Perimenopause represents a critical transitional period in a woman’s life, marking the shift from reproductive capability to menopause. According to the STRAW +10 criteria, it is characterized by a gradual decline in ovarian function and is divided into 2 substages: early perimenopause (Stage −2), marked by increased menstrual cycle variability and elevated early follicular phase FSH, and late perimenopause (Stage −1), defined by amenorrhea lasting≥60 days, significantly elevated FSH levels, and markedly reduced anti-Müllerian hormone (AMH) and antral follicle count (AFC) (1). This stage is accompanied by pronounced hormonal fluctuations and multisystem dysregulation, collectively manifesting as perimenopausal syndrome (PMS) (2). Among the most common and distressing symptoms is insomnia, affecting 30–50% of perimenopausal women. Emerging evidence suggests that the impact of insomnia extends beyond quality of life impairment, exerting a significant influence on cardiovascular and cognitive health through multifactorial pathways. Chronic sleep disturbances may contribute to dysregulation of the hypothalamic–pituitary–adrenal (HPA) axis, heightened systemic inflammation, and increased sympathetic nervous system activity, all of which can promote atherogenesis and elevate the risk of myocardial infarction (MI). A longitudinal population-based study using data from the UK Household Longitudinal Study found that individuals with elevated levels of social dysfunction and anhedonia—core dimensions of psychological distress—had a significantly increased risk of developing MI over a 10-year period (3). Moreover, in patients already diagnosed with coronary heart disease (CHD), significantly worse scores across multiple domains of mental health, including depression, anxiety, and loss of confidence, were observed (4). These findings suggest a bidirectional interplay between cardiovascular vulnerability and psychological distress, wherein insomnia may act both as a consequence and as a precipitating factor through its neurobiological and behavioral effects. Understanding these interconnections underscores the importance of early intervention in sleep problems to mitigate downstream cardiovascular and cognitive complications.

Current therapeutic approaches vary between Western pharmacological interventions and Chinese herbal medicine. Pharmacotherapy typically includes sedative-hypnotics and menopausal hormone therapy (MHT), which are effective in relieving vasomotor symptoms and preventing osteoporosis. However, long-term MHT use carries risks such as thromboembolic events and breast cancer (5, 6). In contrast, acupuncture as a core modality of (Traditional Chinese Medicine) TCM has garnered increasing attention for its multimodal therapeutic mechanisms. Evidence has already suggested that acupuncture exerts its effects through 3 primary pathways: (1) neuroendocrine regulation via the hypothalamic–pituitary-ovarian (HPO) axis to stabilize hormonal levels (7); (2) neuromodulation, enhancing central neurotransmitters such as serotonin and γ-aminobutyric acid (GABA), which contribute to improved mood and sleep quality (8); (3) immunomodulation, reducing inflammatory cytokines like interleukin-6 (IL-6), which are linked to neuroendocrine dysfunction and sleep disturbances (9). Recent advances in neuroimmunology have uncovered striking similarities in the pathophysiological pathways connecting neuroinflammation with concurrent sleep and cognitive disturbances. Studies of anti-CASPR2 encephalitis demonstrate that autoantibody interference with potassium channel complexes induces dual pathology—disrupting both memory consolidation (notably causing retrograde amnesia) and sleep architecture through combined hippocampal damage and inflammatory mediator release (10). This phenomenon bears remarkable resemblance to the cytokine-driven hippocampal sensitization seen in PMI, where elevated IL-6 and other inflammatory markers similarly impair memory networks and sleep–wake regulation. Such cross-condition parallels highlight immunoneuroendocrine pathways as critical therapeutic targets for sleep-cognitive comorbidities.

Currently, integrated AP has emerged as a promising strategy that may enhance therapeutic efficacy while making up the limitations of conventional pharmacological approaches (11). Preclinical studies have suggested that acupuncture may influence neuropharmacological pathways, potentially affecting drug metabolism, receptor activity, and central responsiveness (12). In areas such as chronic pain and addiction, acupuncture has been shown to reduce dependence on opioids, lower required dosages, and minimize adverse drug reactions (13, 14). Although such evidence is primarily drawn from these fields, similar neuromodulatory effects may be relevant to the treatment of hormonal insomnia.

However, pharmacotherapy alone often encounters issues such as tolerance, withdrawal symptoms, and residual sedation (15). Despite increasing interest in integrated approaches, current clinical evidence for AP in treating PMI remains limited. Systematic reviews have identified significant methodological inconsistencies, including heterogeneity in outcome measures, variability in intervention protocols (e.g., frequency, duration, and acupoint selection), and insufficient long-term follow-up. These limitations hamper the development of standardized clinical guidelines.

This study aims to systematically evaluate the efficacy of AP interventions for PMI by synthesizing data from RCTs. By standardizing outcome measures and comparing the effects of various interventions, this meta-analysis aims to provide an evidence-based foundation for integrative treatment strategies in the management of PMS.

