Efficacy and Underlying Mechanism of Acupuncture in the Treatment of Posttraumatic Stress Disorder: A Systematic Review of Animal Studies

Acupuncture is a nonpharmacological intervention that can be useful in the clinical management of posttraumatic stress disorder (PTSD), especially in situations with a lack of medical resources, including large-scale PTSD events such as disasters. Some clinical studies have reported the clinical effect of acupuncture in improving PTSD symptoms, but the underlying therapeutic mechanism has yet to be explored. Therefore, this review summarized the underlying therapeutic mechanisms of acupuncture in animal PTSD models. A comprehensive search was conducted in 14 electronic databases, and two independent researchers performed study selection, data extraction, and the methodological quality assessment. Twenty-four relevant studies were included in this review and summarized according to the proposed main mechanisms. In behavioral evaluation, acupuncture, including manual acupuncture and electro-acupuncture, reduced anxiety and fear responses and weakened fear conditioning, improved sleep architecture, reduced depressive symptoms, and alleviated disturbance of spatial learning and memory of PTSD animal models. The therapeutic mechanisms of acupuncture proposed in the included studies could be classified into two categories: (1) regulation of stress responses in the neuroendocrine system and (2) promotion of neuroprotection, neurogenesis, and synaptic plasticity in several brain areas. However, the methodological quality of the included animal studies was not high enough to produce robust evidence. In addition, mechanistic studies on specific aspects of acupuncture that may affect PTSD, including expectancy effects, in human PTSD subjects are also needed.


Introduction
Direct or indirect exposure to exceptionally threatening or terrifying events can lead to a debilitating mental disorder, i.e., posttraumatic stress disorder (PTSD). Characteristic signs and symptoms of PTSD include intrusion symptoms, avoidance, negative alterations in cognitions and mood, and alterations in arousal and reactivity [1]. In primary care, the prevalence of this disorder is common, and a systematic review in 2017 found that the lifetime prevalence of PTSD confirmed by diagnostic interviews ranged from 2% to 39.1%, with a median point prevalence of 12.5%, similar to that of depression [2]. Studies have shown that risk factors for PTSD include witnessing someone being killed, having no regular income after an earthquake, lower social support, drug and/or alcohol abuse, current smoking, history of chest pain, being female, low educational or socioeconomic status, and prior psychological trauma [3][4][5].

Quality Assessment
The risk of bias of the included studies was assessed using Systematic Review Centre for Laboratory Animal Experimentation (SYRCLE)'s risk of bias tool [19]. This tool contains 10 items related to selection bias (sequence generation, baseline characteristics, and allocation concealment), performance bias (random housing and blinding), detection bias (random outcome assessment and blinding), attrition bias (incomplete outcome data), reporting bias (selective outcome reporting), and other biases (any important concerns about bias not covered by other domains). Each item was evaluated as three categories: 'Yes', 'No', and 'Unclear'. Two researchers (CYK and BL) independently evaluated the risk of bias for the included studies, and any disagreements were resolved through discussion with a third researcher (SHK). Review Manager version 5.3 software (Cochrane, London, UK) was used to generate the risk of bias figure.

Data Analysis
The descriptive analysis of basic information, characteristics of animal models, acupuncture method, outcome, results, and the proposed mechanism of acupuncture for each included study was conducted. Due to the significant clinical heterogeneity between the studies included, quantitative synthesis was not performed.

Search Results
A total of 2603 studies were identified from the databases, and there were no studies from other sources. Duplicate studies were removed, and then, another 2312 articles were excluded through title and abstract screening. During the process of reviewing the full texts of the remaining 31 articles, three duplicate articles [20][21][22] and four studies with abstracts but no full texts [23][24][25][26] were excluded. Therefore, a total of 24 studies  were included in this review ( Figure 1). excluded through title and abstract screening. During the process of reviewing the full texts of the remaining 31 articles, three duplicate articles [20][21][22] and four studies with abstracts but no full texts [23][24][25][26] were excluded. Therefore, a total of 24 studies  were included in this review ( Figure 1).

