Integrated Chinese and Western Medicine for Coronary Heart Disease Prevention in Polycystic Ovary Syndrome: A 19-Year Nationwide Cohort Study

Introduction

Polycystic ovary syndrome (PCOS) is a prevalent endocrine disorder affecting an estimated 5–18% of women of reproductive age worldwide.¹ It is a multifaceted disorder characterized by a constellation of symptoms, including ovulatory dysfunction, hyperandrogenism, and polycystic ovaries.² However, PCOS transcends the boundaries of a reproductive disorder and has profound implications for metabolic and cardiovascular health.³

Numerous studies have established a substantial link between PCOS and a heightened risk of metabolic syndrome, including insulin resistance, obesity, dyslipidemia, and type 2 diabetes mellitus.^4,5 The combination of these factors places women with PCOS at a significantly increased risk of cardiovascular diseases, and coronary heart disease (CHD) is a critical concern.^6,7 CHD results from the buildup of plaque in the coronary arteries of the heart, leading to a reduction in blood flow to the heart muscle.⁸ The symptoms of CHD include chest pain and shortness of breath and, in more severe cases, can lead to heart attacks, heart failure, and even death, is the leading cause of death globally, accounting for millions of deaths annually.⁹

The association between PCOS and elevated CHD risk has been corroborated by various epidemiological studies.^10–12 Women diagnosed with PCOS are more likely to possess conventional cardiovascular risk factors, such as obesity, hypertension, dyslipidemia, and insulin resistance, leading to an increased risk of CHD development.^13,14 Furthermore, inflammation, a key player in the pathophysiology of atherosclerosis, has been observed to be chronically elevated in patients with PCOS, which could provide an additional link to their elevated CHD risk.^15–17 Despite our growing understanding of the PCOS–CHD connection, the gap in developing effective strategies for the prevention and management of CHD in patients with PCOS remains substantial. The current therapeutic options, such as lifestyle modifications and pharmacological interventions targeting insulin resistance and dyslipidemia, have yielded some benefits but are often hampered by limited adherence, side effects, and inadequate risk reduction.^18–20

Traditional Chinese medicine (TCM), with a rich history spanning several thousand years, offers a wide range of Chinese herbal medicine (CHM) treatments that have been increasingly integrated into modern healthcare. Several integrated CHM products have demonstrated cardiovascular protective effects.^21–23 For example, a study demonstrated that Bushen Jieyu Tiaochong Formula (BJTF) effectively treats PCOS in a rat model. This is achieved by reducing cystic follicle dilation, normalizing hormone levels, and lessening stress-induced behavioral changes. It involves modulating the PERK-ATF4-CHOP signaling pathway and cerebral monoamine neurotransmitters, offering a novel approach for treating PCOS associated with psychological stress.²⁴ Another study revealed that Bu-Shen-Tian-Jing Formula (BSTJF) significantly improved the outcomes of in vitro fertilization (IVF) in Chinese patients with PCOS. This included a higher number of retrieved and fertilized oocytes, increased pregnancy rates, and decreased anti-Müllerian hormone (AMH) levels in follicular fluids, without affecting fertilization rates or increasing complications.²⁵ However, their long-term benefits in reducing CHD risk in patients with PCOS remain largely unexplored.

To investigate the effect of CHM treatment on the long-term risk of CHD in women diagnosed with PCOS, we conducted a retrospective cohort study using data extracted from a comprehensive population-based database in Taiwan. Moreover, we examined the prescription patterns of CHM used in the treatment of PCOS, which may be beneficial for preventing CHD.

