Introduction
In the current world population, some disorders like liver diseases, chronic obstructive pulmonary disease, cancer, diabetes, and CVDs are the leading causes of mortality. The prevalence of these disorders is closely correlated with lifestyle choices like physical activity, diet, smoking, and alcohol use(Reference Nakamura and Omaye1–Reference Williams4). Alanine transaminase (ALT) and aspartate transaminase (AST) are two important enzymes in diagnosing liver diseases. An increase in the level of these enzymes indicates liver damage. It has been demonstrated that variations in the amounts of these two enzymes have negative effects on one’s health(Reference Aragon and Younossi5,Reference Lee, Kim and Poterucha6) . On the other hand, the importance of glycaemic management for human health has been demonstrated in earlier research(Reference Klein and Klein7,Reference Wheeler, Dempsey and Grace8) . Garcinia cambogia (native to Southeast Asia) is one of the plants that has been shown to play an important role in human health. This plant has anti-inflammatory and antioxidant characteristics and is effective in the management of liver abnormalities and glycaemic control. A bioactive compound called hydroxycitric acid (HCA) is mainly responsible for these therapeutic effects(Reference John, Brown and Panchal9).
Previous studies on the use of Garcinia cambogia in the adult population have yielded conflicting results when examining metabolic indicators such as insulin, blood glucose, ALT, and AST. It seems that age, sex, BMI, and intervention duration were the main potential sources of heterogeneity. In one study, it was found that orlistat in combination with Garcinia cambogia or each alone did not have a significant effect on postprandial glucose and fasting blood glucose in obese men(Reference Al-Kuraishy and Al-Gareeb10). Other studies have also shown that Garcinia cambogia extract supplementation does not have a considerable effect on glycaemic control and liver transaminases in healthy individuals(Reference Vasques, Schneider, Klein, Falavigna, Piazza and Rossetto11–Reference Chong, Beah, Grube and Riede16). A study conducted on obese adults in Taiwan showed that the serum AST level in the group receiving only Garcinia cambogia significantly decreased, but insulin and glucose levels and serum ALT did not change significantly(Reference Lu, Yang and Wu17). On the contrary, Garcinia cambogia extract along with a calorie-restricted diet has led to increased liver enzymes and failure to control glycaemic parameters(Reference Arefhosseini, Tutunchi, Nomi-Golzar, Mahboob, Pouretedal and Ebrahimi-Mameghani18).
Therefore, considering the conflicting and heterogeneous results of previous studies, the objective of this study is to systematically review and meta-analyse the existing research on the impact of Garcinia cambogia supplementation on glycaemic control and liver enzymes in the adult population.
Methods
Search strategy
Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed when conducting the current systematic review and meta-analysis(Reference Moher, Shamseer and Clarke19) and were registered at PROSPERO (CRD42023428039). The systematic literature search of Scopus, Web of Science, PubMed/Medline, Google Scholar, and Cochrane Library databases was conducted up to 9 February 2023. In order to conduct our search technique, the following MeSH and non-MeSH phrases were combined: (‘Garcinia cambogia’ OR ‘Hydroxycitric Acid’ OR Hydroxycitrate OR brindleberry OR ‘Malabar tamarind’ OR kudampuli) AND (intervention OR RCT OR Randomly OR randomised OR random OR Placebo OR trials OR trial OR randomised). There was no time limit, and we included only English-language articles. Two independent authors (MRA, ST) reviewed the titles and abstracts to find related publications. Additionally, to guarantee that all potentially relevant papers were found, the reference lists of the included research and pertinent reviews were carefully searched.
Selection criteria
To determine eligibility, the population, intervention, comparison, outcome, and study design (PICOS) criteria(Reference Richardson, Wilson, Nishikawa and Hayward20) were used.
Studies were selected based on the following inclusion criteria: (1) designed as randomised controlled trials (RCT) (either parallel or crossover); (2) examined the effects of Garcinia cambogia in humans (either healthy or unhealthy subjects); (3) investigated fasting blood sugar (FBS), insulin, ALT, and AST levels for both treatment and control groups at baseline and at the end of the study; (4) involved adult individuals (aged ≥ 18 years); and (5) provided means and SDs of desired outcomes or any other effect sizes being convertible to means and SDs. We excluded articles with non-randomised study designs, animal studies, observational studies, in vitro research, papers without a placebo group, studies conducted on children or adolescents, conference abstracts, and reviews.
Data extraction
Two different researchers (MRA and ST) independently extracted the pertinent information from each included article using common data extraction forms, such as the name of the first author, the year that the work was published, the nation, the study’s design, the participant’s health status, their gender, the sample size, their average age and BMI, the duration of the intervention, the dosage of Garcinia cambogia supplementation (and its equivalent in HCA), the comparison group, and the averages and standard deviations of their FBS, insulin, ALT, and AST levels both before and after the intervention for the treatment and control groups. Our primary outcome was FBS levels. The secondary outcomes were insulin, ALT, and AST levels. In studies where relevant information was not provided, the respective authors were emailed to ask for more details.
Quality assessment
Two reviewers (MRA and ST) independently assessed the methodological quality of the chosen papers using the Cochrane risk-of-bias tool for randomised trials (RoB 1),(Reference Higgins21) which rate the quality of studies based on random sequence generation, allocation concealment, selective reporting, blinding of participants, and staff, blinding of outcome assessment, incomplete outcome data, and other likely sources of bias. Any disagreements were handled through conversation. Methodological defects affecting the findings of the study gave rise to a ‘high risk’ score for each domain, while a ‘low risk’ score was given to each defect-free domain. When the information was not sufficient to evaluate the impact, the domain was scored ‘unclear risk’. According to the Cochrane Handbook’s guideline, the studies were finally classified as having a low risk of bias, a high risk of bias, or an uncertain bias(Reference Higgins21) (Table 1).
