Acupuncture for insomnia

Objectives

This is a protocol for a Cochrane Review (intervention). The objectives are as follows:

To assess the benefits and harms of acupuncture in the treatment of insomnia.

Background

Description of the condition

Insomnia is a subjective complaint of poor sleep, which may produce impairment in daytime function or mood. The 5th edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM‐V) defines insomnia disorder as subjective difficulty in initiating or maintaining sleep or non‐restorative sleep lasting at least three months (APA 2013). Insomnia is very common and can be acute, intermittent, or chronic (AASM 2014). More than 100 differential diagnoses of insomnia have been listed in the revised edition of the International Classification of Sleep Disorders (AASM 2014). In order to meet the diagnostic criteria for insomnia disorder, one must have sleeping difficulties at least three nights per week for at least three months despite adequate opportunity for sleep and resulting in at least one associated daytime impairment affecting social, occupational, academic, or other essential areas of functioning (AASM 2014; APA 2013). Since there is no universally established definition of 'normal' sleep, the estimates of insomnia prevalence vary widely due to differences in case definitions, assessment procedures, sample characteristics, and time intervals between assessments. The American Academy of Sleep Medicine (AASM) has estimated that about 30% of adults experience symptoms of insomnia (AASM 2005).

In Hong Kong and the USA, the prevalence of insomnia disorders was found to be 22.1%, 3.9%, and 14.7% for DSM‐IV, ICD‐10 (International Classification of Diseases, 10th Revision), and ICSD‐2 (International Classification of Sleep Disorders, 2nd Edition), respectively (Chung 2015). Women and older adults tend to report higher rates of sleep problems (Crowley 2011; Jaussent 2011; Suh 2018). The COVID‐19 pandemic further exacerbated the burden of insomnia (Halsøy 2021). There has been evidence from multiple sources suggesting that the world is facing an increasing risk of insomnia, with 37% of adults, 39% of healthcare workers, 58% of university students, 54% of people with COVID‐19 infection, and 53% of pregnant women suffering from insomnia‐related symptoms (Morin 2021; Pappa 2020; Yuan 2022).

Insomnia is caused by a variety of physiological, psychological, and environmental factors. Physiological determinants include pain, toxic and metabolic disturbance, effects of medications, and sleep‐disordered breathing, while anxiety and depression are prominent psychological causes of poor sleep. Environmental factors such as the physical and social features of the surrounding neighbourhood and its atmosphere could affect sleep health (Billings 2020).

The sleep deprivation of insomnia leads to a myriad of adverse effects. Insomnia sufferers report lower quality of life than good sleepers (Léger 2012). Insomnia may result in cognitive deficits, fatigue, irritability, impaired concentration, an increased risk of traffic accidents, impaired job performance, and absenteeism (Buysse 2013; Kucharczyk 2012; Laugsand 2014; Riemann 2015; Soong 2021). The combined direct and indirect costs of insomnia in the USA exceed USD 100 billion annually (Wickwire 2016). Moreover, continued unresolved insomnia is associated with significant psychiatric morbidity, such as depression (Freeman 2020; Manber 2009), reduced pain tolerance (Liu 2021; Todd 2022), and impaired immune function (Garbarino 2021). In addition, there is a robust correlation between insomnia and various psychological disorders, such as anxiety disorders, alcohol and drug abuse or dependence, and suicide (Perlis 2020; Taylor 2007). Insomnia might also be a risk factor for obesity (Crönlein 2016), hypertension (Lanfranchi 2009; Li 2015; Vgontzas 2009a), diabetes (Vgontzas 2009b), cardiovascular disease (Javaheri 2017), and mortality, but the data are inconclusive (Taylor 2007). Since short‐term insomnia, if left untreated, can progress to long‐term insomnia (Ancoli‐Israel 2000), which might be more difficult to treat, it is important to evaluate and treat insomnia early (Perlis 2020).

Description of the intervention

Insomnia, a multifaceted condition, can be addressed through a range of treatment modalities, categorised broadly as pharmacological and non‐pharmacological treatments. Pharmacological approaches encompass a host of agents such as benzodiazepines, lemborexant, doxepin, seltorexant, and zaleplon, among others. However, evidence pertaining to the efficacy and safety of some pharmacological interventions remains inadequate, limiting definitive conclusions (De 2022). Notably, drug‐induced improvements in sleep parameters often evaporate post‐discontinuation (Sateia 2017), and extended use of hypnotic drugs could result in tolerance, addiction, and ancillary side effects (Wilson 2019). Non‐pharmacological strategies for insomnia typically involve lifestyle modifications and behavioural strategies to foster beneficial sleep habits, which include but are not limited to Cognitive Behavioural Therapy for Insomnia (CBT‐I), incorporating cognitive therapy, sleep restriction, stimulus control, sleep hygiene, relaxation techniques, and, notably, acupuncture (Chan 2021; Soong 2021). This review concentrates on the efficacy of acupuncture as a treatment for insomnia.

