Effectiveness and Safety of Chinese Herbal Medicine in Treatment of Perimenopausal Insomnia

Abstract

This study aimed to evaluate the efficacy and safety of Chinese herbal medicine (CHM) for treating perimenopausal insomnia using data from randomized controlled trials (RCTs). A comprehensive search was conducted in PubMed, Web of Science, Embase, Cochrane Library, Chinese Biomedical Literature Service System (SinoMed), Chinese Science Journals Database (VIP), and Wanfang Database through March 2024. Two reviewers independently performed literature screening, data extraction, and risk-of-bias assessment. The meta-analysis incorporated 20 RCTs with a total of 1700 patients. The results demonstrated that CHM, both as monotherapy and in combination with Western medicine (WM), significantly improved sleep quality and clinical total effective rate in patients with perimenopausal insomnia. Specifically, outcomes such as the Pittsburgh Sleep Quality Index (PSQI) scores, total effective rate, and serum follicle-stimulating hormone (FSH) levels favored CHM, either as monotherapy or combined with WM, over WM alone. However, CHM and WM showed no significant differences in PSQI sleep efficiency, PSQI sleep disorders, or serum luteinizing hormone (LH) levels. Moreover, CHM combined with WM did not demonstrate superiority over WM alone in PSQI overall score or serum estradiol (E₂) levels. No severe CHM-related adverse events were reported. These findings suggest that CHM may serve as a safe and effective therapeutic option for managing perimenopausal insomnia. Nonetheless, these conclusions should be interpreted cautiously, and further research with larger sample sizes is needed to validate the efficacy and safety of CHM in this population.

INTRODUCTION

Perimenopausal syndrome is a cluster of physical and psychological symptoms, relatively prominently insomnia, caused by decreased sex hormone production due to declining ovarian function during the menopausal transition. Insomnia is one of the most common symptoms among menopausal women.1 It is typically characterized by difficulty falling asleep, frequent awakenings, and poor sleep maintenance. Tension, anxiety, irritability, palpitations, excessive perspiration, and various other symptoms are commonly associated with insomnia.2 Studies conducted in the United States and other regions have demonstrated significant deterioration in both sleep quantity and quality among perimenopausal women, with sleep disturbances affecting 33% to 51% of this population.3 Long-term insomnia negatively affects patients’ quality of life, compromises immunity, and contributes to metabolic dysfunction, potentially leading to various diseases.4

In traditional Chinese medicine (TCM), insomnia is categorized as “sleeplessness.” Nowadays, non-pharmaceutical methods, Western medicine (WM), and Chinese herbal medicine (CHM) are all used to treat perimenopausal insomnia. Hormone replacement therapy (HRT) is frequently employed in WM to improve sleep quality and alleviate endocrine-related symptoms. However, HRT is associated with an increased risk of cardiovascular disease, endometrial cancer, and ovarian cancer.5 The first-line medications for treating insomnia, according to guidelines, are benzodiazepine receptor agonists and melatonin receptor agonists.6 Although benzodiazepines are effective in alleviating symptoms, they are associated with risks of addiction, withdrawal, and other related disorders. Therefore, exploring alternative therapies, such as TCM, is essential. Chinese medicine treatments have been reported to be promising results in treating perimenopausal insomnia in clinical settings, which may help patients sleep better and have a higher quality of life .7,8

Sleep disorders in TCM are primarily caused by changes and imbalances in the movement of Yang Qi, as well as static changes. Yang Qi enters a state of sleep when it automatically transforms to static, and it enters a state of wakefulness when it transforms from static to dynamic. During menopause, a decline in kidney energy disrupts the balance of Yang and Yin, leading to excessive fire in the heart and liver. This imbalance disturbs the mind and causes difficulty sleeping or restless sleep.9 Furthermore, insomnia is associated with the liver, and menopausal women often experience emotional disorders and depression. Menopausal insomnia is caused by Yin deficiency, Qi deficiency, Blood deficiency, and Blood stagnation, and the pathology is Yin deficiency, Qi deficiency, Blood deficiency, and Blood stagnation.10 While the causes of insomnia are complex and multifaceted, the liver and kidneys are the primary targets for treatment, as they are most affected.11

The studies included in this review exhibited diverse pharmacological effects and complex herbal compositions. CHM compositions contain ingredients such as Dioscorea opposita Thunb. [Dioscoreaceae, Dioscoreae Rhizoma], Scutellaria baicalensis Georgi. [Labiatae, Scutellariae Radix], Ziziphus jujuba Mill. var. spinosa (Bunge) Hu ex H. F. Chou. [Rats and Liidae, Ziziphi Spinosae Semen], Glycyrrhiza uralensis Fisch. [Leguminosae, Glycyrrhizae Radix et Rhizoma], Angelica sinensis (Oliv.) Diels. [Umbelliferae, Angelicae Sinensis Radix], Ligustrum lucidum Ait. [Myrtaceae, Ligustri Lucidi Fructus], Cuscuta australis R. Br. [Convolvulaceae, Cuscutae Semen], Epimedium brevicomu Maxim. [Nambuco, Epimedii Folium] and Curculigo orchioides Gaertn. [Lycopodiaceae, Curculiginis Rhizoma], Taxillus chinensis (DC.) Danser. [Moraceae, Taxilli Herba], which can stimulate hormone secretion, boost the immune system, and have anti-fatigue effects. As an aid, take tranquilizing medicinal substances, such as Cinnabaris. [Sulphide minerals], Polygala tenuifolia Willd. [Yuanzhi Branch, Polygalae Radix], Ostrea gigas Thunberg. [Oysteridae, Ostreae concha], Magnetitum [Magnetite minerals, Magnetitum]. These are capable of treating neurological disorders as well as soothing and protecting the liver.12