Materials and Methods

Data Source

Since 1995, the National Health Insurance (NHI) program has been instrumental in providing healthcare coverage to both Taiwanese and foreign residents. It ensures access to medical services, medications, and treatments. The National Health Insurance Research Database (NHIRD) was established to support this program. It serves as an extensive repository of anonymized health data gathered from the NHI system, which encompasses the entire Taiwanese population. This retrospective cohort study utilized data from the Longitudinal Generation Tracking Database (LGTD) of the NHIRD for the year 2000. The LGTD 2000 contains claims data from a random sample of two million beneficiaries within the NHIRD, spanning from 2000 to 2018. Within the LGTD 2000, diseases were diagnosed using the International Classification of Diseases, 9th Revision, and 10th Revision, Clinical Modification (ICD-9-CM and ICD-10-CM, respectively) codes. Before 2016, diseases were identified using ICD-9-CM codes. However, since 2016, the ICD-10-CM has been used for this purpose. Medications were classified using anatomical therapeutic chemical codes, and CHMs were defined using the database’s unique codes for CHM. Due to privacy concerns for insured individuals, events reported in counts of one or two are not disclosed. This study received approval from the Institutional Review Board of China Medical University (CMUH111-REC2-109(CR-1)). Informed consent was waived, as the study utilizes de-identified data.

Study Populations

We selected female participants who were diagnosed with PCOS (ICD-9-CM:256.4; ICD-10-CM: E28.2) and had a minimum of two ambulatory or inpatient claims from January 1, 2000, to December 31, 2017, in the LGTD 2000. The following exclusion criteria were applied: (1) age < 20 years (n=1875) and (2) a history of CHD (ICD-9-CM: 410–414; ICD-10-CM: I20-I25) before the index date (n=388). Patients were divided into two cohorts during 2000–2017: the CHM cohort (n = 17,285) and the non-CHM cohort (n = 2959). The CHM cohort included patients who received CHM after the initial diagnosis of PCOS. The index date for this cohort was defined as the date when the patients first began receiving CHM. The non-CHM cohort included patients who did not receive CHM after their initial diagnosis of PCOS. The index date for this cohort was selected randomly between 2000 and 2017. Lastly, we employed 1:1 propensity score matching to balance potential confounders between the groups, such as demographic characteristics, comorbidities, and medications.

Covariate Assessment

The demographic factors evaluated included age, urbanization level, and income. Female patients were further divided into three age subgroups as follows: 20–29, 30–39, and ≥40 years. Urbanization was classified into three levels based on the population density in residential areas, with level 1 being the most urbanized and level 3 being the least urbanized. Income was defined by an individual’s monthly earnings and divided into three categories as follows: ≤19,999; 20,000–38,999; and ≥39,000 NT$. Baseline comorbidities were considered if the patients had these conditions before the index date. These comorbidities included diabetes mellitus (ICD-9-CM: 250; ICD-10-CM: E08-E13), hypertension (ICD-9-CM: 401–405; ICD-10-CM: I10-I15), hyperlipidemia (ICD-9-CM: 272; ICD-10-CM: E71.30, E75.21, E75.22, E75.24, E75.3, E75.5, E75.6, E77, E78.0, E78.1, E78.2, E78.3, E78.4, E78.5, E78.6, E78.70, E78.79, E78.8, and E78.9), congestive heart failure (ICD-9-CM: 428; ICD-10-CM: I50), stroke (ICD-9-CM: 430–438; ICD-10-CM: I60-I69), chronic kidney disease (ICD-9-CM: 585, 586, 587, 588, 589; ICD-10-CM: N18, N19, N25, N26, and N27), female infertility (ICD-9-CM: 628; ICD-10-CM: N97), mental disorders (ICD-9-CM: 290–319; ICD-10-CM: F01–F99), and obesity (ICD-9-CM: 278, 783.1; ICD-10-CM: E66.09, E66.1, E66.8, E66.9, E66.01, E66.2, E65, E67.0, E67.1, E67.3, E67.2, E67.8, E68, and R63.5). We also included covariates if the participants were prescribed common medications associated with PCOS during the study period. These medications include metformin, oral contraceptive pills (OCPs), clomiphene, and spironolactone.