Acupuncture, a traditional medicine modality, involves the insertion of extremely fine needles through the skin at specific body points, denominated as acupuncture points (Zhuang 2013). It has been employed to treat a diverse array of chronic conditions including, but not limited to, back pain, arthritis, headaches, asthma, and psychiatric disorders (Asher 2017; Kelly 2019; McCarney 2004). The underlying premise of acupuncture is to equilibrate the flow of vital energy or 'qi', believed to course through body pathways. It is posited that needle insertion at these points can re‐establish energy balance, thereby facilitating healing and promoting wellness (Huang 2009; Zhang 2012).

Acupuncture encompasses a variety of forms, including electroacupuncture, laser acupuncture, and injection acupuncture (Chan 2017). Furthermore, it can also be classified into body acupuncture, auricular acupuncture, scalp acupuncture, and their combinations, contingent on the specific stimulation of body parts and acupuncture points (Zhao 2013). Acupuncture is a practice involving the insertion of fine needles into acupoints for therapeutic purposes, in which the needles may be stimulated in multiple forms. Manual acupuncture is a form of needle manipulation technique to induce needling sensation, including rotation, lifting, or thrusting. Electroacupuncture is the use of small electric currents to stimulate pairs of acupuncture needles. Laser acupuncture is the photonic stimulation of acupuncture points to initiate therapy effects similar to that of needle acupuncture. Injection acupuncture is a procedure in which drugs, vitamins, herbal extracts, or other fluids are injected into acupoints using a syringe and needle (Ang 2024; Cheuk 2012).

In relation to insomnia, acupuncture is proposed to help by stimulating the nervous system, which in turn might influence the production of neurotransmitters and biochemicals integral to sleep regulation. Various studies indicate that acupuncture may impact serotonin and melatonin levels, both critical to the sleep‐wake cycle (Zhao 2013). Selection of acupuncture points could be tailored to the underlying aetiology of insomnia, as discerned by traditional medicine principles. For instance, points on heart, kidney, or spleen meridians might be selected to treat insomnia stemming from emotional distress, anxiety, or overthinking.

The broad clinical applicability of acupuncture for insomnia is underscored by the variety of acupuncture treatment types and the extensive range of insomnia disorders it addresses. However, it should be emphasised that outcomes may vary among individuals, and research is ongoing to unravel the comprehensive effects and potential benefits of acupuncture.

For many individuals attracted to acupuncture, its potential adverse effects, whether concerning presence or minor nature, have been both the principal concern and the appeal. However, the current understanding and evidence on the adverse effects of acupuncture is fragmented. Studies have been performed systematically to examine the adverse effects of acupuncture and suggest that most complications are minor, and the complication occurrence rate is low (Vincent 2001; White 2001). The most common complications were bleeding, needling pain, nausea, dizziness, bruising, and aggravation of existing symptoms (Adams 2011; MacPherson 2001; White 2001). Although major complications are uncommon, they also require attention as serious tissue or organ injuries and even mortality may occur due to incompetent practitioners. Serious complications of acupuncture include pneumothorax, injury to the spinal cord, nerve injuries, infections, and death (Chan 2017; Vincent 2001). Similar to adults, the majority of the complications associated with paediatric acupuncture were also mild with few serious adverse effects reported (Adams 2011). Nevertheless, the balance of benefits and harms of acupuncture remains to be determined.

How the intervention might work

Acupuncture has shown promise in treating insomnia through several potential mechanisms. Acupuncture improves sleep structure by inhibiting sympathetic activity and down‐regulating the hypothalamic‐pituitary‐adrenal axis (Lin 2022). Hypothalamic adrenocorticotropic hormone‐releasing hormone, serum adrenocorticotropic hormone, and corticosterone levels were reduced in insomnia rats assigned to acupuncture, which showed that the hypothalamic‐pituitary‐adrenal axis influences sleep mechanisms (Wu 2017). Functional magnetic resonance imaging studies suggest that activity in certain regions of the brain is associated with the experience of sleep and that acupuncture can modulate activity in these regions (Wang 2020). The clinical efficacy of acupuncture for insomnia may be mediated by a variety of neurotransmitters, including norepinephrine, melatonin, γ‐aminobutyric acid, and β ‐endorphins. Stimulation of certain acupoints is found to increase nitric oxide in the brain and the blood, which is associated with sleep improvement clinically (Li 2003; Zhao 2013). Acupuncture influences sleep mechanisms by regulating amines, amino acids, and peptides, which are the three different types of central neurotransmitters in the brain that influence sleep (Wang 2023). Acupuncture could improve sleep quality by upregulating serum γ‐aminobutyric acid (GABA) and 5‐HT levels in individuals with insomnia (Li 2022a). In an animal model, acupuncture also increased the expression of 5‐HT1AR and 5‐HT2AR in the hippocampus, which improved the symptoms of insomnia in rats (Li 2022b). Acupuncture can also regulate the sleep‐wake cycle by modulating TNF‐α, IL‐6, and IL‐1β in the hypothalamus and alleviating the inhibitory release of 5‐HT (Rockstrom 2018; Tang 2019).