This study included 20 articles published between 2010 and the present, with 8 comparing CHM to WM alone and 12 comparing CHM combined with WM to WM alone. A combinatorial analysis was conducted for CHM as both an independent and adjunct intervention. Quantitative indicators, including the Pittsburgh Sleep Quality Index (PSQI) score (overall and component scores), total effective rate, and endocrine hormone levels (E₂, FSH, and LH), were assessed. This study aims to provide evidence-based insights into the clinical application of CHM for treating perimenopausal insomnia.

MATERIALS AND METHODS

Registration

We used the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) Statement to perform a systematic review and meta-analysis.13 This study has been registered with the PROSPERO (International Prospective Register of Systematic Reviews) (CRD42023407255).

Literature search

Two researchers searched electronic databases such as PubMed, Web of Science, Embase, Cochrane Library, Chinese Biomedical Literature Service System (SinoMed), Chinese Science Journals Database (VIP), and Wanfang Database from their conception until March 2024. We included studies that were written in either English or Chinese. A comprehensive search strategy is presented in Table S1, available at: http://links.lww.com/HNP/A28.

Eligibility criteria

Inclusion criteria

The inclusion criteria encompass patients diagnosed with perimenopausal insomnia according to authoritative diagnostic criteria, interventions involving CHM alone or combined with WM, comparison with WM alone, specific outcomes including PSQI scores and other indicators, and a study design limited to randomized clinical trials (RCTs).

1. Patients: patients diagnosed with perimenopausal insomnia using authoritative diagnostic criteria (eg, Guidelines for Clinical Research on New Chinese Medicine [Zheng XY, 2002] or Gynecology of Chinese Medicine [Luo SP, 2015], etc14,15);

2. Interventions: CHM stand-alone or CHM combined with WM;

3. Comparison: use WM alone;

4. Outcomes: the primary outcome was PSQI score (including both the overall and component scores), the secondary outcomes included total effective rate, serum levels of LH, FSH and E₂, and adverse events;

5. Study design: RCTs.

Exclusion criteria

We excluded the studies according to the following criteria:

1. Non-RCT articles, review articles, animal experiments, etc;

2. The research content does not conform to the theme;

3. The subjects had other diseases and medications that might affect the statistical results.

Data extraction quality and validation

The literature review, data extraction, quality evaluation, and cross-checking of the final results were conducted independently by 2 researchers. Discrepancies were resolved through consensus. Data extracted from the original studies included the following: (1) The initial author, the publication year, the nation, the sample size, and the type of clinical study; (2) Subject characteristics such as age, disease duration, and disease diagnosis; (3) information about the intervention, including the substance, dosage, dosing form, and length of the study and control groups’ treatments; (4) information on the outcome, including a definition and specifics; (5) safety: number and description of adverse events.

Each included study was evaluated using the Cochrane Risk of Bias (RoB) tool, which evaluates factors such random sequence generation, allocation concealment, participant and staff blinding, outcome assessment blinding, insufficient outcome data, selective reporting, and other biases. Each item received 1 of 3 ratings: uncertain, low, or high. The graphic illustrating the risk of bias was created using Review Manager (RevMan, version 5.4, Nordic Cochrane Center, Cochrane Collaboration, 2012 Copenhagen, Denmark).

Quantitative data analysis

All the statistical analyses were performed by RevMan 5.4 software. Because there were differences between the papers, we used either the random or fixed effects model for the meta-analysis. For the statistical analysis of counting indicators, we used the risk ratio (RR), and for the statistical analysis of quantitative indicators, we used mean difference (MD). For both kinds of data, the 95% confidence intervals (CIs) were calculated. The Chi-square and I² tests were employed in each meta-analysis to gauge statistical similarity. Applied was a fixed effects model when I² <50% and P >.1. The random effects model was used for research when I² ≥50% and P ≤.1. Heterogeneity refers to the differences that exist between different studies included In meta-analysis. I² values of 25%, 50%, and 75% represented low, moderate, and high heterogeneity. Stata version 15.0 was used to conduct Egger’s tests and sensitivity analysis (STATA Corp., College Station, TX, USA).