Methodological Qualities of Included Studies
For the sequence generation domain, 7 experiments in 6 studies [42,43,46,[48][49][50] were rated as having a low risk of bias ("Yes") because they used appropriate random sequence generation methods, such as computer-generated random numbers. The remaining 19 experiments [27][28][29][30][31][32][33][34][35][36][37][38][39][40][41]44,45,47,48,50] were rated as having an unclear risk of bias because they did not describe the method. For baseline characteristics, allocation concealment, and blinding domains, all studies rated them as having an unclear risk of bias because they had no information regarding them. For the random housing domain, 5 experiments in 4 studies [37,44,48,50] randomly housing animals during the experiment were evaluated as having a low risk of bias, and 1 experiment [46] assigning cages according to groups was assessed as having a high risk of bias ("No"). One experiment [46] selecting animals at random for outcome assessment and 2 experiments [34,37] conducting blinding of outcome assessor were assessed as low risk of bias in random outcome assessment and outcome assessor blinding domains, respectively. Five experiments [42][43][44][45] that did not include all of the animals tested in the analysis were evaluated as having a high risk of bias in the incomplete outcome data domain. Eight experiments [28,30,41,44,[46][47][48]50] that did not mention any experimental results or relevant mechanisms were evaluated as having a high risk of bias in the selective outcome reporting domain. Nine experiments [27,28,[30][31][32][33]35,39,40] were evaluated as having a high risk of bias in other sources of bias domain because acupuncture treatment methods were not described in sufficient detail ( Figure 2 and Appendix E).

Anxiety & Fear
A total of nine studies [27,[32][33][34][35][36][37]39,40] targeted anxiety or fear symptoms of PTSD by using EA. Among them, 8 studies [27,32,33,[35][36][37]39,40] conducted an elevated plus-maze (EPM) test, and all studies showed that the EA group spent significantly more time in the open arms and more entries in the open arms than the model group, meaning reduced anxiety levels. Five studies [33,35,37,39,40] conducted open field tests (OFTs), and all studies reported that the EA group had significantly more entries in the center and more distance traveled in the center than the model group, although there was no difference between the two groups in total distance. In the case of time spent in the center, there was inconsistent results between studies, showing results in favor of the EA group in 4 studies [33,37,39,40] and no difference between the two groups in 1 study [35]. Two studies [39,40] conducted fear conditioning tests, and the contextual freezing time and cued freezing time were significantly shorter in the EA group than in the model group.
One study [34] assessed scene-and cue-conditioned fear response detection, and the results showed that memory extinction and reconstruction of each had significantly lower values in the EA group, although there was no significant difference between groups in memory acquisition. One study using MA [44] conducted the EPM test and showed that entries and time spent in open arms and the anxiety index were significantly higher in the MA group than in the model group, although there was an inconsistent result in entries in closed arms. Three studies [43,44,47] performed OFT and showed that there were favorable results to the MA group in entries in the center, horizontal crossing grid number, vertical frequency, and the number of fecal particles. However, there were inconsistent results in travel distance between the two groups.

Sleep Disturbance
One study [38] assessed sleep disturbance in PTSD using EA. There were significantly higher values of rapid eye movement sleep (REMS) and slow wave sleep stage 2 in the EA group than in the model group, although there was no difference in slow wave sleep stage 1 between the two groups. Two studies [42,49] assessed sleep disturbance in a PTSD model using MA. They showed that non-REMS and REMS latency and awakening period were significantly shorter in the MA group than in the model group. In addition, the total sleep time was significantly longer in the MA group.

Cognitive Symptom
Two [29,30] targeted spatial learning and memory symptoms of PTSD using EA. They [29,30] performed the Morris water maze (MWM) test, and the escape latency was significantly shorter in the EA group than in the model group. Three studies [43,45,47] targeted cognitive symptoms by conducting the MWM test using MA. In the MA group, the daily latency during the positioning navigation experiment was significantly shorter, and the space exploration experiment times crossing the platform or effective areas were significantly longer than those in the model group.

Depression
One study using MA [44] included a forced swim test (FST) to test the depression level, and the results showed that immobility time was significantly shorter and climbing time was significantly longer in the MA group than in the model group, although there was no difference between the two groups in swimming time.