Why it is important to do this review

Acupuncture is widely used for the treatment of insomnia. Anecdotal reports suggest that acupuncture may improve sleep and relieve insomnia. Over the past decade, acupuncture has been widely used to treat patients with insomnia. Many new clinical trials have also been performed to study the efficacy of acupuncture for insomnia (Fu 2017; Jiang 2024; Yin 2022; Zhang 2020). However, it remains uncertain whether the existing evidence is rigorous enough to reach a definitive conclusion. A systematic review of randomised controlled trials of acupuncture therapy for insomnia was carried out in 2006 (Cheuk 2007) and updated by the authors in 2012 (Cheuk 2012). However, this question remains unresolved. Many systematic reviews have been published concerning this topic (Cao 2019), for example, involving scalp acupuncture and auricular acupuncture (Lan 2015; Liu 2021b), and a network meta‐analysis has been published to assess the efficacy of acupuncture for cancer‐related insomnia (Chen 2024). However, these reviews focused on certain acupuncture forms or disorder‐related insomnia, included studies with suboptimal quality, or did not include only populations that met clear diagnostic criteria. The results of these reviews are therefore not adequate to provide strong support for the use of acupuncture in the treatment of insomnia. Considering the substantial new evidence available, a comprehensive up‐to‐date review is needed. This review will ensure an adequate systematic review process is followed using the latest Cochrane methodology.

Objectives

To assess the benefits and harms of acupuncture in the treatment of insomnia.

Methods

Criteria for considering studies for this review

Types of studies

We will include all randomised controlled trials, including parallel or cluster designs. We will include studies comparing acupuncture or its variants with at least one control group that uses no treatment, placebo treatment, or sham treatment. Both published and unpublished studies will be included.

We will exclude cross‐over trials, as they are generally unsuitable for assessing longer‐term outcomes in chronic diseases, and tend to focus on interventions with transient effects on stable, chronic conditions during the study period (Higgins 2022a). We will exclude quasi‐randomised studies to reduce sources of potential heterogeneity because the allocation methods used in these studies are not truly random (e.g. resident record number or alternation). If the randomisation method used in the study is unclear, making it impossible for us to determine whether it is a true randomised controlled trial, we will contact the original authors to obtain the relevant information.

Types of participants

Participant characteristics

We will include studies with participants of any age, regardless of their gender, ethnicity, or religion.

Diagnosis

We will include participants with insomnia explicitly documented by standardised measures (e.g. the Pittsburgh Sleep Quality Index; Buysse 1989), objective measures in the sleep laboratory (e.g. polysomnography, actigraphy), or by reports/diaries kept by participants, partners, other informants or nursing staff; or participants with insomnia diagnosed by standard diagnostic criteria such as Diagnostic and Statistical Manual of Mental Disorders (APA 2013), International Classification of Sleep Disorders (AASM 2014), or International Classification of Diseases (WHO 2019), or with a complaint of sleep difficulties.

Comorbidities

We will also include participants with comorbid mental disorders (e.g. anxiety, depression, bipolar disorder, and post‐traumatic stress disorder) or health conditions (e.g. parkinsonian syndromes, cancer, sleep apnoea, pain, and epilepsy).

If stratified randomisation is used to randomly group relevant insomnia participants, only studies that include a subset of insomnia participants and report their data separately will be eligible.

Types of interventions

Trials evaluating all forms of acupuncture therapy, including acupressure, laser acupuncture, and electroacupuncture will be included in the review. There will be no restrictions on the number of treatments or the length of the treatment period. We will include either traditional acupuncture in classical meridian points or contemporary acupuncture in non‐meridian or trigger points, regardless of the source or methods of stimulation (e.g. hand, needle, laser, or electrical stimulation).