Outcome Measures

The follow-up period was extended from the index date to December 31, 2018, the date of the patient’s death, withdrawal from the NHI system, or occurrence of CHD, whichever occurred first. We compared the incidence of CHD between the two groups of patients with PCOS during the follow-up period. In the CHM user group, we further examined the prescribed treatments, including calculating the frequency of single Chinese herbs and Chinese herbal formulas used for PCOS. Additionally, to ascertain the individual effects of specific CHM on CHD risk, we analyzed the association between CHD incidence and the most commonly used single herbs and herbal formulas. To understand the relationships between different single herbs and herbal formulas used for PCOS and to identify a core prescription pattern, we conducted a social network analysis of commonly combined CHM.

Statistical Analyses

Categorical variables are presented as frequencies and percentages, whereas continuous variables are reported as means and standard errors. To test the difference between the means of the two groups, we used t-tests for continuous variables and chi-square tests for categorical variables. Person-years were calculated by adding the follow-up times (in years), and the incidence rates of CHD were computed for each study cohort. Univariate and multivariate Cox models were used to estimate the hazard ratios (HR) and 95% confidence intervals (CI) to investigate the risk of CHD among patients with PCOS. Kaplan–Meier curves were plotted to depict the cumulative incidence of CHD in the two cohorts. The Log rank test was used to compare the survival data of patients with PCOS according to the use of CHM and non-CHM users to determine a significant difference in the survival curves. To understand the prescription pattern of CHM in PCOS, the frequency distribution, pattern of single herbs/herbal formula, and HR associated with the risks of CHM among patients with PCOS were determined. Social network analysis is a methodological approach that examines the connections and interactions between variables within a network. In this study, the network focused on single herbs and herbal formulas of PCOS and aimed to illustrate their relationships. The significance level was set at 0.05 for a two-tailed test. All statistical analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA). Kaplan–Meier curves and social network plots were generated using R studio (3.5.2).

Results

Baseline Characteristics of the Study Population

After applying the inclusion and exclusion criteria, we identified 17,285 CHM and 2959 non-CHM users with PCOS from the LGTD 2000 dataset. Using 1:1 propensity score matching, the final study population was reduced to 2913 individuals in each cohort. A flowchart illustrating this process is presented in Figure 1. Table 1 presents the baseline characteristics of the cohorts. After propensity score matching, a balanced distribution of demographic characteristics and comorbidities was observed between the CHM and non-CHM cohorts. The mean age of the study population was approximately 32 years, with the majority of participants in the 30–39-year age group. The most common baseline comorbidity in both groups was female infertility, followed by mental disorders. More than half of the patients with PCOS used OCPs and clomiphene, with similar usage rates observed between the cohorts.

Table 1 Characteristics of Patients with Polycystic Ovary Syndrome According to Reception of Chinese Herbal Medicine

Figure 1 Flowchart illustrating participant selection and cohort formation process in the study.

CHD Risk and Covariates in Patients with PCOS

Table 2 presents a comprehensive analysis of crude and adjusted HRs for CHD and its covariates in patients with PCOS. Among patients with PCOS, the adjusted Cox model demonstrated a significant overall benefit for CHM users compared with non-CHM users (adjusted HR, 0.06; 95% CI, 0.04–0.1) after adjusting for adjusted for age, insurance amount, urbanization level, all comorbidities, and medication used. This indicates that CHM users had a 94% reduced risk of developing CHM compared to non-CHM users, suggesting that CHM treatment may play a protective role against the development of CHD in patients with PCOS. The risk of CHD was significantly higher in patients aged ≧40 years (adjusted HR, 1.8; 95% CI, 1.25–2.59) than in those aged 20–29 years. Patients with PCOS and baseline comorbidities such as hypertension, hyperlipidemia, stroke, female infertility, and mental disorders experienced a significantly increased risk of CHD (aHR > 1 and p < 0.05). No increase or decrease in the CHD risk in patients with PCOS who used metformin, OCPs, clomiphene, or spironolactone was observed (p > 0.05).

