the third author (BL).

The primary outcomes were blood lipid parameters [LDL-C, TG, total cholesterol (TC), and high-density lipoprotein (HDL)] and major adverse cardiovascular events (MACEs) including cardiovascular mortality, non-fatal myocardial infarction, non-fatal stroke, angina pectoris, heart failure, and revascularization (percutaneous coronary intervention or coronary artery bypass graft surgery). The secondary outcomes were individual components of the primary outcomes.

The methodological quality of the selected studies was assessed by two authors (LXH and JJ) separately according to the seven-item checklist of the Cochrane risk of bias tool (16). The studies that included sufficient domain information were regarded as low risk, while studies with inadequate information were considered as unclear risk. In addition, evaluations were also conducted based on the loss to follow-up and comparability between groups. Furthermore, the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach was used to assess the quality of evidence for primary outcomes (17). We followed the GRADE handbook to conduct the quality assessment, and the GRADEpro Guideline Development Tool (GDT) was used to generate the results (18). Any disagreements were settled through consultation with the third author (JiaL).

Statistical analyses were conducted using Review Manager (RevMan) version 5.4 (The Cochrane Collaboration, 2020). The rate ratios (RR) and weighted mean difference (WMD) were applied to analyze the dichotomous data and continuous data. The heterogeneity across the included studies was measured by I² statistic and Cochran's Q-test. I² > 50% or P < 0.05 was considered as having statistically significant heterogeneity.

Publication bias was detected by funnel plots and Egger's test. To explore the potential source of heterogeneity, the meta-regression analyses, subgroup (Age: <60, ≥60; Statins: used statins, not used statins; course of treatment: <6, 6–12, ≥12 months; type of disease: post-PCI, CCS, ACS) analyses, and sensitivity analysis by omitting one trial at a time were performed as there was significant heterogeneity.

A total of 1,624 records were initially identified, of which 114 were duplicated. Subsequently, 1,182 were excluded after carefully screening titles and abstracts because they were (1) review, meta-analysis, meeting abstract, and irrelevant literatures; (2) CHM in control or non-oral CHM in trial group; and (3) non-clinical study. Further, 305 were excluded after reviewing the full text because: (1) the sample size was less than 30; (2) no lipid parameters and cardiovascular events were recorded; and (3) CHM was used in the control group. Ultimately, 23 studies were identified to be eligible and subsumed in the meta-analysis (Figure 1).

Of the selected studies, 22 were published in Chinese and 1 in English between 2004 and 2022. The baselines of all the studies were comparable. A total of 7,316 participants were included, with 3,670 participants in the CHM group and 3,646 in the control group. The subjects of CHD included stable angina (SA), unstable angina (UA), acute coronary syndrome (ACS), myocardial infarction (MI), non-ST segment elevation myocardial infarction (NSTEMI), ST segment elevation myocardial infarction (STEMI), variant angina pectoris (VAP), post-percutaneous coronary intervention (post-PCI), and borderline lesion CHD. Twenty-two out of the 23 trials were conducted for comparisons of CHM combined conventional treatment therapy vs. a conventional treatment alone (19–40), one trial performed the comparisons of CHM therapy vs. a placebo (41). The CHM in the intervention group was administered orally (decoction, tablets, capsules, granules, or pills). The duration of follow-up varied from 15 days to 4.5 years. The levels of LDL-C were reported in 23 studies, the levels of HDL-C in 17 studies, the levels of TG in 20 studies, and the levels of TC in 20 studies. The overall characteristics are provided in Table 1.

All 23 included studies met the criterion of random allocation, 12 of the included 23 studies reported using a random number table, one applied random numbers generated by a computer, and 10 studies used randomization without providing methodological details. In addition, one study was conducted with blinding of participants and personnel, one study was conducted with blinding of outcome assessment, and four studies did not report the prespecified outcomes—the details are shown in Supplementary Material S2. Furthermore, the results of GRADE analysis demonstrated low confidence in the findings on CHM for lipid levels and CHM for MACEs stratified by lipid levels, the details are shown in Supplementary Material S2.