The control interventions will be no treatment, placebo acupuncture, or sham acupuncture. Placebo acupuncture refers to a needle attached to the skin surface (not penetrating the skin but at the same acupoints) (Mu 2020). Sham acupuncture refers to a needle placed in an area close to but not in acupuncture points (Mu 2020) or subliminal skin electrostimulation via electrodes attached to the skin (SCSSS 1999).

The comparisons that we will investigate are listed below:

1. Acupuncture only versus no treatment

2. Acupuncture only versus placebo or sham treatment

3. Acupuncture adjunctive to other treatment versus other treatment alone

4. Acupuncture adjunctive to other treatment versus placebo or sham treatment adjunctive to other treatment

'Other treatment' refers to any treatment, including medication, psychological treatment, or alternative complementary treatment, provided that both the intervention and the control groups received the same treatment. We will exclude trials that compared only different forms of acupuncture or compared acupuncture with other forms of treatment.

Minimum duration of intervention

There is no minimum duration of intervention.

Minimum duration of follow‐up

The minimum duration of follow‐up will be from baseline to immediately post‐treatment.

Types of outcome measures

We will include studies that meet the above inclusion criteria, regardless of whether they report on the following outcomes.

Primary outcomes

Our primary outcome will be the improvement in sleep quality.

• Change in sleep index, as measured by any self‐rated validated scale, including the Pittsburgh Sleep Quality Index (PSQI) (Buysse 1989), the Insomnia Severity Index (ISI) (Bastien 2001), or the Leeds Sleep Evaluation Questionnaire (LSEQ) (Parrott 1980; Zisapel 2003). If other standardised scales are used in some trials, we will use them in the absence of PSQI, ISI, or LSEQ.

• Number of participants who dropped out due to any adverse event.

• Number of participants experiencing at least one adverse event.

Secondary outcomes

Where data are available, the secondary outcomes that we will include are as follows:

• Sleep parameters, as measured by sleep diaries or other objective measurements, such as actigraphy, electroencephalography, or polysomnography

  • Sleep onset latency

  • Total wake time

  • Total sleep duration

  • Wake after sleep onset (WASO)

  • Early morning wakening (defined by the trialist)

  • Other sleep efficiency (ratio of time asleep to time in bed)

• Daytime functioning, as measured by attentional tasks tests, self‐report using a standardised measure, e.g. the Stanford Sleepiness Scale (Hoddes 1973) or the Epworth Sleepiness Scale (Johns 1991).

• Quality of life (QoL), measured using standardised scales, such as the Short Form‐36 (SF‐36) (Hays 1993), the World Health Organization Quality of Life (WHOQOL) assessment (WHO 1995), or Health of the Nation Outcome Scales (HoNOS) (Wing 1999).

• Frequency of improvement in sleep quality (proportion of participants satisfied with insomnia improvement), measured as a dichotomous outcome of improvement. Since improvement in sleep quality is subjective, for the purposes of this review it could be variably defined, with or without the use of a sleep score or other sleep parameters (e.g. sleep onset latency, total sleep duration, total wake time, wake after sleep onset).

• Number of dropouts from study or treatment (all‐cause dropout) within trials, where data on reasons for dropout will be collected and summarised.

• Adverse effects ‐ we will define adverse effects as defined by individual studies (e.g. unfavourable symptoms that occurred during the course of the study).

Timing of outcome assessment

We will summarise the results of the intervention at the end and the follow‐up points of each report. We will divide outcomes, where possible, into immediate post‐treatment, short‐term (up to three months), medium‐term (three to six months), and long‐term (more than six months). If a study reports multiple time points within one of the prespecified time periods, we will select the latest time point. We will report outcomes at the first assessment post‐intervention in the abstract and summary of findings table.

Hierarchy of outcome measures

When an included study reports more than one measurement scale for the same outcome, we will only include validated scales. When more than one validated scale is used, priority will be given to the most commonly used scale in clinical settings, but we will include data from all validated scales. If a study assesses insomnia symptoms through multiple instruments, we will prioritise these methods in the following order: clinician rating scale, informant rating scale, and self‐rating scale.

Minimally important difference

​​​​​​If possible, we will compare the pooled estimates with the minimally important difference (MID) values for continuous outcomes to aid interpretation. We will use published MIDs when available. When multiple MID estimates are available for an outcome, we will use the smallest validated MID.