Regulation of Stress Responses in the Neuroendocrine System
The altered function of the HPA axis has a significant impact on the pathology of PTSD and can all be related to stress reactivity, fear learning and extinction, and inflammation [51]. A study using EA suggested the mechanism of acupuncture against PTSD as downregulation of HPA axis activity [27]. Specifically a study found that EA regulated the levels of tumor necrosis factor (TNF)-α and interleukin (IL)-4 in the hypothalamus of the PTSD animal model [38]. Moreover, for the locus coeruleus, which plays an important role in stress responsiveness as a nucleus in the brainstem [52], one study reported that EA can reduce excessive stress responses of PTSD model by downregulating nitric oxide synthase (nNOS) expression in this region [31]. Finally, one study [29] reported upregulation of hippocampal mineralocorticoid receptor expression and downregulation of hippocampal glucocorticoid receptor expression after EA. Both are adrenal steroid receptors that affect the HPA axis through negative glucocorticoid feedback [53]. Importantly, the activation of these receptors is related to memory, behavioral responsibility, anxiety, and fear [54]. Moreover, another study [28] reported downregulation of hippocampal neuronal nNOS expression, which regulate the behavioral effects of glucocorticoids [55], after EA. These results suggest the role of acupuncture in neuroendocrine regulation.

Promotion of Neuroprotection, Neurogenesis, and Synaptic Plasticity in Brain
(1) Hippocampus: In PSTD, the hippocampus forms a link between situational stimuli and aversive events and plays an important role in memory recall [56]. Thus, hippocampal dysfunction is considered an important factor that causes PTSD or maintains it by interacting with traumatic experiences [56]. Furthermore, since chronic stress is known to cause atrophy of the hippocampus, interventions targeting adult neurogenesis and synaptic plasticity improvement in the hippocampus may be useful for PTSD and other mental disorders [57]. In this respect, several studies included in this review have reported the potential impact of acupuncture on the hippocampus: upregulation of thioredoxin (Trx) expression in the hippocampus [32], restoration of messenger ribonucleic acid (mRNA) levels of sirtuin 1 and monoamine oxidase A in the hippocampus [33], improvement and repair of synaptic plasticity in hippocampus [36], improvement of hippocampal neurogenesis [40], upregulation of the protein synthesis required for synaptic plasticity via the mammalian target of rapamycin pathway in the hippocampus [44], restoration of the structure of hippocampal neuronal cells [42,49], reverse of the discharge activity of neurons in the hippocampus [49], and regulation of abnormal neuronal cluster electrical activity in the hippocampal CA1 and CA3 regions [48,50]. (2) Amygdala: In addition to the hippocampus, the amygdala is the most clearly altered limbic region in PTSD, which can lead to an overactive reaction to fear stimuli [58]. In the fear circuit, the hippocampus is thought to play an important role in the explicit memory of traumatic events and mediate learned responses to contextual cues. The PFC reactivates past emotional associations, and the amygdala is activated in PTSD subjects and associated with oversensitivity to stress, generalized fear responses, and impaired extinction [58]. Thus, interventions that promote synaptic plasticity in the amygdala could potentially contribute to fear control and extinction [58]. In this respect, some studies included in this review have reported on the potential impact of acupuncture on the amygdala: upregulation of brain-derived neurotrophic factor (BDNF), tropomyosin-related kinase B (TrkB), phosphorylated mitogen-activated protein/extracellular signal-regulated kinase (p-MEK), phosphorylated extracellular signal-regulated kinase (p-ERK) expression in the amygdala [34], prevention of increase in tyrosine hydroxylase (TH) level and decrease in BDNF level in the amygdala [41], and improvement and repair of synaptic plasticity in the amygdala [36]. (3) Anterior cingulate cortex: In the anterior cingulate cortex (ACC), c-Fos expression is related to emotional regulation and attention and cognitive control [59,60], which are all related to major pathologies of PTSD. c-Fos is a gene involved in cell proliferation and differentiation after extracellular stimuli [61]. In ACC, which plays a role in regulating conditioned fear responses [62], c-FOS expression was significantly increased after EA [37].
(4) Prefrontal cortex: Endocannabinoids regulate various forms of synaptic plasticity in the adult brain, and enhancement of cannabinoid type 1 receptor (CB1R) is highly expressed in the forebrain [63]. A study found that EA could upregulate expressions of CB1R and diacylglycerol lipase α (DAGLα), which generate 2-arachidonoylglycerol (the most abundant ligand for the CB1R), in the PFC [35]. (5) Others: One study [46] indicated the regulatory effect of MA on the blood oxygen level in the cerebral cortex.

