The control of body weight through regulation of energy balance relies on a complex neuroendocrine interplay of integrating peripheral metabolic signals by the central nervous system (CNS) controlling feeding behaviour. Energy balance and food intake are controlled through a coordinated feedback system of neuropeptides and neurotransmitters in the CNS and peripheral signals arising from the gastrointestinal cells in the stomach, adipose tissue, pancreas, intestine as well as nutrients and microbiome (11, 13, 54). See Figure 1 for this dynamic interplay in the pathogenesis of obesity.

A state of chronic low-grade systemic inflammation and immune dysfunction is involved in the pathogenesis of obesity and obesity-related metabolic disorders, like MetS and T2DM (90). Adipose tissue is not only a fat storage organ, but also an important neuroendocrine organ; besides energy storage and expenditure it performs essential functions in metabolic regulation (91). In obesity, adipocyte hypertrophy, hypoxia, mechanical stress, adipocyte apoptosis, and oxidative stress together initiate an inflammatory response with infiltration of immune cells like macrophages (2). This infiltration is associated with a cell population shift from anti-inflammatory M2 macrophages to pro-inflammatory M1 macrophages, and expression of pro-inflammatory cytokines. These cytokines like Tumor Necrosis Factor α (TNFα) and Interleukin 6 (IL-6) and adipokines (leptin and resistin) then spill over into the circulation and contribute to systematic low grade chronic inflammation and dysfunction of the adipose tissue (2, 92). Pro-inflammatory adipokines secretion and decreased release of anti-inflammatory adipokines contribute to obesity-induced inflammation and promotes systemic inflammation and IR (91–93). IL-6 has a controversial role as its action varies in different cells and tissues. It can both exert a proinflammatory as well as an anti-inflammatory action (94). An acute increase due to fasting or exercise mobilizes lipids, but in contrary to this physiological elevated IL-6 level a chronic increase triggers proinflammatory action and leads to pathological conditions like dysregulation of the glucose metabolism, hepatic steatosis and insulin resistance (94).

In addition to the pro-inflammatory milieu in the adipose tissue, research has also revealed obesity related inflammatory changes perturbing brain function, especially affecting brain areas like the hypothalamus-controlling neuroendocrine functions that integrate and regulate energy homeostasis and systemic metabolism. Adipocyte hypertrophy and hyperplasia induce a state of systemic chronic low-grade systematic inflammation resulting in activation of brain immune cells such as astrocytes, microglia, and oligodendroglia. This neuroinflammation further facilitates weight gain and obesity-associated IR (95, 96). Quantitative magnetic resonance imaging (qMRI), aimed at quantifying brain water content, has recently emerged as a tool to characterise the pathophysiological process of inflammation in the brain. The hypothalamus displays the most prominent alterations in water content, but other surrounding brain regions also show signs of brain inflammation (95–97). This neuroimmune cross-talk is crucial for the understanding and pathogenesis of obesity and developing innovative therapeutic strategies. In conclusion, the intersection between the nervous system, immune system and metabolism is an important factor in the context of obesity and has recently been defined as neuro-immuno-metabolism. As mentioned earlier, in obesity a complex interactive multi-directional relationship exists between the adipose tissue, the sympathetic nervous system (SNS) and innate immune cells (72, 98), as shown in Figure 1 . The scheme is completed by a missing link: the gut microbiota. Below, the potential causal effect of a gut microbiota disbalance on the pathogenesis of human obesity is explained.