Search methods for identification of studies

Electronic searches

We will search the Cochrane Central Register of Controlled Trials (CENTRAL, latest issue) via the Cochrane Library, MEDLINE via Ovid (1966 to date), PsycINFO via Ovid (1887 to date), CINAHL via EBSCO (1982 to date), and AMED via Ovid (the Allied and Complementary Medicine Database, 1985 to date). In addition, we will search the following electronic bibliographic databases: China National Knowledge Infrastructure (CNKI) (https://www.cnki.net); Chinese Science and Technology Periodical Database (VIP) (https://www.cqvip.com); WANFANG (China Online Journals) (https://www.wanfangdata.com.cn). We will not include Embase in our search, as RCTs indexed in Embase are now prospectively added to CENTRAL via a highly sensitive screening process (Higgins 2022b).

We will search trial register databases for registered trials: ClinicalTrials.gov (https://clinicaltrials.gov/); the World Health Organization International Trials Registry Platform (WHO ICTRP) (https://www.who.int/clinical-trials-registry-platform); the Australian and New Zealand Clinical Trial Registry (https://www.anzctr.org.au/); Chinese Clinical Trial Registry (https://www.chictr.org.cn/indexEN.html); the Trials Register of the Cochrane Complementary Medicine Field; and the Cochrane Depression, Anxiety and Neurosis Group Controlled Trials Register (CCDANCTR).

Searching other resources

We will search the reference lists of all relevant papers for further studies. The process of searching many different sources may bring to light direct or indirect references to unpublished studies. We will seek to obtain copies of any such unpublished material.

Before carrying out analyses, we will search for retractions and publication corrections within our set of eligible studies using the software EndNote, which is integrated with the Retraction Watch database (http://retractiondatabase.org).

There will be no language restriction in the search and inclusion of studies. However, we will exclude multiple publications reporting the same group of patients or their subsets.

Data collection and analysis

Selection of studies

Two review authors (LA and ES) will independently review titles and abstracts of references retrieved from the searches and will select all potentially relevant studies. Any disagreements will be discussed with a third review author (MSL). We will obtain copies of these articles, and they will be reviewed independently by the same authors against the inclusion criteria of the review. The review authors will not be blinded to the names of the authors, institutions, or journals of publication. In addition, the review authors will contact colleagues and experts in the field to identify any unpublished or ongoing studies. We will also contact the author of the included studies to obtain any unpublished data and missing information, if necessary. We will collate the decisions made regarding the studies for inclusion using the Covidence software (Covidence).

Dealing with duplicates and companion publications

In the event of duplicate publications, companion documents, or multiple reports of a primary study, we will maximise the information yield by collating all available data, and we will use the most complete data set aggregated across all known publications. We will list duplicate publications, companion documents, multiple reports of a primary study, and trial documents of included trials (such as trial registry information) as secondary references under the study ID of the included study; these will not be regarded as independent included studies.

Data extraction and management

Two review authors (TYC and JHJ) will independently extract data such as study characteristics, methodological details, and treatment effects from reports of each study using a piloted data extraction form. Any disagreements will be discussed with a third review author (LA).

We will extract the following data:

1. General information, including author's last name, year of publication, title, journal information (title, volume, pages), or the unpublished source, country, and language of publication.

2. Study methods, including study design (e.g. parallel or cross‐over design), randomisation method, method of allocation concealment, blinding method, stratification factors used if stratified randomisation was employed, losses to follow‐up, and washout period in cross‐over studies.

3. Participants, including inclusion/exclusion criteria, number (total/per group), age and sex distribution, specific diagnosis/diagnostic subtypes, associated physical or neuropsychiatric diseases, duration of the disorder, previous treatments, and comorbidities.

4. Intervention and control, including the type of acupuncture, details of treatment regime (including procedures, who provided, how, when, frequency, tailoring, and modifications) (Hoffmann 2014), type of control, details of control treatment including drug dosage, and details of co‐interventions.

5. Follow‐up data, including duration of follow‐up, dates of treatment withdrawal and reasons for treatment withdrawal, withdrawal rates.

6. Outcome data, including change in sleep index (measured by any self‐rated validated scale such as PSQI, ISI, or LSEQ), sleep parameters on sleep onset latency, total wake‐time, total sleep duration, wake after sleep onset, early morning wakening, and other sleep efficiency (measured by sleep diaries or other objective measurements), daytime functioning (measured by attentional tasks tests, self‐report using a standardised measure), quality of life (measured by standardised scales such as SF‐36, WHOQOL, or HoNOS), sleep quality, which is the proportion of participants satisfied with insomnia improvement (measured as a dichotomous outcome of improvement), number of dropouts from study or treatment (all‐cause dropout), and adverse effects.

7. Analysis data, including pretest means and post‐test means or change scores and standard deviations (SD), for all groups for all outcomes in Primary outcomes and Secondary outcomes, baseline differences, and statistical techniques

Two review authors will independently use WebPlotDigitizer or Engauge Digitizer to extract data from graphs or figures (Engauge Digitizer 2019; WebPlotDigitizer 2021). Data will be entered into RevMan by one review author and then checked by the second review author.