Discussion
In this review, previously investigated therapeutic mechanisms of acupuncture for various symptoms of PTSD were collected and analyzed from preclinical studies. Through comprehensive literature searches, 24 relevant animal studies  were included in the analysis. All but one study used the SPS and/or SPS & S model to establish the PTSD model, which is a representative model that mimics the pathology of human PTSD [64], and it is considered to faithfully reflect the PTSD criteria presented in the Diagnostic and Statistical Manual of Mental Disorders (DSM)-5 [16]. One study [30] that used the compound stress model of restraint, electric shock, and exhaustive swimming is also considered to reflect the DSM-5 criteria [16]. Therefore, PTSD models of all included animal studies can be considered reliable and potentially relevant to human outcomes. In these studies, the effects and/or underlying mechanisms of acupuncture on anxiety and fear behaviors, depression behavior, sleep disturbance, cognitive symptoms (especially spatial learning and memory), and stress responses in PTSD animal models have been reported ( Figure 3 and Table 2).    In behavioral evaluation, acupuncture, including MA and EA, reduced anxiety and fear responses and weakened fear conditioning, improved sleep architecture, reduced depressive symptoms, and alleviated the disturbance of spatial learning and memory in PTSD animal models. The therapeutic mechanisms of acupuncture proposed in the included studies could be classified into two categories: (1) regulation of stress responses in the neuroendocrine system and (2) promotion of neuroprotection, neurogenesis, and synaptic plasticity in several brain areas.

Regulation of Stress Responses in the Neuroendocrine System
In PTSD animal models, acupuncture modulated the reactivity of the HPA axis, including reduction of inflammation in the hypothalamus. The axis plays an extremely important role in stress management, and chronic stress or psychological trauma can lead to pathological consequences of this axis. Although the relevance of dysregulated HPA axis functionality in patients with PTSD has not yet been firmly proven [65], the HPA axis-modulating effect of acupuncture treatment may also gain more attention given that the HPA axis cannot be overlooked in the pathology of PTSD. Moreover, anti-inflammatory effects in the hypothalamus may be related to the observed improvement of sleep architecture and even memory consolidation. Disruption of sleep architecture is a common clinical manifestation of PTSD, and some sleep stages, especially non-REM slow wave sleep, play an important role in memory consolidation and may affect other symptoms of PTSD [66]. In addition, disturbed REMS or non-REMS may contribute to maladaptive stress and trauma responses [67]. Although there is limited certainty, there is some evidence that acupuncture can improve sleep quality in patients with insomnia [68]. In a recent randomized controlled trial [69], acupuncture significantly decreased non-REMS stages 1 and 2, increased non-REMS stage 3, and improved episodic memory in patients with chronic insomnia disorder. Given these results, acupuncture may improve sleep quality for PTSD and indirectly affect other symptoms of PTSD, such as memory consolidation. Neuroinflammation is a promising research field in PTSD [70], and some studies have reported that acupuncture reduces neuroinflammation in areas of the brain that play a key role in sleep regulation, such as the hypothalamus [42,49], but further studies are needed. As other aspects of the stress response, acupuncture downregulated nNOS expression in the locus coeruleus and hippocampus, which are associated with stress responses, upregulation of hippocampal mineralocorticoid receptor expression, and downregulation of hippocampal glucocorticoid receptor expression. These results suggest the role of acupuncture in the regulation of stress responses by neuroendocrine regulation.

Promotion of Neuroprotection, Neurogenesis, and Synaptic Plasticity in Several Brain Areas
Acupuncture has been reported to promote neuroprotection, neurogenesis, and synaptic plasticity in several areas of the brain in the PTSD model. In the pathogenesis of PTSD, BDNF has been suggested to be responsible for fear learning by controlling fear circuit plasticity [71]. The main areas associated with anxiety and fear responses in this hypothesis include the hippocampus, PFC, and amygdala [71]. In the fear circuit, the hippocampus is thought to play an important role in the explicit memory of traumatic events and mediate learned responses to contextual cues; the PFC reactivates past emotional associations; and the amygdala is activated in subjects with PTSD and associated with oversensitivity to stress, generalized fear responses, and impaired extinction [58]. Among these, the hippocampus and amygdala play a key role in conditioned fear and associative learning [58]. According to our findings, acupuncture is thought to modulate neuroplasticity in several brain regions, including hippocampus, amygdala, ACC, and PFC. Particularly, effects on the hippocampus have been reported most often, suggesting that acupuncture may have the potential to reduce anxiety and fear symptoms by controlling the fear circuit plasticity of PTSD, which is related to conditioned fear and associative learning clinically.