RESULTS

Literature search and characteristics of included studies

A total of 4478 studies were retrieved, 2082 of which were duplicated and 2396 were acquired after deletion. After reading the abstracts and the full texts, 2351 were removed. Finally, 20 studies16-35 with publication years ranging from 2010 to 2024 comprised 1700 participants were included in the meta-analysis. Tables 1 and 2 summarize the essential characteristics and composition of CHM formula of the included studies, and the PRISMA flow chart summarizes the study selection in Figure 1.

TABLE 1.

The Characteristics of Studies Included in the Systematic Review and Meta-Analysis

Study ID	Intervention		No. of Patients (T/C)	Age (Year)		Course of Disease (Month)		Duration (Week)	Outcome Measurements
	T	C		T	C	T	C
CHM vs WM
Dong, 202216	Shenci Zaoren Soup	Alprazolam	25/25	49.56 ± 7.22	48.87 ± 7.15	14.12 ± 4.11	13.89 ± 3.99	1	1, 2, 4, 5, 6, 7
Qi, 201917	Zishen Ningshen Soup	Eszopiclone	48/47	49.4 ± 4.1	49.6 ± 4.3	12.12 ± 3.40	11.75 ± 3.20	8	1, 2, 3, 4, 5, 6, 7
Zhang, 201718	Huanglian Ejiao soup and Suanzaoren soup	Eszopiclone	45/41	49 ± 4.3	48 ± 3.1	3-60		4	2
Wang, 201719	Qingxin Zhengan soup	Glutathione + Vitamins	55/55	49.2 ± 7.1	48.6 ± 6.9	1-36	1-38	12	1, 9, 10
Lai, 201720	Guipi soup and Maidong	Alprazolam	30/30	35-60	35-39	6-20	6-21	2	1
Yang, 201621	Baizi Yangxin Jiawei soup	Glutathione	50/50	48.7 ± 2.08	46.3 ± 3.01	10.37 ± 0.8 5	12.63 ± 0.71	3	1
Bai, 201422	Bushen Lemian soup	Alprazolam	45/45	49.8 ± 7.1	49.6 ± 7.2	18.00 ± 3.60	19.20 ± 2.40	4	1, 3, 4, 5, 6, 7
Yang, 201323	Andu Gengnian Pills	Eszopiclone	50/50	45-55	44-55	2-36	3-36	8/4	2
Xu, 201324	Gengnianan Fang	Glutathione	42/41	48.21 ± 1.35	47.64 ± 2.19	NR	NR	2	2
Chen, 201225	Nvzhen Anshen Fang	Scholastin	50/48	47.95 ± 4.57	48.32 ± 3.56	11.24 ± 7.13	10.74 ± 5.9 5	2	1, 4, 5, 6
Zhang, 201226	Chaihu Jia Longgu Muli Soup	Diazepam	35/38	48.7 ± 3.10	48.1 ± 3.12	21.60 ± 18.24	21.60 ± 19.32	4	2
Ju, 201127	Self-prepare d herbal formula	Scholastin	46/46	50.6 ± 6.6	51.4 ± 6.4	12.3 ± 5.20	11.70 ± 4.90	4	2, 11
CHM + WM vs WM
Zhai, 202228	Danzhi Xiaoyao Pills + Dexzopiclone Tablets	Dexzopiclone	48/48	48.72 ± 2.47	48.35 ± 2.24	16.22 ± 3.5	17.39 ± 4.31	8	1, 2, 4, 5, 6
Wang, 202229	Chaihu Jia Longgu Muli Soup + Eszopiclone	Eszopiclone	30/30	49.6 ± 2.10	49.8 ± 2.20	9.20 ± 1.20	9.4 ± 1.30	2	1, 2, 3, 4, 5, 6, 12
Chen, 201830	Gualnlong Mixture +Dexzopiclone Tablets	Dexzopiclone	26/26	49.04 ± 2.91	48.31 ± 2.45	37.68 ± 24. 12	38.04 ± 25. 20	8	2
Li, 201831	Guoju Dihuang Pills + Cloxacin Potassium	Cloxacin Potassium + Eszopiclone	58/58	50.40 ± 3.50	50.10 ± 3.70	6.20 ± 1.60	6.40 ± 1.50	4	1, 2, 12
Xiao, 201732	Shbul gan Jieyu Capsules + Alprazolam tablets	Alprazolam	50/50	50.39 ± 2.11	50. 28 ± 2.06	13.28 ± 4.21	13.41 ± 4.28	4	1, 2, 12
Yuan, 201633	Jiawei Xiaoyao Pills + Eszopiclone Tablets	Eszopiclone	40/36	50.21 ± 0.76		3-24		4	2
Wang, 201134	Xiaoyao Pills + Clonidine	Clonidine	45/42	50.22 ± 2.31	50.37 ± 2.47	3.20	2.80	3	2, 8
Liu, 201035	Jiawei Xiaoyao Pills + Mirtazapine Tablets	Mirtazapine	40/36	50.21 ± 0.76		2-24		3	2

Note: T, trail group; C, control group; CHM, Chinese herbal medicine; WM, Western medicine; NR, not report.