Recent research has revealed the role of the host gut microbiome in the development of obesity and its influence on host metabolism (99). The gut microbiome consists of micro-organisms present in the human gastrointestinal tract which have co-evolved into a complex system, contributing to the health status of their host (54, 100, 101). The gut microbiota comprise of a multitude of bacterial taxa, of which Firmicutes and Bacteroidetes are the most abundant phyla, accounting for over 90 per cent of the bacterial community in the large intestine (102). Normal gut microbiota exert several functions including protection against pathogens, increasing dietary energy harvest through their metabolism of carbohydrates and lipids, as well as directing energy utilisation through generation of metabolic substrates accessible to different tissues (103). Moreover, the gut microbiota affect the immune system and play an essential role in regulating the structural integrity of the gut mucosal barrier (104). Dysbiosis, described as an imbalance of microbial populations, is implicated in the pathogenesis of a large range of diseases such as inflammatory bowel disease, irritable bowel disease, neurological disorders, and cancer (105). Diet induced disruptions of the gut microbiome promote white adipose tissue inflammation and expansion, and this microbiome dysbiosis is critical in developing T2DM, MetS and obesity (105).

Indeed, in addition to its well-described contribution to digestion of food, gut microbiota composition is thought to critically influence host health and disease in the context of the gut brain axis (106, 107), where dysbiosis with concomitant increase in pathogens can lead to inflammation and disease (88, 108, 109). The regulatory system linking the CNS and enteric nervous system (ENS) to the peripheral intestinal functions is referred to as the “gut brain axis”. This complex dynamic bidirectional system includes the parasympathetic and SNS (110, 111), immune and endocrine factors as well as the gut microbiota, all regulating gastrointestinal homeostasis (110–114). The vagal nerve, as main component of the parasympathetic nervous system, is important in the regulation of the gastrointestinal system (112), exerting both excitatory as well as inhibitory control over the gastrointestinal tract (115). The role of the SNS is mainly inhibitory (111) and it is likely that its activation via release of stress-hormones such as epinephrine, norepinephrine and cortisol, contributes to a dysregulation of gut motility (54, 116).

In the context of metabolic diseases such as obesity, the gut brain axis integrates cerebral and gastrointestinal functions in a bidirectional manner, including gut motility, appetite, and body weight control (108). From the gut to the brain, the secretion of metabolites by the gut microbiota promotes the release of intestinal peptide hormones and neurotransmitters, such as PYY, GLP-1, 5-hydroxytryptamine (5-HT) and gamma-aminobutyric acid (GABA) (117–119). In addition to contributing to regulating peripheral metabolic status (108), these metabolites are thought to influence the CNS via the ENS (120). Via the hypothalamic ARC, they can control the balance of appetite (108, 119, 121, 122). Although the role of 5-HT in obesity has not yet been fully clarified, changes in central and peripheral levels of 5-HT have been implicated in obesity as mediators of feeding behavior and feeling of satiety (119, 122–124).

Pathways in the gut brain axis include alterations in secretion of gastrointestinal fluids, gut motility, modulating gut permeability and mucosal immune response (111). The gut microbiota is affected both directly and indirectly by neuroendocrine efferent systems comprising of the CNS and the HPA-axis (111, 121). In the context of metabolic disorders, the nervous system and HPA axis are thought to influence the gut microbiota composition, specifically the Firmicutes-to-Bacteroidetes ratio, which is of potential relevance in the development of obesity (88, 125). Initial key research in obese mice has shown a lower gut microbiota diversity and an increased Firmicutes-to-Bacteroidetes ratio compared to lean counterparts (126), and germ-free mice colonised with obesity-associated gut microbiota developed an associated phenotype (125). In human adult twins, obesity was associated with a lower Bacteroidetes abundance, and fecal transplantation thereof into mice revealed further transmissible modifying effects of human gut microbiota on obesity phenotype (127). However, although promising, this microbiota profile does not translate as such in clinical studies in adults with obesity (128). A systematic review of six placebo controlled RCTs reported that fecal microbiota transplantations (FMT) from lean donors into adults with obesity show no improvement of parameters including hepatic insulin sensitivity, BMI, fasting plasma glucose, or cholesterol level, other than a reduction of HbA1c levels at six weeks post-FMT (129). Thus, the regulation of energy metabolism of the host is influenced by many metabolites produced and modified by the gut microbiota as part of the gut brain axis (128). However, this complex interaction is not yet fully explained. Moreover, inherent to the complexity of this bi-directional system, it is likely only one of several ways in which gut microbiota contribute to host health in the context of the gut brain axis. Nevertheless, demonstrating whether specific gut microbiota-based interventions have beneficial effects on weight management and subsequently improved cardiometabolic risk remains a challenge (109, 128). Thus, other interventions should be taken into account.