Assessment of risk of bias in included studies

Two review authors (LA and ES) will independently carry out the assessment of risk of bias according to the Cochrane Handbook for Systematic Reviews of Interventions using the RoB 2 tool (Higgins 2022a; Sterne 2019). If the authors of this review are also authors of any included studies, these authors will not participate in the risk of bias assessment of their own studies. We will resolve any disagreement by discussion among all review authors. We will assess the risk of bias according to the following domains:

1. Bias arising from the randomisation process

2. Bias due to deviations from intended interventions

3. Bias due to missing outcome data

4. Bias in the measurement of the outcome

5. Bias in the selection of the reported result

We will judge each domain to be at low risk of bias, have some concerns, or be at high risk of bias. When the risk of bias is unclear due to the lack of information, or we are uncertain of the potential for bias, we will contact the trial investigators to seek clarification. We will use the data available in the published report if we do not receive a response within four weeks. We will also reach an overall risk of bias judgement for each prespecified outcome by evaluating the risk of bias across all five domains. We will assess the trial to be at overall low risk of bias if all domains of the trial are assessed as low risk of bias, to be at overall high risk of bias if at least one domain is assessed as high risk of bias, or if multiple domains of this trial are assessed to have some concerns, and to have overall some concerns if only one domain is assessed to have some concerns and the other domains are assessed as low risk of bias.

We will assess the risk of bias of a specific result of cluster‐randomised trials using the Cochrane RoB 2 tool (Higgins 2022a), Bias arising from the timing of identification and recruitment of individual participants within clusters in relation to timing of randomisation, as outlined in the Cochrane Handbook for Systematic Reviews of Interventions.

For this review, we will evaluate the intervention assignment effect (intention‐to‐treat effect) for the main outcomes. We will perform an evaluation using the RoB 2 tool for the following outcomes:

• Change in sleep index: average endpoint or change score on a general depression scale (e.g. PSQI, ISI, LSEQ) – medium‐term

• Daytime functioning: average endpoint or change score on a functioning scale – medium‐term

• Number of dropouts from the study or treatment within trials ‐ medium‐term

• Quality of life (QoL): average endpoint or change score on quality of life scale – medium‐term

• At least one adverse event – long‐term

We will use the Excel tool to implement RoB 2 (available at www.riskofbias.info/welcome/rob-2-0-tool/current-version-of-rob-2). We will store the RoB 2 data, and make the data available as supplemental files. We will use Review Manager (RevMan) to input and visualise the risk of bias results (RevMan Web 2022). We will conduct the review according to this published protocol and report any deviation in the 'Differences between protocol and review' section of the review.

Measures of treatment effect

We will use odds ratio (OR) or risk ratio (RR) with 95% confidence intervals (CI) as appropriate for binary outcomes. We will use the mean difference (MD) for continuous outcomes, analysing with the inverse variance method. If there are outcomes that could be reported as time‐to‐event data, we will analyse them as hazard ratios. All analyses will include all participants in the treatment groups to which they were allocated.

Unit of analysis issues

If there are two different intervention or control groups, we will report pairwise comparison results under different subgroups of intervention or control groups in a particular comparison, and their results will not be combined into a single summary measure; we will only present subgroup summaries to preserve the identification of different intervention subgroups, as well as to avoid an incorrect unit of analysis. If a study has multiple treatment arms, we will only use comparisons between the acupuncture intervention and the control/treatment as usual group. If there is a different acupuncture method at different points, we will conduct different subgroup analysis in a particular comparison, and these results will not be combined into a single summary measure. In the case of cluster‐RCTs, we will use the intra‐cluster correlation coefficient to adjust the data, then combine with the data from individual RCTs for analysis.

Dealing with missing data

We will contact study authors to request data when missing. If data remain unavailable, we will try to estimate the missing data using the available information from the study, such as CIs, based on the methods outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2022). Where possible, we will attempt to impute missing data. Where this is not possible, we will report the study narratively and discuss its impact on the overall assessment of results. For studies in which the standard deviation (SD) of the outcome is not available at follow‐up, we will standardise by the mean of the pooled baseline SD from studies that reported this information for that scale.

Assessment of heterogeneity

We will assess clinical and methodological heterogeneity by examining the characteristics of the studies. We will assess clinical heterogeneity by noting the difference in the distribution of important participant factors between trials (age, gender, specific diagnosis/diagnostic subtypes, duration of disorder, associated diseases) and trial factors (randomisation concealment, blinding, losses to follow‐up, treatment type, co‐interventions). We will assess statistical heterogeneity visually from the forest plots, and using the I² and Chi² statistical tests.

As recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2022), the interpretation of an I² value of 0% to 40% may not be important; 30% to 60% may represent moderate heterogeneity; 50% to 90% may represent substantial heterogeneity; and 75% to 100% may represent considerable heterogeneity.

We will avoid the use of absolute cutoff values, but interpret I² in relation to (a) the size and direction of effects, and (b) the strength of evidence for heterogeneity (e.g. P value from the Chi² test, or CI for I²).

Assessment of reporting biases

We will assess funnel plot asymmetry visually and statistically by means of the Bee and Mazumdar (Begg 1994) and the Egger tests (Egger 1997), only if we are able to include at least 10 studies in a meta‐analysis, as the power of the tests will be low with fewer studies. If asymmetry is suggested by visual assessment or detected in any of these tests, we will examine the possible reasons (Deeks 2022). We will attempt to minimise the potential for publication bias through our comprehensive search strategy that includes evaluating published and unpublished literature. We will contact authors for information about multiple publications in case of doubts.

Data synthesis

We will include all eligible studies in the primary analysis. A meta‐analysis will be conducted when there are data from at least two included studies and in the absence of substantial heterogeneity. We will use a random‐effects model by default because a certain degree of heterogeneity is expected among studies. We will present data using forest plots, when possible. If it is inappropriate to pool data statistically due to an insufficient number of studies or substantial heterogeneity between studies, we will provide a narrative description of study characteristics and findings according to the Synthesis Without Meta‐analysis guideline (Campbell 2020). We will use the latest version of RevMan Web to analyse the data (RevMan Web 2022).

Subgroup analysis and investigation of heterogeneity

Considering that effect modifiers influence how well interventions work in affecting outcomes (Deeks 2022), we plan to conduct the following subgroup analyses based on factors that may lead to heterogeneity in intervention effectiveness:

• Age (under 65 versus over 65 years old): the effectiveness and acceptability of intervention measures are affected as a person becomes older (De Crescenzo 2022). We plan to test the differences in results between participants under 65 years old and those over 65 years old.

• Baseline severity of insomnia: this may impact on the primary outcome of interventions (Su 2024). We plan to test for differences between different severities of insomnia at baseline using the ISI scale.

• Length of study/number of sessions: two reviews on acupuncture in the management of insomnia have demonstrated that the effectiveness of acupuncture intervention may be related to higher sessions of treatment and length of study (Xu 2024; Zhao 2022). We will test for differences between different numbers of sessions for acupuncture intervention and lengths of study.

• Types of acupuncture therapies: the 2012 Cochrane systematic review on acupuncture for insomnia showed that different therapies may differ in effect size and acceptability to participants (Cheuk 2012). We will test for differences between different types of acupuncture therapy (e.g. needle acupuncture, electroacupuncture, acupressure, magnetic acupressure).

• Types of comparators: a 2022 review on acupuncture in the management of insomnia suggested that different controls may yield differing effect sizes (Zhang 2022). Therefore, we will examine the effect of any acupuncture therapy compared to each distinct other treatment.

In particular, subgroup analyses with fewer than two studies per category are unlikely to be sufficient to identify a valid effect difference and therefore will not be highlighted in our results. We will use the formal test for subgroup interactions in RevMan and we will use caution in the interpretation of subgroup analyses, as advised in Chapter 10 of the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2022; RevMan Web 2022).

Sensitivity analysis

We plan to conduct sensitivity analyses to assess the impact of study quality, provided that sufficient studies are available. These include:

1. excluding those assessed with overall high risk of bias;

2. excluding those with a lower than 70% follow‐up rate as > 20% attrition leads to high bias and poses a threat to the study’s validity.

To assess the effect of dropouts on the dichotomous outcomes, we will perform another sensitivity analysis including the following.

1. Best‐case scenario:

• We will assume that all the dropouts from the treatment group had a positive outcome and all the dropouts from the control group had a negative outcome.

2. Worst‐case scenario:

• We will assume that all the dropouts from the treatment group had a negative outcome and all the dropouts from the control group had a positive outcome.

Summary of findings and assessment of the certainty of the evidence

We will create a summary of findings (SoF) table, according to the guidelines in Chapter 14 of the Cochrane Handbook for Systematic Reviews of Interventions (Schünemann 2022). We will prepare separate SoF tables for the comparisons listed below:

• Acupuncture only versus no treatment

• Acupuncture only versus placebo or sham treatment

• Acupuncture adjunctive to other treatment versus other treatment alone

• Acupuncture adjunctive to other treatment versus placebo or sham treatment adjunctive to other treatment

Each table will include the following primary outcomes, measured at the end of the intervention and short‐term follow‐up (with sufficient data available).