Table 2 Crude and Adjusted Hazard Ratios for Coronary Heart Disease with Chinese Herbal Medicine and Covariates Among Patients with Polycystic Ovary Syndrome by Multivariable Cox Proportional Hazards Regression Analysis

Stratified Analyses of CHM Treatment and CHD Risk in Patients with PCOS

Table 3 compares the incidence rates and HRs of CHD between CHM users and non-CHM users, stratified by demographics and clinical factors. The beneficial impact of CHM treatment on CHD incidence was observed across various subgroups, encompassing all ages, income, and urbanization, as well as patients with or without baseline comorbidities and medication usage, compared with non-CHM users (all aHR < 1 and p < 0.01).

Table 3 Comparison of Incidence Rates and Hazard Ratios for Coronary Heart Disease Between Patients with Polycystic Ovary Syndrome Who Received Chinese Herbal Medicine Treatment and Those Who Did Not

Association Between CHD Incidence and Cumulative Use Days of CHM in Patients with PCOS

We categorized the cumulative use days of CHM treatment into three subgroups:1–29 days, 30–59 days, and ≧60 days and compared them to those of the non-CHM treatment group in terms of the association of the risk of CHD with cumulative use days of CHM during the follow-up period among the patients PCOS. The patients with PCOS who received CHM treatments for cumulative 1–29 days exhibited a reduced risk of CHD (adjusted HR [aHR] = 0.13; 95% CI, 0.06–0.27) compared to the non-CHM users (Table 4). Similar benefits were observed in those who received CHM treatments for cumulative 30–59 days (aHR = 0.14; 95% CI, 0.06–0.35), while the risk of CHD was minimized in patients who underwent CHM treatments for cumulative ≧60 days (aHR = 0.02; 95% CI, 0.01–0.06) compared to the non-CHM users (Table 4).

Table 4 Association Between Coronary Heart Disease Incidence and Cumulative Use Days of Chinese Herbal Medicine in Patients with Polycystic Ovary Syndrome

Cumulative Incidence of CHD in Patients with PCOS

Figure 2 presents the results of the Kaplan–Meier analysis, which demonstrates the cumulative incidence of CHD in patients with PCOS. During the 19-year follow-up period, the CHM cohort demonstrated a significantly lower cumulative incidence of CHD than the non-CHM cohort (Log rank test, p < 0.001). This finding further supports the potential benefits of CHM treatment as a therapeutic intervention to reduce long-term CHD risk in patients with PCOS.

Figure 2 Kaplan–Meier curve demonstrating the cumulative incidence of coronary heart disease in patients with polycystic ovary syndrome in the Chinese herbal medicine (CHM) and non-CHM cohorts.

Prescription Patterns of Single Herbs and Herbal Formulas for PCOS and Their Association with CHD Risk

Table 5 presents the top 10 most frequently prescribed single Chinese herbs and herbal formulas for PCOS, along with information of the number of person-days, average daily dose (g), and average duration for prescription (days). The three most frequently prescribed single Chinese herbs were Xiang Fu (6396 prescriptions), Yi Mu Cao (5878 prescriptions), and Da Huang (5072 prescriptions). Jia Wei Xiao Yao San (8556 prescriptions), Gui Zhi Fu Ling Wan (5075 prescriptions), and Dang Gui Shao Yao San (5069 prescriptions) were the most commonly used Chinese herbal formulas. Moreover, compared to patients who did not receive CHM, patients with PCOS who received these single herbs and herbal formulas had a significantly lower risk of CHD.

Table 5 Top 10 Most Frequently Prescribed Single Herbs and Herbal Formulas for Polycystic Ovary Syndrome, and Their Association with Coronary Heart Disease Risk