Multiple pathways have been studied to understand the underlying mechanism of acupuncture to counter obesity. Evidence showed that acupuncture regulates the secretion of related biomolecules (lipid hormones, neuromodulators, peptide hormones, and neurotransmitters), modulates inflammation, suppresses appetite, regulates lipid metabolism, and promotes browning of white adipose tissue (37, 133, 145).

As discussed before, the immune system and chronic low-grade inflammation play an essential role in the pathophysiology and development of obesity (163). Mounting evidence has gradually recognised that acupuncture modulates and regulates the immune system in multiple systems and multiple diseases. This non-invasive technique has emerged as a promising treatment for immunomodulation (133, 135, 145, 164–166). In murine models, general anti-inflammatory effects of acupuncture involve the regulation of multiple types and functions of the innate immune system, including macrophages, mast cells and granulocytes, as well as the adaptive immune system, such as lymphocytes. Several animal studies show that the pro-inflammatory M1 macrophages are downregulated and the anti-inflammatory M2 macrophages are upregulated through MA and EA on ST36 (167). This leads to the associated release of anti-inflammatory cytokines like TNF-β, IL-10, and the inhibition of the expression of pro-inflammatory cytokines such as TNF-α, IL-6, and IL-1, and promotes the expression of anti-inflammatory tissue repair factors. Apart from macrophages, acupuncture also regulates mast cells and the quantity of neutrophils and increases the activity of NK cells. Acupuncture controls the adaptive immune system by regulating the lymphocyte Th-cell balance as well as adaptive immune cytokines to initiate the immune response and repair damage (41, 164, 167–169). Acupuncture also seems to regulate oxidative stress, which is associated with chronic inflammation in the adipose tissue. In a recent metanalysis of 12 animal studies, acupuncture appears to reduce oxidative stress through the activation of the antioxidant enzyme system and lowering of lipid peroxidation (170). Another rat model shows that acupuncture, treatment of acupoints GB34, LR3, ST36 and SP10, exerts an antioxidant action and showed neuroprotective properties (171).

Moreover, the hypothalamus is crucial for both innate and adaptive immunity. Acupuncture has also shown to be effective to reduce neuro-inflammation through the inhibition of the activation of glial cells (astrocytes, microglia and oligodendrocytes). More specifically for obesity and obesity metabolic related disorders like T2DM, acupuncture regulates several signaling pathways affecting the production of inflammatory cytokines to inhibit inflammation. In a review of Li et al., several animal studies report the effect of acupuncture on DIO rats. The most used acupoints were ST36, ST37, ST39, SP6, LI4, LR3, CV4, KI1, ST40, GB34 (164). Among other signaling pathways, it downregulates the pro-inflammatory factor TNF-α and inhibits the macrophage proliferation and infiltration of macrophages into adipose tissue and, in turn, stimulates the anti-inflammatory adipokine balance lowering the adiponectin/leptin ratio. In addition, the chronic low-grade inflammation is reduced, and insulin sensitivity and glucose tolerance improved, and blood lipid content lowered (164).