Comparison with Previous Clinical Studies
The findings of this review suggest promising potentials of acupuncture for treatment of PTSD at the preclinical level. However, reproducibility in clinical studies involving human PTSD subjects is considered highly insufficient. For example, a systematic review that evaluated the effectiveness of acupuncture on PTSD, including seven randomized controlled trials of 3-12 weeks, not only pointed out extremely low quality of the body of evidence (QoE) but also no significant difference on depressive symptoms, anxiety symptoms, and sleep quality between the acupuncture and control groups, at post-treatment [11]. However, although the QoE was low, significant improvement in PTSD and depressive symptoms was observed in longer follow-up (1 to 6 months post-intervention) [11]. There are also differences arising from the limitations of animal studies. For example, a systematic review was published in 2017 reporting the effectiveness of emotional freedom techniques (EFT) using cognitive and exposure therapies simultaneously with manual stimulation of acupoints for PTSD [72]. At the same time, a study published in 2018 concluded that manual stimulation of acupoints during EFT could not be explained by just placebo or nonspecific effects [73]. Thus, EFT includes aspects of cognitive therapy that are difficult to experiment in animal experiments, and stimulation of acupoints is thought to have a unique therapeutic effect. Moreover, although the level of evidence is very insufficient [74], factors such as patient-practitioner interactions are likely to influence the clinical outcome of acupuncture [75]. Therefore, the findings of this review may not fully support several acupoint-based interventions for human patients with PTSD.

Strengths and Limitations of this Review
Acupuncture is a nonpharmacological intervention that can be useful in the clinical management of PTSD, especially in situations where medical resources are prone to lack, including large-scale PTSD events such as disasters. However, since acupuncture, especially EA, is generally considered a time-consuming treatment, a simpler acupuncture type, such as ear acupuncture, could be considered in large-scale PTSD events [14]. Our results show that this intervention has some preclinical evidence of the therapeutic mechanism, suggesting the potential to improve different clinical aspects of PTSD. Additionally, although not the outcome of interest in this review, other clinical effects of acupuncture, such as pain control, which may be important in PTSD patients, are well known. Therefore, the results of this study may help elucidate the possibility of using acupuncture for PTSD in terms of evidence-based medicine. However, the following limitations should be considered in this review. First, this review presents results based on animal studies, and it is not possible to directly correlate these results with therapeutic mechanisms of acupuncture in human PTSD patients. To our knowledge, few studies have investigated the therapeutic mechanisms of acupuncture in human PTSD subjects [13]. Therefore, the results of this review could be used as data for designing a future acupuncture mechanism study in PTSD patients. Second, although acupuncture was largely classified into EA and MA in this review, the difference in treatment sites, called acupoints, was not considered. Analysis of the frequency has provided some treatment site data, such as Baihui (GV 20), that may be important in future studies, so future studies will be able to investigate acupoint-specific effects on the main acupoints. Third, the methodological quality of the included animal studies was not high enough to produce robust evidence. The included studies were evaluated as unclear in most SYRCLE risk of bias items, and insufficient reporting was suggested as a major problem. Fourth, in the treatment of mental disorders, acupuncture is composed of a treatment procedure that includes a patient-practitioner interaction context, so some other effects, including expectancy effects, cannot be ignored [76,77]. Such effects of acupuncture are difficult to elucidate in animal studies; that is, it is suggested that there may be other therapeutic mechanisms that are not covered by the results of this review. Lastly, this systematic review does not have a preregistered protocol, reflecting a potential for reporting bias.