Analysis of Common CHM Combinations and Social Correlation Network for PCOS

Table 6 lists the top 10 most frequently prescribed dual combinations of CHM for PCOS. The most common combination was “Xiang Fu plus Yi Mu Cao”, followed by “Nu Zhen Zi plus Han Lian Cao”, “Xiang Fu plus Jia Wei Xiao Yao San”, “Yi Mu Cao plus Jia Wei Xiao Yao San”, and “Yi Mu Cao plus Gui Zhi Fu Ling Wan”. Figure 3 illustrates a social correlation network graph representing the relationships between various Chinese herbs and herbal formulas used to treat PCOS. This graph was created by analyzing the 100 most commonly prescribed CHM combinations using RStudio. In the graph, each node represents a herb or herbal formula, and the edges (connecting lines) denote their co-prescription for treating PCOS. The thickness of each edge reflects the correlation strength or co-prescription frequency between two nodes. Combining this social correlation network with previous results from the most frequently prescribed single Chinese herbs and herbal formulas for reduced CHD risk in patients with PCOS, we identified Xiang Fu, Yi Mu Cao, Jia Wei Xiao Yao San, and Gui Zhi Fu Ling Wan as key elements in CHM prescription patterns (nodes marked in red in Figure 3).

Table 6 Top 10 Most Frequently Prescribed Dual Combinations of Chinese Herbal Medicines for Polycystic Ovary Syndrome

Figure 3 Social correlation network depicting relationships between various single herbs and herbal formulas used in Chinese herbal medicine for polycystic ovary syndrome treatment.

Discussion

This population-based cohort study presents a significant association between CHM use and a long-term 94% reduction in the risk of CHD in patients with PCOS. This is a key finding, as the high prevalence of cardiovascular disease in women with PCOS is a well-documented phenomenon,^26,27 and therapeutic strategies to mitigate this risk are continually being investigated. Using the comprehensive LGTD 2000 dataset, we identified specific patterns of CHM prescription and usage that were associated with significantly reduced CHD risk in patients with PCOS. These insights could not have been achieved through clinical trials owing to their inherent limitations, such as shorter duration, smaller sample size, and the challenge of isolating variables in complex conditions, such as PCOS. Moreover, our longitudinal approach provided a more robust and realistic picture of the effects of CHM over time, particularly in relation to CHD, a disease that typically develops over many years. This further enhances the validity of our findings and their applicability in real-world clinical settings. Furthermore, through the application of big data analytics, we uncovered correlations and generated hypotheses on a scale that is not feasible in traditional clinical trials. For example, our analyses of the most common CHM combinations and their efficacy in reducing CHD risk could pave the way for future investigations into the mechanisms of action of these herbal formulations.

The remarkable number of CHM users identified in this study (>85%) reflects the high prevalence of CHM use among patients with PCOS. This is consistent with previous studies demonstrating the popularity of complementary and alternative medicines, including CHM, in women with gynecological conditions.^28–30 PCOS, a chronic and complex disorder, often requires long-term management, which may explain why many patients seek complementary treatment options such as CHM that emphasize holistic care and fewer side effects. The high usage rate of CHM in our study population could also be considered a reflection of its perceived effectiveness by patients and doctors, which is in line with previous research suggesting positive patient outcomes and satisfaction with CHM treatment for gynecological conditions.^31,32 Moreover, the high utilization of CHM may be bolstered by the integration of TCM into Taiwan’s healthcare system, facilitating easy access to these treatments. Interestingly, the high prevalence of CHM use did not diminish the use of conventional treatments such as OCPs and clomiphene, as more than half of the patients in our study reported using these medications. This suggests that patients seek a complementary approach to managing PCOS by combining the benefits of traditional and modern medicine.