The CNS plays a pivotal role in integration of acupuncture-driven information. Several main regulatory pathways can be distinguished through which acupuncture exerts its anti-inflammatory effect: cholinergic anti-inflammatory pathway (CAIP), vagal-adrenal medulla-dopamine pathway, spinal sympathetic pathway, HPA-axis, and the brain-gut axis (BGA). First, looking at the cholinergic anti-inflammatory pathway (CAIP), upon adrenalectomy and vagotomy, the anti-inflammatory effect of EA was blocked, implying the anti-inflammatory effect of the activation of the CAIP. Second, in relation to the vagal-adrenal medulla-dopamine pathway, EA can modulate body-physiology through somatosensory autonomic reflexes. These reflexes start with the stimulation of peripheral nerves followed by transmission of sensory information to the brain and subsequent activation of autonomic pathways such as inhibiting systemic inflammation (164, 172). A High-Fat-Diet (HFD) induced rat model examined the relation between the autonomic nervous system and the anti-inflammatory effect of EA treatment of acupoint ST36. Here it was concluded that EA inhibits low-grade inflammation and this mechanism was closely related to the cholinergic anti-inflammatory pathway as well as enhancement of the vagal activity (42). It has been demonstrated that through acupuncture signals the vagal-adrenal medulla reflex is being triggered and dopamine is being released, exerting an anti-inflammatory effect. Neurons innervating the deep hindlimb fascia, the so-called PROKR2^Cre-marked sensory neurons, are fundamental for driving the vagal-adrenal axis (173). Experimental studies show that low-frequency EA of acupoint ST36 stimulates dopamine secretion, modulating vagal activity and thereby exerting an anti-inflammatory effect (174). Catecholamines from the adrenal gland are released by acupuncture stimulation and by acting on peripheral dopamine D1 receptors produce systemic anti-inflammatory effects (175). Moreover, the spinal sympathetic pathway plays an important role because its activation through acupuncture has an anti-inflammatory action. After sympathectomy the anti-inflammatory effect of acupuncture significantly reduces, also indicating a role of the latter pathway. Furthermore, it has been shown that acupuncture bi-directionally regulates the HPA-axis restoring the normal anti-inflammatory action. One of the possible explanations is that acupuncture exerts its anti-inflammatory effect through the regulation of intestinal flora, which is further explored below (135, 164, 168). Lastly, acupuncture exerts an effect on the bi-directional brain-gut axis (BGA). A somatosensory-autonomic reflex pathway controls intestinal inflammation, thereby restoring BGA balance. In addition, the changes in the gut microbiota are transmitted to the brain via the sympathetic nerves or the vagus nerve (135). This is also further reviewed below. In conclusion, accumulated evidence shows that acupuncture exerts a strong anti-inflammatory effect and reduces chronic low inflammation as shown in Figure 2 , and human studies are needed to validate these findings.

Although limited, some animal studies have suggested a role for acupuncture via regulation of the ENS as part of the gut-brain axis. The SNS, influencing the ENS, exerts a mainly inhibitory role over the gastro-intestinal tract (115) and it is likely that its activation via release of stress-hormones such as epinephrine, norepinephrine and cortisol, contributes to a dysregulation of gut motility (116). Among the peripheral gastrointestinal hormones, 5-HT is an important neurotransmitter in the gut brain axis (117). Although the function of 5-HT has not yet been fully explained, several studies describe a role on secretory and peristaltic reflexes, by stimulating enteric neurons (110, 117). A study in mice suggests improved colonic motility after treatment with EA via regulation of the ENS, through alteration of both excitatory and inhibitory enteric neurons (183). In two rat models with diarrhea-prominent irritable bowel syndrome (IBS-D), EA at ST25, ST36 and LR3 alleviated IBS-D symptoms and lowered the levels of NPY and 5-HT in the gut brain axis, suggesting restored balance of the gut brain axis (184). Moreover, in adult patients with constipation dominant-IBS (IBS-C), an RCT comparing EA to mild-warm moxibustion treatment at ST25 and ST37 showed improvement of defecation frequency compared to baseline up to three months after treatment, and the effects of EA were proposed to be mediated through alleviation of visceral hypersensitivity and improved brain-gut function as visualised with fMRI (185). Additionally, several studies in animals and humans show altered levels of stress hormones such as cortisol and epinephrine, and changes in 5-HT associated with acupuncture, although mostly in the context of visceral pain management (186–189). In mice and rats with IBS-associated visceral hypersensitivity, treatment with EA significantly reduced 5-HT and 5-HT receptor mRNA expression levels compared to sham-treatment and was associated with attenuated visceral hypersensitivity (186, 187). In the context of obesity, a potential role for 5-HT on thermogenesis of brown adipose tissue in mice has been reported (189) and gut microbiota are directly implicated in the process of peripheral 5-HT biosynthesis in their host, both animals and human adults with obese phenotypes (122, 190, 191). However, the clinical relevance and changes in these hormonal levels related to acupuncture treatment have not (yet) been studied in the context of acupuncture treatment in mice or humans with obesity.