Outcome measurements: 1. Pittsburgh Sleep Quality Index (PSQI) score; 2. Total effective rate; 3. Kupperman Menopausal Index (KMI) score; 4. Serum levels of oestradiol (E₂); 5. Serum levels of follicle-stimulating hormone (FSH); 6. Serum of levels of luteinizing hormone (LH); 7. Syndrome score; 8. Withdrawal success rate; 9. Serum of levels of 5-HT; 10. Serum of levels 5-HIAA; 11. Spiegel Questionnaire score; 12. Adverse events.

TABLE 2.

CHM in the Included Studies

Study ID	CHM Formula	Composition of Formula
Dong, 202216	Shenci Zaoren Soup	Taizishen 15 g, Cilingshi 30 g, Zhimu 15 g, Suanzaoren 15 g, Fuling 15 g, Chuanqiong 6 g, Meigui 10 g, Hehuanpi 15 g, Gancao 3 g. Add and subtract drug ingredients according to the symptom and sign.
Qi, 201917	Zishen Ningshen Soup	Taizishen 20 g, Shudihuang 20 g, Shanyao 15 g, Suanzaoren 12 g, Danggui 12 g, Yuanzhi 12 g, Hehuanhua 12 g, Gouqizi 10 g, Fuxiaomai 8 g, Duzhong 15 g, Fushen 12 g, Ejiao 10 g, Maidong 9 g. Add and subtract drug ingredients according to the symptom and sign.
Zhang, 201718	Huanglian Ejiao soup and Suanzaoren soup	Suanzaoren 30 g, Chuanqiong 10 g, Zhimu 10 g, Fuling 15 g, Huanglian 6-10 g, Huangqin 10 g, Baishao 15 g, Gancao 5 g, Hehuanpi 10 g, Yejiaoteng 15 g, Ejiao 12 g, Jizihuang 1. Add and subtract drug ingredients according to the symptom and sign.
Wang, 201719	Qingxin Zhengan soup	Huanglian 20 g, Huangqin 20 g, Zhizi 20 g, Dandouchi 20 g, Zhenzhumu 20 g, Shengdi 20 g, Shichangpu 20 g, Yuanzhi 20 g, Longgu 20 g, Muli 20 g, Danggui 20 g, Fushen 20 g, Fuling 20 g. Add and subtract drug ingredients according to the symptom and sign.
Lai, 201720	Guipi soup and Maidong	Dangshen 15 g, Dazao 15 g, Fuling 15 g, Yuanzhi 15 g, Suanzaoren 15 g, Muxiang 8 g, Baizhu 12 g, Danggui 10 g, Guiyuan 10 g, Shengjiang 10 g, Huangqi 20 g, Maidong 20 g.
Yang, 201621	Baizi Yangxin Jiawei soup	Baiziren 15 g, Zhiyuanzhi 6 g, Suanzaoren 10 g, Renshen 15 g, Fushen 10 g, Chuanqiong 10 g, Wuweizi 10 g, Duanlonggu 20 g, Duanmuli 20 g, Fuling 15 g, Zhigancao 6 g. Add and subtract drug ingredients according to the symptom and sign.
Bai, 201422	Bushen Lemian soup	Shudi 20, Danpi 20 g, Shanyurou 15 g, Suanzaoren 15 g, Huaishanyao 15 g, Hehuanpi 15 g, Baizhu 15 g, Xianlingpi 10 g, Gouteng 10 g, Xianmao 10 g, Tianma 10 g, Yuanzhi 10 g, Buguzhi 10 g. Add and subtract drug ingredients according to the symptom and sign.
Yang, 201323	Andu Gengnian Pills	Zhishouwu 30 g, Fupenzi 30 g, Tusizi, 15 g, Shanyurou 15 g, Buguzhi 9 g, Danggui 15 g, Fuling 15 g, Huainiuxi 15 g, Gegen 15 g, Gancao 3 g.
Xu, 201324	Gengnianan Fang	Shengdi 12 g, Mohanlian 12 g, Nvzhenzi 12 g, Longchi 30 g, Baishao 12 g, Zhimu 9 g, Wuweizi 6 g, Baihe 9 g, Chaozaoren 15 g, Danshen 12 g. Add and subtract drug ingredients according to the symptom and sign.
Chen, 201225	Nvzhen Anshen Fang	Nvzhenzi 30 g, Hanliancao 15 g, Shenghhuangqi 30 g, Taizishen 15 g, Chaobaizhu 15 g, Danggui15 g, Shengmuli 30 g, Suanzaoren 30 g, Baishao 20 g, Yuanzhi 12 g, Danshen 20 g, Muxiang 6 g. Add and subtract drug ingredients according to the symptom and sign.
Zhang, 201226	Chaihu Jia Longgu Muli Soup	Chaihu 10 g, Huangqin 10 g, Fuling 10 g, Guizhi 5 g, Shenglonggu 30 g, Shengmuli 30 g, Banxia10 g, Cishi 30 g, Shengjiang 5 g, Dazao 5g, Baishao 10 g, Shengdi 15 g, Shouwuteng 30 g, Chaozaoren 15 g. Add and subtract drug ingredients according to the symptom and sign.
Ju, 201127	Self-prepared herbal formula	Chaihu10g, Yujin 15 g, Shenglonggu 30 g, Shengmuli 30 g, Baishao 15 g, Shengdihuang 15 g, Zhimu 15 g, Yinyanghuo 15 g, Suanzaoren 15 g, Yejiaoteng 30 g. Add and subtract drug ingredients according to the symptom and sign.
Zhai, 202228	Danzhi Xiaoyao Pills	Danzhi Xiaoyao Pills (Fuzhou Haiwang Jinxiang Traditional Chinese Medicine Pharmaceutical Co., Ltd, Guo-Yao-Zhun-Zi—Z10930036)
Wang, 202229	Chaihu Jia Longgu Muli Soup	Longgu 30 g, Muli 30 g, Tusizi 20 g, Shanyao 20 g, Suanzaoren 20 g, Chaihu 15 g, Hehuanpi 15 g, Yujin 15 g, Yejiaoteng15g, Guizhi 10 g, Fuling 10 g, Chuanqiong 9 g, Gancao 15 g.
Chen, 201830	Guanlong Mixture	Shendihuang 30 g, Gouqizi 20 g, Danggui 20 g, Maidong 15 g, Beishashen 15 g, Chuanjizi 10 g, Longdan 15 g, Huangqin 10 g, Zhizi 10 g.
Li, 201831	Gouju Dihuang Pills	Gouju Dihuang Pills (Henan Wanxi Pharmaceutical Co., Ltd, Guo-Yao-Zhun-Zi-Z41021905)
Xiao, 201732	Shugan Jieyu Capsules	Shugan Jieyu Capsules (Chengdu Kanghong Pharmaceutical Group Co., Ltd, Guo-Yao-Zhun-Zi—Z20080580)
Yuan, 201633	Jiawei Xiaoyao Pills	Chaihu 15 g, Fushen 15 g, Baishao 15 g, Baizhu 15 g, Zhigancao 15 g, Jiangcan 15 g, Danggui 10 g, Bohe 10 g, Hehuanpi 20 g, Yejiaoteng 20 g,Shengmuli 30 g, Shenglonggu 30 g.
Wang, 201134	Xiaoyao Pills	Chaihu 15 g, Danggui 10 g, Bohe 10 g, Baishao 15 g, Baizhu 15 g, Zhigancao 15 g, Shengjiang 10 g.
Liu, 201035	Jiawei Xiaoyao Pills	Chaihu 15 g, Fushen 15 g, Baishao 15 g, Baizhu 15 g, Zhigancao 15 g, Jiangcan 15 g, Danggui 10 g, Bohe 10 g, Hehuanpi 20 g, Yejiaoteng 20 g,Shengmuli 30 g, Shenglonggu 30 g.