Our results echo the outcomes of numerous clinical trials and observational studies,^33–35 providing additional real-world evidence on CHM therapy as an effective intervention for patients with PCOS. CHM, with its rich repository of bioactive components and multi-targeted action, is often lauded for its potential therapeutic benefits against various diseases, including PCOS and its associated complications.^36–38 The most common single herbs and herbal formulas identified in our study, namely, Xiang Fu (Cyperi Rhizoma), Yi Mu Cao (Leonuri Herba), Jia Wei Xiao Yao San (Augmented Rambling Powder), and Gui Zhi Fu Ling Wan (Cinnamon and Poria Pill), were frequently prescribed to patients with PCOS and correlated with a lower CHD risk. Xiang Fu plays a crucial role as an important herb in Cangfu-Daotan decoction (CFDTT). A previous study has demonstrated that CFDTT treatment in obese PCOS rats led to dose-dependent improvements in lipid profiles, hormone levels, and inflammation markers. Furthermore, CFDTT enhances the expression of OATP2B1 and OATP3A1 in the ovarian and uterine tissues, which may have a positive impact on pregnancy outcomes.³⁹ Another study identified Yi Cao’s potential for treating PCOS by harnessing active substances, such as quercetin and kaempferol, to regulate endocrine disorders and reduce inflammation. Therapeutic effects are achieved by targeting key pathways, including lipid metabolism, atherosclerosis, endocrine resistance, and AGE–RAGE signaling in diabetic complications, as validated by successful molecular docking experiments.⁴⁰ In a pharmacoepidemiological study conducted in Taiwan, Jia Wei Xiao Yao San was identified as the most frequently prescribed Chinese herbal formula among the population.⁴¹ Jia Wei Xiao Yao San is also frequently used in women’s health. For example, one study compared the efficacy of Jia Wei Xiao Yao San with that of hormone replacement therapy, Premelle, for managing menopausal symptoms. Findings revealed that Jia Wei Xiao Yao San had fewer side effects and was as effective as Premelle in alleviating most menopausal symptoms, with its benefits not being mediated by hormone-like effects, thus demonstrating good compliance and safety.⁴² A systematic review and meta-analysis of clinical trials revealed that Xiao Yao San, when combined with conventional medicines, significantly improved ovulation and pregnancy rates and decreased fasting insulin levels and insulin resistance in women with PCOS.⁴³ A previous study has revealed that Gui Zhi Fu Ling Wan alleviates inflammation and improves insulin resistance in rats with PCOS and insulin resistance (PCOS-IR). The underlying mechanism was demonstrated to be through regulating intestinal flora, which in turn controls inflammation, thus supporting the therapeutic use of Gui Zhi Fu Ling Wan for PCOS-IR.⁴⁴ Another study has demonstrated that Gui Zhi Fu Ling Wan inhibits autophagy in granulosa cells, which is a key component of PCOS. This process is mediated by the upregulation of miR-29b-3p and the downregulation of H19, consequently reducing the expression of autophagy-related proteins and the formation of autophagosomes and autophagic polymers.⁴⁵ Furthermore, one study has reported that the Fuling–Guizhi herb pair, a traditional Chinese herbal formula, has the potential to treat CHD. This potential is realized through the regulation of inflammatory responses and the reduction of TNF-α and IL-6 expressions, as identified by a combined approach of network pharmacology prediction and experimental verification.⁴⁶

The significant reduction in CHD risk associated with CHM use in our study suggests that CHM could be considered as a complementary therapeutic option in the clinical management of PCOS, particularly for patients at elevated risk of cardiovascular diseases. However, this study has several limitations. First, we utilized administrative health data from the LGTD 2000 dataset, which may lack information on crucial clinical and lifestyle variables such as body mass index, dietary patterns, and physical activity levels. These factors significantly influence both PCOS and CHD; hence, their absence may have led to residual confounding factors. Second, our study did not consider instances in which patients may have purchased CHM out-of-pocket outside of health insurance coverage. This scenario may have led to significant variations in the dosage, quality, and consistency of treatment among patients, introducing variability that our study was unable to capture. Third, although we applied propensity score matching to address potential confounding factors, there may still be unmeasured confounders that could not be adjusted for in our analysis. Finally, the generalizability of our findings may be limited because of the nature of our study population, which comprised Taiwanese women only. Our findings may not apply to populations with different ethnic backgrounds, healthcare systems, or cultural attitudes toward CHM. Future prospective randomized studies should be conducted to address these limitations and further explore the potential benefits and mechanisms of CHM in reducing CHD risk in patients with PCOS.

Conclusion

This study highlights TCM’s effectiveness in lowering long-term CHD risk in PCOS patients. Key herbs and formulas, notably Xiang Fu, Yi Mu Cao, Jia Wei Xiao Yao San, and Gui Zhi Fu Ling Wan, emerge as beneficial. These results encourage integrating TCM in PCOS treatment, warranting further research to confirm these findings and understand underlying mechanisms.