PMI involve multiple physiological and psychological factors (30, 31). Women in the perimenopausal stage often experience hormonal fluctuations, particularly in estrogen and progesterone, which play a key role in promoting neurotransmitter balance, improving circadian rhythm, adjusting sleep structure, and indirectly influencing mood. When hormonal levels become disrupted, it may lead to sleep disturbances, irritability, and other symptoms (3, 32, 33). Acupuncture modulates the HPO axis by stimulating estrogen receptor (ER)-positive neurons in the hypothalamus, promoting endogenous E₂ secretion, while downregulating gonadotropin-releasing hormone (GnRH) pulsatility to reduce elevated FSH/LH levels (34). Additionally, acupuncture enhances β-endorphin release from the arcuate nucleus, further stabilizing hormonal fluctuations (35, 36). Sleep is not only related to hormonal changes but is also closely linked to autonomic nervous function (37). Due to hormonal fluctuations, the imbalance between the sympathetic and parasympathetic nervous systems results in overactive sympathetic activity at night, leading to issues like rapid heart rate, hot flashes, night sweats, and anxiety, which in turn affect falling asleep and maintaining deep sleep (38). Acupuncture counteracts this by increasing heart rate variability (HRV), reflecting enhanced parasympathetic tone, and reducing nocturnal norepinephrine (NE) release (39, 40). Pharmacological agents like clonidine (an α2-adrenergic agonist) may further suppress sympathetic outflow, but acupuncture provides sustained autonomic nervous system (ANS) rebalancing without drug dependence (41). In TCM, PMS is categorized under conditions like “disorders before and after menopause” and “organ restlessness” associated with both internal and external factors. Clinically, herbal treatments such as Gan Mai Da Zao Decoction for nourishing yin and blood, and Chai Hu Long Gu Mu Li Decoction for relieving depressive fire are commonly used (42, 43). From a biomedical perspective, these formulations may exert effects via anti-inflammatory pathways (downregulating NF-κB and IL-6) and antioxidant activity (enhancing superoxide dismutase [SOD]), which are also targeted by acupuncture (44). Evidence from studies has indicated that acupuncture is effective in managing PMI, and acupuncture combined with Western medication, as an alternative therapy, has shown advantages across multiple outcome measures (45). Acupuncture works by regulating the HPO axis, improving hormone levels in perimenopausal women, significantly reducing LH and FSH levels, and increasing E₂ levels post-treatment, indicating a positive effect on promoting endogenous hormone secretion and restoring hormonal balance (46). Moreover, acupuncture upregulates serotonin synthesis in the raphe nuclei and GABAergic activity in the hypothalamus, addressing neurotransmitter deficiencies linked to hyperarousal and mood disturbances (47, 48). Acupuncture also shows notable effects in neurological regulation. Research indicates that it can improve anxiety and depression by modulating neurotransmitter levels, such as 5-hydroxytryptamine (5-HT) and dopamine (DA), effectively reducing HAMA scores (49). According to TCM theory, PMI is often caused by liver and kidney yin deficiency or heart and spleen deficiency. Acupuncture enhances TCMS scores by unblocking meridians and harmonizing qi and blood, aligning closely with the holistic concept of TCM (50). Modern studies correlate acupuncture points (GV₂₀, HT₇, SP₆) with vagal stimulation, 5-HT release, and HPO axis modulation, bridging traditional mechanisms with biomedical evidence (51). The role of Western medication in combination therapy is mainly reflected in its impact on GABA receptors, thereby improving sleep quality, with significant advantages observed in PSQI and KMI improvement (52, 53). AP not only leverages the strengths of both approaches, acupuncture regulates the endocrine system, corrects neurotransmitter imbalances, and alleviates anxiety and depression, while Western medication provides rapid symptom control. The integration of AP produces a synergistic effect, with acupuncture partially mitigating the side effects of Western medication (45), making it a safe alternative therapy.

In this meta-analysis, the majority of included studies employed well-defined diagnostic criteria (mostly internationally recognized standards), while only a minority did not specify their diagnostic methods. This rigorous selection process enhances both the validity and clinical applicability of our findings. By focusing on studies that adopt standardized diagnostic thresholds endorsed by major clinical guidelines, we improved cross-study comparability, minimized diagnostic heterogeneity, and reduced misclassification bias. This methodological consistency ensures that our pooled results are both reliable and generalizable to patient populations meeting these widely accepted criteria. Moreover, the use of clearly defined diagnostic criteria enables more meaningful subgroup analyses and enhances the reproducibility of our study in future research.