Figure 1.

Flow diagram of the study selection process.

Risk of bias

The Cochrane RoB tool was used to assess the included studies. All of the research included studies reported “randomization,” 5 studies refer to “random number table method.” None of the participants mentioned whether or not they utilized allocation concealment or blinding. Two studies16,23 lacked information on diagnostic criteria and Yang’s21 total efficiency data reported with errors, thus putting them at risk of bias of reporting bias. Figures 2 and 3 show detailed information on each of the included study with RoB assessment.

Figure 2.

Risk of bias graph.

Figure 3.

Risk of bias summary.

Meta-analysis

PSQI score

In clinical practice, the PSQI score is commonly used to assess the sleep status of patients with insomnia. It evaluates the time it takes to fall asleep, the length of sleep, the quality of sleep, sleep disorder and daytime dysfunction to provide a comprehensive assessment of sleep conditions. The higher the score, the worse the sleep conditions.

1. Overall score: A total of 7 RCTs16,17,19,20,22,23,25 (CHM compared with WM) reported outcome measures for PSQI score (603 participants), and fixed effects model meta-analysis indicated that CHM was associated with lower PSQI score than that of WM alone (MD: −2.22, 95%CI: −2.65 to −1.79) (Figure S1-A, available at: http://links.lww.com/HNP/A25). There are 3 RCTs29,30,32 (CHM + WM compared with WM) reported PSQI score (212 participants), and random effects model meta-analysis showed that CHM combined with WM was comparable to that of WM alone (MD: −1.55, 95%CI: −6.45 to 3.35) (Figure S1-B, available at: http://links.lww.com/HNP/A25).

2. PSQI score—the time to fall asleep: A total of 3 RCTs20-22 (CHM compared with WM) reported outcome measures for PSQI score—time to fall asleep (250 participants), and random effects model meta-analysis indicated that CHM was associated with lower PSQI score—time to fall asleep than that of WM alone (MD: −0.69, 95%CI: −1.01 to −0.38) (Figure S2-A, available at: http://links.lww.com/HNP/A25). There are 4 RCTs28,30-32 (CHM + WM compared with WM) reported PSQI score—the time to fall asleep (364 participants), and random effects model meta-analysis showed that CHM combined with WM was associated with lower scores in combination with WM alone (MD: −0.42, 95%CI: −0.67 to −0.17) (Figure S2-B, available at: http://links.lww.com/HNP/A25).