In human trials acupuncture emerges as a potentially effective complementary treatment of obesity as it is thought to induce loss of appetite, down-regulates insulin and leptin resistance, and lowers ghrelin as well as glucose levels (31, 32, 192). A systematic review of acupuncture in adults with obesity, which reviewed 11 sham-controlled RCTs published in English, suggests that acupuncture, including EA, ear acupuncture and MA, was indeed effective in reducing BMI, hip- and waist circumference, body fat mass, but not body weight. The most used acupoints are LI4, LI11, ST25, ST36, ST44, GB28, CV4, CV9, CV12, SP6, LV3 (193). Due to the limited sample size, heterogeneity and varying methodologic quality, further assessments are necessary. Acupuncture treatment of acupoints LI4, ST36, ST44, SP6, ST25, GB28, CV4, CV6, CV12, SP6, LI11, ST40, SP9 alone or combined with low-calorie diet had beneficial effects on serum leptin levels (52, 194). Due to methodological limitations, lack of homogeneity of the included studies and small effect sizes, these findings should be taken with caution. Park et al. included eight RCTs in their systematic review and concluded that a combination of acupuncture and diet therapy or exercise was more effective than diet and exercise alone in reducing serum leptin levels. Between verum acupuncture and sham acupuncture, however, there was no significant difference, although compared to oral anorexiant therapy or no treatment, acupuncture was proven to be more effective. Again, due to methodological limitations, the results should be used with care (32). In a systematic review and meta-analysis of 20 RCTs Wu et al. showed that acupuncture could be a viable option in treating IR by reducing fasting insulin (FINS) levels, fasting blood glucose (FBG), and 2h postprandial blood glucose (2hPG). Methodological flaws, however, limit these conclusions (47, 71). A small clinical study of 16 obese women showed that EA treatment of acupoints BL18, BL20, BL21, BL23, LI10, LI11, ST36 was effective in reducing glucose level, whereas auricular acupuncture did not (195). This finding is in line with the animal models where EA was shown to act on the glucose metabolism. Again, this study has limitations as it was an uncontrolled pilot study. On the other hand, a systematic review of 33 RCTs performed by Chen et al. concludes that the effects of acupuncture on blood glucose are uncertain. They did report a significant, albeit small reduction of body weight, BMI and waist circumference compared to the non-acupuncture groups. Furthermore, reduction of serum lipid parameters has been found (48). The evidence is of moderate to very low quality and well-designed large trials with long-term follow up are needed.

Although there is substantial animal evidence, human trials studying the effects of acupuncture and chronic low-grade inflammation in obesity are limited. Acupuncture has shown to be effective in several inflammatory disorders, like inflammatory bowel disease (IBD), effectively regulating the inflammatory factors (196). An RCT reviewing EA performed in obese women noted no difference in the high-sensitivity C-reactive protein (hs-CRP) levels, but positive effects in the anti-heat shock protein (anti-Hsp), concluding that acupuncture exerts immunomodulatory effects on the human immune system but not anti-inflammatory effects. In a sham-controlled RCT of 196 obese subjects a significant decrease in preoxidant-antioxidant balance (PAB) was found in the treatment group (acupuncture combined with low-calorie diet) compared to the control group (diet alone). Here the EA treatment used acupoints ST25, GB28, CV6, CV9, CV12, SP6, LI11, ST40, SP6, SP9. This is in line with animal studies showing that oxidative stress is reduced by acupuncture (197).