We also analyzed each component of the MACEs and found that additional CHM treatment was associated with a significant risk reduction in cardiovascular mortality, myocardial infarction, revascularization, angina pectoris, and heart failure, but not with non-fatal stroke (P = 0.05) (Supplementary Material S7). As a meta-regression analysis, the reduction in risk of total MACEs treated by CHM therapy was associated with course of treatment, P = 0.03 for interaction, and might not be attributable to age, statin used, and type of disease (Supplementary Material S8).

Funnel plots and Egger's test were asymmetric for the MD of lipid levels and RR for MACEs, presenting potential publication bias (Supplementary Material S9). A further sensitivity analysis was conducted to show the influence of each trial by deleting each in turn. Deleting each trial successively did not lead to significant deviation from the original overall estimate, indicating that the association of CHM for reducing the risk of MACEs is robust (Supplementary Material S10).

The adverse events during the treatment were reported by three studies. However, there is no significant difference of adverse event rate between the CHM and the Western medicine groups [OR 1.10, 95% CI (0.64–1.89), P = 0.72, I²= 7%] (Supplementary Material S11).

Given the formulae were different due to the type and severity of patients, the frequency statistics of CHM were analyzed to determine the commonly used drugs among different trials. Sichuan lovage rhizome (Chuanxiong, Rhizoma Ligustici Chuanxiong), Danshen root (Danshen, Radix Salviae Miltiorrhizae), Snakegourd fruit (Gualou, Fructus Trichosanthis), Astragalus (Huangqi, Radix Astragali Membranacei), and Peony root (Chishao, Radix Paeoniae Rubra) were the most commonly used in the treatment of CHD (Supplementary Material S12).

In this study, a systematic review and meta-analysis of 23 randomized controlled trials (RCTs) of CHM for regulating lipid levels and lowering cardiovascular events among patients with CHD were conducted, covering 7,316 participants, and 377 cardiovascular events occurred in the 3,670 patients in the CHM group, while 717 events were reported in the 3,646 patients in the Western medicine alone group. First, CHM therapy compared with Western medicine alone was associated with significantly improved lipid profiles.

Second, compared with Western medicine alone, the addition of CHM was associated with a lower risk of cardiovascular events, and the strength of the association varied according to baseline LDL-C and TG levels, and the magnitude in TG and TC reduction. Greater risk reduction in MACEs was observed in patients with baseline LDL-C levels of 2.59–3.34 mmol/L and higher baseline TG level.

Third, CHM therapy was also associated with greater reduction in risk of cardiovascular mortality, myocardial infarction, angina pectoris, revascularization, and heart failure. By contrast, the reduction in non-fatal strokes did not appear to be related to CHM therapy. Considering the probability that the relatively short duration of follow-up is implicated in the lack of MACEs, it could be hypothesized that the course of treatment of CHM and duration of follow-up are too short to fully demonstrate the clinical benefits. It may also be due to the small number of participants. Therefore, further observation is needed for the long-term prognosis.

The development of dyslipidemia, characterized by plasma hypertriglyceridemia, remnant lipoprotein of cholesterol levels (especially elevated LDL-C levels), and decreased HDL-C levels, is a critical driver for atherosclerotic plaques, which further exacerbate risk of MACEs. In a meta-regression analysis, TG lowering is associated with cardiovascular risk reduction, even after adjusting for LDL-C lowering (8). Recent genetic data highlight that interventions of lowering TG levels are also associated with cardiovascular benefits (42). Specifically, it is found that a decrease in 40 mg/dl of LDL-C should lower by approximately 20% the risk of cardiovascular events, but a similar 40 mg/dl reduction in TG levels could generate a 4%–5% reduction of cardiovascular events. Furthermore, the association between TG and MACEs is well demonstrated from epidemiological studies (43, 44) and Mendelian randomization studies (45, 46). Therefore, it will be difficult to achieve a cardiovascular benefit with a purely LDL-C or TG lowering.