3. PSQI score—the length of sleep: A total of 3 RCTs20-22 (CHM compared with WM) reported outcome measures for PSQI score—the length of sleep (250 participants), and random effects model meta-analysis indicated that CHM was comparable to that of WM alone that of (MD: −0.15, 95%CI: −0.81 to 0.52) (Figure S3-A, available at: http://links.lww.com/HNP/A25). There are 4 RCTs28,30-32 (CHM + WM compared with WM) reported PSQI score—the length of sleep (364 participants), and random effects model meta-analysis showed that CHM combined with WM was associated with lower scores in combination with WM alone (MD: −0.50, 95%CI: −0.69 to −0.31) (Figure S3-B, available at: http://links.lww.com/HNP/A25).

4. PSQI score—the quality of sleep: A total of 3 RCTs20-22 (CHM compared with WM) reported outcome measures for PSQI score—the quality of sleep (250 participants), and random effects model meta-analysis indicated that CHM was associated with lower PSQI score—the quality of sleep than that of WM alone (MD: −0.70, 95%CI: −1.27 to −0.13) (Figure S4-A, available at: http://links.lww.com/HNP/A25).

There are 3 RCTs28,30,32 (CHM + WM compared with WM) reported PSQI score—the quality of sleep (248 participants), and random effects model meta-analysis showed that CHM combined with WM was associated with lower scores in combination with WM alone (MD: −0.55, 95%CI: −0.84 to −0.25) (Figure S4-B, available at: http://links.lww.com/HNP/A25).

5. PSQI score—the efficiency of sleep: A total of 3 RCTs20-22 (CHM compared with WM) reported outcome measures for PSQI score—the efficiency of sleep (250 participants), and random effects model meta-analysis showed that CHM was comparable to that of WM alone (MD: 0.06, 95%CI: −0.70 to 0.83) (Figure S5-A, available at: http://links.lww.com/HNP/A25). There are 4 RCTs28,30-32 (CHM + WM compared with WM) reported PSQI score—the efficiency of sleep (364 participants), and random effects model meta-analysis showed that CHM combined with WM was associated with lower scores in combination with WM alone (MD: −0.45, 95%CI: −0.70 to −0.21) (Figure S5-B, available at: http://links.lww.com/HNP/A25).

6. PSQI score—sleep disorder: A total of 3 RCTs20-22 (CHM compared with WM) reported outcome measures for PSQI score—sleep disorder (250 participants), and random effects model meta-analysis showed that CHM was comparable to that of WM alone (MD: −0.15, 95%CI: −0.30 to 0.01) (Figure S6-A, available at: http://links.lww.com/HNP/A25). There are 4 RCTs28,30-32 (CHM + WM compared with WM) reported PSQI score—the efficiency of sleep (364 participants), and random effects model meta-analysis showed that CHM combined with WM was associated with lower scores in combination with WM alone (MD: −0.46, 95%CI: −0.69 to −0.23) (Figure S6-B, available at: http://links.lww.com/HNP/A25).

7. PSQI score—daytime dysfunction: A total of 3 RCTs20-22 (CHM compared with WM) reported outcome measures for PSQI score—daytime dysfunction (250 participants), and fixed effects model meta-analysis indicated that CHM was associated with lower PSQI score—daytime dysfunction than that of WM alone (MD: −0.36, 95%CI: −0.51 to −0.21) (Figure S7-A, available at: http://links.lww.com/HNP/A25). There are 4 RCTs28,30-32 (CHM + WM compared with WM) reported PSQI score—the efficiency of sleep (364 participants), and fixed effects model meta-analysis showed that CHM combined with WM was associated with lower scores in combination with WM alone (MD: −0.44, 95%CI: −0.56 to −0.32) (Figure S7-B, available at: http://links.lww.com/HNP/A25).

The total effective rate

A total of 10 RCTs16,18,20-27 (CHM compared with WM) reported the total effective rate (846 participants), and fixed model meta-analysis indicated that the total effective rate of CHM (90.19%) was greater than that of WM alone (69.14%) with statistically significant difference (RR: 1.30, 95%CI: 1.21 to 1.40) (Figure S8-A, available at: http://links.lww.com/HNP/A26). There were 6 RCTs28,29,31-33,35 (CHM + WM compared with WM) reported the total effective rate (524 participants), and fixed effects model meta-analysis showed that the total effective rate of CHM combined with WM (93.61%) was higher than the WM group (74.42%), with statistically significant difference (RR: 1.26, 95%CI: 1.17 to 1.36) (Figure S8-B, available at: http://links.lww.com/HNP/A26).