As described previously, some evidence in animal models of acupuncture on obesity and modulation of gut microbiota composition is presented. In humans, the evidence is still scarce but emerging changes in species abundance and diversity of the gut microbiome have been documented in patients undergoing treatment with EA (49, 53, 198). In perimenopausal women with abdominal obesity, EA for eight weeks in combination with diet decreased the abundance of the species and increased the species diversity of gut microbiota (49). After EA treatment, notably the two species Klebsiella and Kosakonia were increased in perimenopausal patients (49). In adult studies using acupuncture for various conditions, emerging evidence points towards an effect of acupuncture on the gut microbiome via the gut brain axis. A recent sham-controlled study in adults with antipsychotic-related constipation treated with EA showed that richness of various bacterial strains such as Klebsiella and Gammaproteobacteria was significantly higher after acupuncture, and was associated with significantly increased spontaneous bowel movements and reduced use of rescue medication. Moreover, Firmicutes-to-Bacteroidetes ratio was increased post-treatment (198). In a population of patients with knee osteoarthritis, eight weeks of EA modified gut microbiota diversity and microbiota richness of species (53). In healthy adults, a recent study on acupoint application treatment - a non-invasive modality of traditional acupuncture - showed increased Firmicutes/Bacteroidetes ratio after three sessions in 24 months compared to sham control group (199). It should be noted that true clinical consequences of the changes in gut microbiota in obese patients are to date not established in humans. As described previously in the section “Microbiota dysbiosis and gut-brain axis”, changes in diversity and richness of the microbiome have been linked to many significant biological changes with consequences for human health (106, 107); in the context of obesity, a dysbiotic Bacteroidetes/Firmicutes ratio is frequently reported (107, 125). However, the clinical relevance of the currently presented changes in microbiota profiles after treatment remain to be evaluated. Indeed, this “signature” dysbiosis is mainly shown in rodent models, as studies in humans report varying results regarding microbiota profiles (128, 129, 200), which could be the result of methodological differences between studies and differences in gut microbiota sample processing (99). The varying duration of interventions, use of different acupoints, and the variety of sex and age distribution potentially impact the results. In acupuncture studies, including microbiota profiles in animal models, mainly male rodent models are used. In the few studies in adults, sex distribution differs between studies, ranging from women only (49) to two-thirds of the population subgroups being women (53, 198, 199). Recent studies reported that changes in the gut microbiota associated to metabolic illnesses are different in men and women (201) and this could be translated to acupuncture treatment (50). To the best of our knowledge, no studies analysing acupuncture in combination with gut microbiota analysis in children have been performed to date. This evidence combined, although limited, suggests that acupuncture may affect constituents of the brain-gut axis, including the enteric and autonomic nervous system and the gut microbiota, and this potential mechanism of action should be investigated further when evaluating acupuncture in patients with metabolic diseases.

An adequate strategy to reduce weight is a Very Low-Calorie Diet (VLCD) (202). As mentioned before, however, losing weight is not as easy as it seems as adherence to diet is low and calorie restriction might increase cortisol levels and perceived stress (203). As previously discussed, acupuncture seems to reduce appetite through its neuro-endocrine regulation and reduces cortisol levels and stress. However, acupuncture alone seems not to be clinically effective on body weight loss (33). Therefore, a combined intervention of VLCD and acupuncture is proposed. In line with previous research (31, 32, 48, 192) acupuncture combined with lifestyle modification (diet alone or diet in combination with physical exercise) showed greater reduction of body weight than acupuncture alone. Also, greater serum leptin levels reduction was found when combining acupuncture with diet (32). A synergetic effect between lifestyle modifications and acupuncture is therefore suggested and this combined intervention may be a more effective and viable option. Recently, a large-scale retrospective chart review was performed of 11,233 obese patients with combined acupuncture and VLCD underscored that the combination can induce weight reduction in obese patients (50). However, due to the lack of a control group, a prospective trial is needed to validate these findings and therefore such RCT is currently executed.