Several outcome measures in this meta-analysis exhibited substantial heterogeneity (I² > 50%), which warrants careful consideration. Potential sources of variability may include differences within patient populations (disease severity, comorbidities), inconsistencies in practitioner technique, and variations in outcome assessment methods (subjective or objective measures) or follow-up durations. Importantly, potential confounding factors such as lifestyle variables (diet, exercise habits), concomitant medication use (hormone therapy, antidepressants), and socioeconomic status were not uniformly reported across studies, which may further contribute to heterogeneity. These factors could independently influence outcomes like sleep quality or mood scores, potentially obscuring the true treatment effect. To address this heterogeneity, we conducted sensitivity analyses by excluding studies with high risk of bias or outliers, which partially reduced inconsistency in some outcomes. Subgroup analyses based on key baseline characteristics (sample size, age, duration, and course) further clarified effect estimates. For outcomes with high heterogeneity, biological mechanisms may offer explanations. For instance, individual variations in hormonal sensitivity (estrogen receptor polymorphisms) or neurotransmitter profiles (serotonin transporter gene variants) could modulate responses to acupuncture or pharmacotherapy. Similarly, variations in sleep outcomes may reflect population differences in how perimenopausal circadian disturbances interact with therapeutic interventions. Nevertheless, residual heterogeneity suggests that unmeasured factors, such as unstandardized co-interventions or publication bias, may still influence results. Given these limitations, the evaluation should be interpreted with caution, particularly for outcomes with high heterogeneity. Future research should prioritize standardized protocols and rigorous reporting to minimize variability and enhance comparability across studies.

Building on established longitudinal methodologies from mental health research, future studies should develop validated clinical prediction tools to identify perimenopausal women most likely to benefit from integrated AP. Three key prognostic domains warrant investigation: (1) biological markers (baseline cortisol, IL-6, and estrogen profiles); (2) sleep architecture parameters (PSQI sub-scores and actigraphy-measured sleep efficiency); (3) psychological phenotypes (HAMA depression cluster scores and stress resilience scales). The proposed framework could adapt linear mixed-effect methods from substance use research to model treatment response trajectories, potentially incorporating dynamic symptom networks mapping insomnia severity to endocrine-immune fluctuations, machine learning analysis of acupoint response patterns from electronic health records, and digital phenotyping via wearable sleep-stage validation —all of which are approaches that would collectively address current evidence gaps in personalized treatment selection for PMI.

Future clinical implementation of acupuncture-pharmacotherapy could benefit from targeted health campaigns and personalized approaches informed by psychological profiles, building on models from vaccination promotion research. Similar to COVID-19 vaccine uptake strategies that considered personality traits and social support, tailored interventions accounting for patients’ stress resilience and health beliefs may optimize treatment adherence. Integration with menopausal health programs could further enhance accessibility and acceptance of this combined therapy.

The strengths of the study are reflected in the following aspects: (1) The study involved a comprehensive search across 8 databases, ensuring a wide scope and thorough content coverage; (2) The analysis included 8 commonly used clinical outcome indicators, making the results more accurate and credible; (3) During the literature inclusion process, strict criteria were applied to select the interventions (with experimental group receiving AP and control group only receiving the corresponding Western medication), which helped to avoid excessive heterogeneity to some extent; (4) The article evaluates the efficacy of AP in PMI, and the analysis results demonstrate that the combination therapy is more advantageous than Western medication alone in treating PMI, highlighting the innovation and unique advantages of TCM combined with pharmacotherapy.

The studies still have some limitations: (1) While our systematic search strategy underwent multiple iterative refinements across 8 databases, the absence of formal peer review by an information specialist represents a potential limitation in search methodology rigor; (2) The limited number of eligible studies, all of which were conducted in Chinese, may lead to potential bias stemming from linguistic or regional influences, and the exclusion of gray literature further restricts the generalizability of findings by omitting potentially relevant unpublished data; (3) When collecting data, the same indicators in different studies had varying units, and some units lack internationally recognized conversion standards, making analysis challenging; (4) Some indicators still showed high heterogeneity, suggesting potential subgroup analyses may be needed; (5) The distinctive characteristics of acupuncture intervention make genuine practitioner blinding methodologically unattainable in clinical research; (6) Currently, high-quality, blinded RCTs are still lacking in clinical practice, which has precluded a comprehensive analysis of the correlation between treatment effects and clinically meaningful thresholds (MCID), and long-term follow-up has also not been achieved; (7) The overall quality of evidence is relatively low; (8) The regression results, while offering exploratory insights, are underpowered due to small subgroup sizes (often <10) and should be viewed as hypothesis-generating given risks of unreliable estimates or spurious associations.

While the combination therapy of AP demonstrates considerable therapeutic potential, its long-term efficacy and MCID warrant further investigation through large-scale, multicenter RCTs with extended follow-up periods, particularly for distinct insomnia subtypes. Future studies should prioritize protocol optimization to facilitate clinical translation.