Elevated levels of LDL-C and TG are associated with MACEs, all clinicians attest to the fact that intensive LDL-C or TG lowering in cardiovascular disease prevention, although helpful, is difficult to achieve in practice (47). Although the omega-3-fatty acids eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) could lower TG level by 10%–50%, LDL-C may increase with EPA + DHA when TG levels are high (>5,000 mg/dl). In addition, the STRENGTH trial showed that high dose EPA + DHA did not reduce risk of cardiovascular events (48). Therapy purely with fibrates was found to reduce risk of cardiovascular events in patients with high TG and low HDL-C levels. It is currently uncertain whether reduction of cardiovascular events is associated with the addition of fibrates to statin therapy, which may increase the risk of muscle symptoms (49).

Different from Western medicine, characterized by multi-components, multi-targets, and multi-pathways, CHM plays an important role in antiatherosclerosis effects, cardioprotective function, and regulation of lipid levels, which is widely used for CHD in China (50–52). The Huazhuo Yixin Yin is a Chinese medicine decoction, which has been used for treatment of CHD over several years. Zhou et al. (22) reported that Huazhuo Yixin Yin could reduce the risk of MACEs, which was partly the result of its lipid metabolism regulation. Lu et al. (41) administered long-term therapy with Xuezhikang to patients with previous myocardial infarction and found that the incidence of MACEs was significantly decreased and the lipoprotein regulation was improved, and also safe and well tolerated. In these trials, additional CHM administration provided additional reduction in MACEs. Yuan et al. (53) found that the Chuanxiong–Chishao herb pair could improve cardiac function and reduce myocardial fibrosis area by regulating the expression of circRNAs and IncRNAs and its intervention mechanism in CHD may be related to the regulation of lipid metabolism, lipid transport, inflammation, and angiogenesis. Tetramethylpyrazine is an active alkaloid in Ligusticum chuanxiong Hort, which can exert protective effects on myocardial ischemia injury, including scavenging oxygen free radicals (OFRs) to inhibit lipid peroxidation, inhibiting apoptosis and anti-inflammatory activity (54). Ma et al. (55) reported that Salvia miltiorrhiza and Tanshinone IIA could attenuate the buildup of plaque and the accumulation of lipid in ApoE^−/− mice and its mechanism may be related to reducing vascular endothelial inflammation and preventing plaque formation via COX-2/TNF-α/NF-κB signaling pathway. Li et al. (56) summarized that the lipid-regulating effect of CHM in CVD may be related to inhibiting intestinal absorption of lipids, reducing endogenous cholesterol synthesis, regulating transcription factors related to lipid metabolism, regulating cholesterol transport, and promoting the excretion of cholesterol in the liver. These findings suggest that CHM may be used as a complementary approach for patients with CHD. Nevertheless, more rigorously designed RCTs are needed to support these findings.

This study has several limitations. First, the analyses were at the trial level, only the association between risk of cardiovascular events and lipid-lowering treatment by CHM within a trial was considered. Considering the differences in patients, comorbidities (e.g., diabetes), and treatment formulae, subgroup analyses might provide additional explanations on sources of heterogeneity. Second, the lipid levels at the end of the treatment period were used rather than at the end of the follow-up period, which may have overlooked the association of the magnitude of lipid lowering during the follow-up. Third, the formulae were different due to the type and severity of illness of the patients; most prescriptions were modified depending on the individual physique and clinical presentations. Fourth, the included trials lacked strict trial design and adequate sample size estimation. The blinding of participants, personnel, and outcome assessment was performed in 2 of 23 studies.

Given the higher reduction in lipid levels and risk of cardiovascular events, CHM therapy could be considered in patients with high baseline LDL-C and TG levels as it also would offer clinical benefits from lipid lowering. In addition, if additional lipid-lowering therapies or additional treatment to reduce adverse effects are considered for patients treated with Western medicine , CHM therapy to reduce cardiovascular risk is recommended. Although the possible mechanism of CHM in CHD treatment has been discussed, further exploration of the active ingredients and cardioprotective mechanisms of CHM in CHD treatment is of great significance. Therefore, because of the residual cardiovascular risk of Western medicine in the treatment of CHD, more RCTs using multiple centers, larger samples, and strict methodologies are necessary to explore the effectiveness and cardioprotective mechanisms of CHM therapy for CHD.