Serum levels of LHFSH and E₂

In 3 RCTs16,17,22 (CHM compared with WM) (235 participants), serum levels of LH, FSH and E₂ were used to assess hormone levels in menopausal women. Due to the existence of inconsistent units of LH and FSH, the standard mean difference (SMD) is used. Random effects model meta-analysis indicated that CHM group was comparable to that of WM alone (SMD: −1.86, 95%CI: −3.82 to 0.11) (Figure S9-A, available at: http://links.lww.com/HNP/A27). Random effects model meta-analysis showed that CHM group was associated with lower FSH level than that of WM alone (SMD: −1.13, 95%CI: −1.78 to −0.48) (Figure S10-A, available at: http://links.lww.com/HNP/A27). In terms of raising E₂, fixed effects model meta-analysis indicated that CHM group was associated with higher level than WM alone (MD: 11.72, 95%CI: 9.83 to 13.62) (Figure S11-A, available at: http://links.lww.com/HNP/A27).

In 3 RCTs28-30 (CHM + WM compared with WM) (208 participants), serum levels of LH, FSH and E₂ were used to assess hormone levels in menopausal women. Fixed effects model meta-analysis indicated that the CHM combined with WM group was associated with lower LH level and lower FSH level (SMD: −0.75, 95%CI: −1.04 to −0.47) (SMD: −0.59, 95%CI: −0.87 to −0.31) (Figure S9-B, available at: http://links.lww.com/HNP/A27; Figure S10-B, available at: http://links.lww.com/HNP/A27). As to serum level of E_2, random effects model meta-analysis indicated that CHM combined with WM was comparable to that of WM alone (MD: 12.51, 95%CI: −1.11 to 26.12) (Figure S11-B, available at: http://links.lww.com/HNP/A27).

Safety assessment

According to Table 3, 3 out of a total of 12 trials comparing CHM to WM and 3 out of a total of 8 trials comparing CHM combined with WM to WM reported adverse events. The adverse reactions were more serious and happened more frequently in the WM group than in the CHM group. Most of the unfavorable side effects, including headache, dizziness, blurred vision, and dry mouth, were experienced by the control group, which got treatment with WM. However, both the CHM combined with WM and the WM groups were at risk for symptoms like weakness, headache, skin redness and rash, dizziness, and gastrointestinal reactions, though the frequency was greater in the WM group. Overall, compared to the CHM stand-alone group, the combined CHM and WM group was more likely to suffer additional side effects.

TABLE 3.

The Adverse Reactions in the Included Studies

Study ID	Intervention vs Control	Adverse Reactions
		Trail Group	Control Group
Qi,201917	CHM vs WM	None	2 cases of dry mouth, 2 cases of stomach discomfort and 1 case of dizziness and headache, with mild symptoms.
Yang,201621	CHM vs WM	1 case of dyscalculia, 1 case of loose stool.	4 cases of night fatigue, 2 cases of weakness, 2 cases of drowsiness and 1 case of dizziness and headache.
Zhang,201226	CHM vs WM	2 cases of dry mouth, 2 cases of dizziness, 1 case of headache.	10 cases of dry mouth,3 cases of blurred vision, 12 cases of dizziness, 2 cases of headache.
Chen,201830	CHM + WM vs WM	3 cases of mild headache, 2 cases of gastrointestinal reactions such as nausea and dyspepsia.	4 cases with mild headache, 3 cases with gastrointestinal reactions such as nausea and dyspepsia.
Li, 201831	CHM + WM vs WM	3 cases of mild skin redness 3 cases of mild skin redness and rash, 2 cases of mild headache and dizziness.	4 cases of nausea and vomiting, 3 cases of mild skin redness and rash, 3 cases of mild headache and vertigo.
Xiao, 201732	CHM + WM vs WM	1 case of weakness, 2 cases of dizziness, 1 case of gastrointestinal discomfort.	4 cases of weakness,3cases of dizziness, 6 cases of gastrointestinal discomfort.

Publication bias

The distribution is almost symmetrical, as shown by the funnel plots of the total effective rate (Figure 4). After quantitative analysis, Begg’s test resulted in Z = 2.24, P = .025, and Egger’s test resulted in P = .015, 95%CI = 0.57 to 3.88, indicating that there is no publication bias.

Figure 4.

Funnel plot of the total effective rate.