Outcomes of systematic reviews on the effectiveness and safety of acupuncture both in adults and in children support the notion that acupuncture is well-tolerated and can be considered a safe intervention (48, 139, 192, 204, 205, 131). The frequency of reported adverse events (AEs) associated with adult and pediatric needle acupuncture is low and most AEs are mild and transient. Reported AEs include mild ecchymosis, bleeding, pain upon needle insertion,dizziness caused by acupuncture and EA and mild skin burns caused by moxibustion (48, 139, 204, 205). Another noteworthy factor when considering acupuncture for obesity is the potential cost-effectiveness. Worldwide, obesity represents a significant economic burden in developed as well as developing countries (206). Although a cost-effectiveness analysis has not yet been performed in the context of obesity, several cost-effectiveness analyses of acupuncture versus usual care for varying conditions (dysmenorrhea, allergic rhinits, osteoarthritis, headache and acute lower back pain) show a benefit for this treatment modality (207, 208). Its cost-effectiveness should be further evaluated in the specific case of obesity, especially considering the current treatment options ranging from dietary changes and prescription medication to bariatric surgery.

In conclusion, acupuncture might be an effective, safe and cost-effective method for obesity. However, the results remain controversial and conclusions should be taken with caution. There are several limitations in the studies regarding presented obese animal models. First, the treated acupoints differed between studies and an explanation for the use of these points was not at all or only partially provided in most studies, basing it on a proposed transposition model of acupoints in rodents and referring to the points used in a human trial by who based the points on expert opinion. Different acupoints might regulate different biological pathways and a combination of points may activate several pathways simultaneously and standardization is lacking. Furthermore, frequencies of EA, durations of the treatment and wave types varied as well as different moxibustion techniques were applied. Among others, the most used acupoints in both animal and human studies in the context of obesity are ST25, ST36, CV9, CV12, CV4, SP6, ST44 (146). The exact mechanism of the single actions of the points and the combination of points, however, must be further elucidated. The myriad of studies thus has individual variabilities which makes it difficult to evaluate systematic reviews or meta-analyses for their accuracy and applicability to clinical practice. With regards to human studies, the quality of most included RCTs is low due to high heterogeneity of interventions (MA, EA, auricular acupuncture, moxibustion) different points were used and studies used with and without diet and/or exercise. Also, inclusion criteria (gender, age, BMI) and studied endpoints differed widely. Most studies that have been reviewed are low in sample size, have a short follow-up and overall high risk of bias. Moreover, there is a lack of consensus of general accepted criteria and international accepted protocols regarding the acupoint selection and technique, caused by a strong doctor variability, and existence of numerous acupuncture traditions. There is also a publication-geographical bias as many RCTs were conducted in China and the question remains whether these findings may be extrapolated to the rest of the world. Due to these methodological flaws the real effects may be distorted and quality of the evidence is weak preventing us from drawing definitive conclusions. Lastly, there is a continuous academic discussion about the use of sham-acupuncture for placebo-controlled studies. Acupuncture is often criticised for having a placebo effect. Compared to “real acupuncture”, sham techniques might exert similar biological effects and its use should be reconsidered (209). These limitations make acupuncture an avenue for treatment but do yet not support mainstream treatment until further rigorous research is conducted in this field. Future research should try to overcome these methodological limitations and therefore well-designed rigorous RCTs and multi-centre trials with large sample size and protocolized standard treatments are needed to draw evidence-based conclusions about the beneficial effect of acupuncture as a treatment for obesity. Furthermore, the understanding of acupuncture’s action mechanism may be further researched thanks to the advancements in systems biology, modern omics technology and emerging techniques to analyse biological information such as tracer technology, i-needles, metabolomics, two-photon technology, and cryo-electron microscopy technology. This will enable acupuncture’s action on the neuro-endocrine-immune network to be elucidated (133, 175). This progress in biomedical research combined with future clinical trial should shed greater light on acupuncture as a viable treatment option for obesity. Lastly, research should also be focused on the accessibility and equity of acupuncture practice. For multiple reasons acupuncture implementation has been limited due to fact of the limitations in the insurance reimbursement as well as the fear for needles as a potential barrier (210).