• PSQI

• ISI or LSEQ

• Number of participants who dropped out due to any adverse event

• Total number of participants experiencing at least one adverse event

We will follow the methods and recommendations described in Chapters 14 and 15 of the Cochrane Handbook for Systematic Reviews of Interventions, and we will use GRADEpro GDT to prepare the SoF tables (GRADEpro GDT). We will assess the overall risk of bias of the studies that we will include in the SoF tables, and use the assessments to inform the overall risk of bias domain of the GRADE approach for assessing the certainty of a body of evidence. Using the five GRADE considerations (study limitations, consistency of effect, imprecision, indirectness, and publication bias), two review authors (JZ and LA) will independently assess the certainty of the body of evidence as it relates to the studies that contribute data to the meta‐analyses for each outcome (Guyatt 2008). Any disagreement will be resolved by discussion between the authors and addressed by consulting a corresponding author (MSL) if no consensus can be reached. If the authors of this review are also authors of any included studies, these authors will not participate in the GRADE assessment of their own studies.

The certainty of evidence may be assessed as high, moderate, low, or very low. We will justify all decisions to downgrade the certainty of evidence using footnotes, and make comments to aid the readers' understanding of the review where necessary.

Appendices

Appendix 1. Preliminary MEDLINE (Ovid) search strategy

1. exp randomized controlled trial/

2. controlled clinical trial.pt.

3. random*.ti,ab.

4. placebo.ab,ti.

5. drug therapy.fs.

6. trial.ab,ti.

7.groups.ab,ti.

8. or/1‐7

9. (animals not (humans and animals)).sh.

10. 8 not 9

11. exp Acupuncture Therapy/

12. exp Acupuncture/

13. exp Electroacupuncture/

14. exp Meridians/

15. exp acupuncture points/

16. exp moxibustion/

17. exp Dry Needling/

18. electroacupuncture.tw.

19. acupressure.tw.

20. acupuncture.tw.

21. meridian$.tw.

22. needling.tw.

23. moxi$.tw.

24. acup$ point$.tw.

25. (trigger adj3 point$).tw.

26. or/11‐25

27. exp Sleep/

28. exp Sleep Wake Disorders/

29. (pre‐sleep$ or presleep$).tw.

30. sleep$.tw.

31. insomnia$.tw.

32. dyssomn$.tw.

33. (awake$ or wake$ or waking or awaking).tw.

34. or/27‐33

35. 10 and 26 and 34

Contributions of authors

Ang L, Lee MS, Zhang J, and Yao L conceived and designed the review.

All authors contributed to the development of the protocol.

Ang L and Zhang J drafted and revised the protocol.

All authors revised the protocol critically for important intellectual content and approved the final version.

Details of the contributions of authors are as follows:

Ang L: conceptualisation, methodology, project administration, writing – original draft, writing – review and editing.

Lee MS: conceptualisation, funding acquisition, investigation, methodology, project administration, supervision.

Choi T: methodology, writing – review and editing.

Jun JH: methodology, writing – review and editing.

Song E: methodology, writing – review and editing.

Lee HW: methodology, writing – review and editing.

Lee B: methodology, writing – review and editing.

Zhang J: conceptualisation, methodology, writing – original draft, writing – review and editing.

Cao L: methodology, writing – review and editing.

Kwon CY: methodology, writing – review and editing.

Wang Q: methodology, writing – review and editing.

Yao L: conceptualisation, methodology, supervision.

Sources of support

Internal sources

• Korea Institute of Oriental Medicine, Korea, South

  This review is supported by the Korea Institute of Oriental Medicine under grant number KSN2312021 and KSN2121211.

External sources

• None, Other

  This review did not receive any external funding.

Declarations of interest

Ang L: no known conflict of interest to declare.

Lee MS: no known conflict of interest to declare.

Choi T: no known conflict of interest to declare.

Jun JH: no known conflict of interest to declare.

Song E: no known conflict of interest to declare.

Lee HW: no known conflict of interest to declare.

Lee B: no known conflict of interest to declare.

Zhang J: no known conflict of interest to declare.

Cao L: no known conflict of interest to declare.

Kwon CY: no known conflict of interest to declare.

Wang Q: no known conflict of interest to declare.

Yao L: no known conflict of interest to declare.

No authors will be involved in the risk of bias assessment for their own study or in the GRADE assessment that uses outcome data from their own study.

New