Future Direction
Considering the abovementioned limitations, future research in this field will be able to consider the following. First, research is needed to investigate the therapeutic mechanism of acupuncture on PTSD in humans. Ideally, tools, such as functional magnetic resonance imaging or functional near-infrared spectroscopy, can be used to investigate the therapeutic mechanism of acupuncture in terms of brain science. Moreover, in a few clinical studies, such an attempt has already been conducted [78]. The findings of this review may be used as a reference in the process. Second, by conducting clinical studies on some important acupoints proposed in the findings of this review, it may be possible to study acupoint-specific mechanisms or standardize the acupuncture method for treating PTSD. For example, scalp acupoints, such as Baihui (GV 20) and Shenting (GV 24), have traditionally been used to treat psychiatric disorders, and EA on Baihui (GV 20) and Shenting (GV 24) and other acupoints present in the head or neck, Shenting (GV 24), and Fengchi (GB 20), have been used in a clinical study to compare the effectiveness with that of paroxetine in the treatment of PTSD [79]. These acupoints may be promising in future research in this field. Third, it was found that most of the included studies were conducted in China. Although these studies were conducted on animal models, acupuncture methods, such as acupoint selection or stimulation method, in the study design stage may be influenced by the researchers. Thus, to reflect the acupuncture method uniquely used in countries other than China, it is necessary to conduct acupuncture studies on PTSD animal models in other countries as well. In addition, animal studies of acupuncture on PTSD need to be further improved, as current animal studies are insufficient to provide robust evidence on underlying mechanisms of acupuncture for PTSD.

Conclusions
This review summarizes the complex underlying therapeutic mechanisms of acupuncture, including MA and EA, in PTSD animal models. In behavioral evaluation, acupuncture, including MA and EA, reduced anxiety and fear responses and weakened fear conditioning, improved sleep architecture, reduced depressive symptoms, and alleviated the disturbance of spatial learning and memory of the PTSD animal models. The therapeutic mechanisms of acupuncture proposed in the included studies could be classified into two categories: (1) regulation of stress responses in the neuroendocrine system and (2) promotion of neuroprotection, neurogenesis, and synaptic plasticity in several brain areas. However, the methodological quality of the included animal studies was not high enough to produce robust evidence. In addition, mechanistic studies on specific aspects of acupuncture that may affect PTSD, including expectancy effects, in human PTSD subjects are also needed.

Data Availability Statement:
The data used to support the findings of this study are included in the article.

Conflicts of Interest:
The authors declare no conflict of interest. Identify the report as a systematic review, meta-analysis, or both. 1

Structured summary 2
Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and implications of key findings; systematic review registration number.

Rationale 3
Describe the rationale for the review in the context of what is already known. 2

Objectives 4
Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons, outcomes, and study design (PICOS).

Protocol and registration 5
Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide registration information including registration number.

Eligibility criteria 6
Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered, language, publication status) used as criteria for eligibility, giving rationale. 3

Information sources 7
Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify additional studies) in the search and date last searched.
3 Search 8 Present full electronic search strategy for at least one database, including any limits used, such that it could be repeated. Appendix A

Study selection 9
State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable, included in the meta-analysis). 3

Data collection process 10
Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes for obtaining and confirming data from investigators. 3

Data items 11
List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and simplifications made. 3

Risk of bias in individual studies 12
Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis. 3

Summary measures 13
State the principal summary measures (e.g., risk ratio, difference in means).

3-4 Synthesis of results 14
Describe the methods of handling data and combining results of studies, if done, including measures of consistency (e.g., I 2 ) for each meta-analysis.

Risk of bias across studies 15
Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., publication bias, selective reporting within studies). NA

Additional analyses 16
Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-regression), if done, indicating which were pre-specified. NA

Study selection 17
Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions at each stage, ideally with a flow diagram. 4, Figure 1 Study characteristics 18 For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up period) and provide the citations. 4-5, Table 1, Table 2 Risk of bias within studies 19 Present data on risk of bias of each study and, if available, any outcome level assessment (see item 12). 13, Figure 2

Results of individual studies 20
For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for each intervention group (b) effect estimates and confidence intervals, ideally with a forest plot. 14-16 Synthesis of results 21 Present results of each meta-analysis done, including confidence intervals and measures of consistency. NA

Risk of bias across studies 22
Present results of any assessment of risk of bias across studies (see Item 15). NA

Additional analysis 23
Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta-regression [see Item 16]). NA

Summary of evidence 24
Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (e.g., healthcare providers, users, and policy makers).

16-19
Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of identified research, reporting bias).