Sensitivity analysis

According to the forest plot of PSQI component scores (the quality of asleep, the efficiency of sleep and sleep disorder), there was a high heterogeneity. In terms of CHM compared with WM, sensitivity analysis revealed that after excluding (Yang, 2016),21 the heterogeneity was reduced to 0 (Figure S4-A, available at: http://links.lww.com/HNP/A25; Figure S5-A, available at: http://links.lww.com/HNP/A25; Figure S6-A, available at: http://links.lww.com/HNP/A25). Compared to the other 2 studies,20,22 the average duration of insomnia in Yang’s21 trial group was less than 1 year, while the other 2 articles were 1.5 years or more than 3 years. In terms of CHM + WM compared with WM, sensitivity analysis revealed that after excluding (Zhai, 2022),28 the heterogeneity was reduced to 0 (Figure S2-B, available at: http://links.lww.com/HNP/A25; Figure S5-B, available at: http://links.lww.com/HNP/A25), or after excluding (Chen, 2018),30 the heterogeneity was reduced to 0 (Figure S4-B, available at: http://links.lww.com/HNP/A25; Figure S6-B, available at: http://links.lww.com/HNP/A25). No adverse reactions were reported in (Zhai, 2022),28 whereas minor adverse drug reactions were observed in the trail and control groups of the other 3 studies.30-32 Compared to the other 3 studies,28,31,32 the average duration of insomnia in Chen’s30 trial group was more than 3 years, while the others were 1 year or so. Sensitivity analysis was also performed on endocrine hormone levels, and found that the length of treatment (Dong, 2022)16 (Figure S10-A, available at: http://links.lww.com/HNP/A27) and average duration of insomnia (Wang, 2022)29 (Figure S11-B, available at: http://links.lww.com/HNP/A27) might cause high heterogeneity.

DISCUSSION

This study conducted a systematic review and meta-analysis to evaluate the effectiveness and safety of CHM in treating perimenopausal insomnia. A total of 20 RCTs were included, comprising 8 RCTs with 663 patients comparing CHM combined with Western medicine (CHM + WM) to Western medicine (WM) alone, and 12 RCTs with 1,037 patients comparing CHM to WM. The meta-analysis findings revealed that either CHM alone or CHM combined with WM can significantly improve sleep quality and clinical total effective rate. Most indicators, such as PSQI score, total effective rate, and serum of FSH level, demonstrated that CHM use alone or in combination with WM outperformed WM alone groups. However, in terms of PSQI score—sleep efficiency, PSQI score—sleep disorder and serum levels of LH, CHM was comparable to the WM group. CHM combined with WM was comparable to the WM in terms of PSQI overall score and serum levels of E₂. No severe adverse CHM-related events were reported.

Studies have shown that FSH and LH work in harmony with the physical biological clock, that rhythm disturbances become more common during perimenopause and affect sleep regulation, and that low levels of ovarian hormone production, especially E₂, lead to sympathetic activation in the body and poorer sleep quality.36 This research also demonstrates that CHM alone significantly influences FSH and E₂ regulation, while CHM in combination with WM significantly influences FSH and LH regulation. Sleep, according to TCM, is dominated by the mind and spirit.

According to this study, CHM demonstrated effectiveness in the treatment of perimenopausal insomnia. This highlights a promising nursing intervention that registered nurses (RNs) may consider incorporating into their practice. RNs play a critical role in health education by informing patients about the potential benefits of CHM as an adjuvant therapy for perimenopausal insomnia. However, the clinical application of CHM necessitates careful consideration to ensure both safety and efficacy. Improper processing or prolonged use of high doses of CHM can pose safety risks, and potential interactions between herbal medicines and prescription drugs should be carefully evaluated.37,38 Our study also found that compared to the CHM stand-alone group, the combined CHM and WM group was more likely to suffer additional side effects. Therefore, nurses should encourage patients to consult qualified Chinese medicine practitioners to determine the suitability of incorporating CHM into their treatment plans. Additionally, limited access to CHM in certain regions where it is not widely available may present challenges for patients. In such cases, RNs can assist patients by providing guidance on identifying certified and reliable CHM suppliers. Future research should focus on assessing the efficacy of CHM for perimenopausal insomnia in larger, more diverse populations. Furthermore, studies aimed at identifying the specific bioactive components of CHM responsible for its therapeutic effects, as well as determining optimal dosing regimens and treatment durations, are needed. These efforts will contribute to refining clinical protocols, optimizing therapeutic outcomes, and strengthening the evidence base for the use of CHM in managing perimenopausal insomnia.

The primary strengths of this study are as follows. It is the first Meta-analysis to systematically evaluate the efficacy and safety of CHM as a standalone and adjunctive intervention for perimenopausal insomnia. This study evaluated a comprehensive range of outcomes, with particular emphasis on a detailed comparative analysis of the PSQI overall score and its component scores. This analysis significantly enhanced our understanding of how CHM functions in treating perimenopausal insomnia. However, several limitations of this study should be acknowledged. First, women’s sleep patterns can be influenced by vasomotor symptoms, anxiety, depression, and osteoporosis associated with perimenopause. These factors were not addressed in this study. When more high-quality RCT studies are added, it is hoped that research on the efficacy of CHM in reducing such symptoms can be broadened. Second, Subjective variables may have influenced the scale scores and symptom ratings, potentially leading to inaccuracies. Third, few studies included long-term follow-up, and the long-term efficacy of CHM remains to be confirmed.

CONCLUSIONS

This systematic review and meta-analysis demonstrated that CHM is an effective and safe alternative therapy for perimenopausal insomnia, both as a stand-alone treatment and as an adjunctive intervention. Its primary benefits are reflected in improved sleep quality and better regulation of sex hormone levels. However, further high-quality RCTs are needed to confirm the effectiveness of this therapy.