Table 1. Risk of bias for randomised controlled trials, assessed according to the revised Cochrane risk-of-bias tool for randomised trials (RoB 1)
L, low risk of bias; H, high risk of bias; U, unknown.
Statistical analysis
Using the random-effects method (DerSimonian and Laird)(Reference DerSimonian and Laird22), standardised mean differences, estimated by Hedges’ g, with the 95% confidence intervals (CIs) of ALT and AST were used as opposed to weighted mean differences (WMDs) with 95% CIs for glucose and insulin. Also, in studies that only provided SEM (standard error of the mean), SDs were derived using the following formula: SD = SEM × √n, where n is the number of participants in each group. When medians and interquartile ranges were presented instead of mean and SD values, SDs were obtained using SD = interquartile range/1.35 (symmetrical data distribution)(Reference Hozo, Djulbegovic and Hozo23). If RCTs did not report the SDs of the change, we calculated them using the following formula in which the correlation coefficient (R) was considered 0.8: SDdifference = Square Root [(SDpre-treatment)2 + (SDpost-treatment)2 − (2 × R × SDpre-treatment × SDpost-treatment)](Reference Borenstein, Hedges, Higgins and Rothstein24). Using the Cochrane’s Q-test and the I-squared (I 2 ) values, the heterogeneity among studies was evaluated. If I 2 values >50% or P<0.05, between-study heterogeneity was regarded as significant(Reference Sheikhhossein, Amini, Shahinfar, Djafari, Safabakhsh and Shab-Bidar25,Reference Sheikhhossein, Borazjani and Jafari26) . We conducted a stratified analysis to determine the cause of the heterogeneity among the examined studies. Sensitivity analyses were carried out by excluding each study individually and rerunning the pooled analysis in order to examine the impact of each study on the overall effect size. To identify the likely publication bias, Egger’s regression test was used(Reference Egger, Smith, Schneider and Minder27). All statistical tests were performed using STATA, version 14 (StataCorp, College Station, Texas, USA). Statistics were considered significant for P-values under 0.05.
Results
Search results
A primary systematic search of databases turned up 569 papers in all, and manual searching turned up one study. After excluding duplicates, 464 articles were reviewed based on the title and abstract screening approach, which resulted in the exclusion of 440 articles due to no relevant or original data (n = 196), animal studies (n = 160), and reviews (n = 84). From twenty-four studies that entered the full-text examination, fifteen studies were excluded because of the following reasons: irrelevant research (n = 9), no control group (n = 2), inadequate data of interest (n = 1), the same participants (n = 2), and combined intervention (n = 1). Nine RCTs(Reference Al-Kuraishy and Al-Gareeb10–Reference Arefhosseini, Tutunchi, Nomi-Golzar, Mahboob, Pouretedal and Ebrahimi-Mameghani18) were finally included in the qualitative and quantitative synthesis (Fig. 1).
Figure 1. Flow chart of the number of studies identified and selected into the meta-analysis.
Study characteristics
The detailed characteristics of selected studies are summarised in Table 2. These trials have involved 444 participants. The range of publication years of selected articles was between 2001 and 2022. Two studies were carried out in Japan(Reference Hayamizu, Tomi, Kaneko, Shen, Soni and Yoshino14,Reference Hayamizu, Ismi and Kaneko15) , one in the Netherlands(Reference Kovacs and Westerterp-Plantenga12), one in Korea(Reference Kim, Jeon and Park13), one in Taiwan(Reference Lu, Yang and Wu17), one in Brazil(Reference Vasques, Schneider, Klein, Falavigna, Piazza and Rossetto11), one in Germany(Reference Chong, Beah, Grube and Riede16), one in Iraq(Reference Al-Kuraishy and Al-Gareeb10), and one in Iran(Reference Arefhosseini, Tutunchi, Nomi-Golzar, Mahboob, Pouretedal and Ebrahimi-Mameghani18). All studies were RCTs that were conducted on adult subjects aged ≥ 18 years. Two studies included only women(Reference Vasques, Schneider, Klein, Falavigna, Piazza and Rossetto11,Reference Arefhosseini, Tutunchi, Nomi-Golzar, Mahboob, Pouretedal and Ebrahimi-Mameghani18) , two included only men(Reference Al-Kuraishy and Al-Gareeb10,Reference Kovacs and Westerterp-Plantenga12) , and the rest of the RCTs were conducted on both genders(Reference Kim, Jeon and Park13–Reference Lu, Yang and Wu17). The mean BMI varied from 21.8 to 38.1 kg/m2 among eligible articles. Except for two studies conducted on obese(Reference Al-Kuraishy and Al-Gareeb10) and non-alcoholic fatty liver disease(Reference Arefhosseini, Tutunchi, Nomi-Golzar, Mahboob, Pouretedal and Ebrahimi-Mameghani18) patients, other studies involved healthy individuals(Reference Vasques, Schneider, Klein, Falavigna, Piazza and Rossetto11–Reference Lu, Yang and Wu17). The sample size of the eligible studies ranged between ten and eighty-four subjects in both treatment and control groups. The dose of Garcinia cambogia extract supplementation ranged between 166 and 2800 mg/d. The control group received a placebo except for two studies in which the control group received orlistat(Reference Al-Kuraishy and Al-Gareeb10) and a calorie-restricted diet(Reference Arefhosseini, Tutunchi, Nomi-Golzar, Mahboob, Pouretedal and Ebrahimi-Mameghani18). The range of intervention length was from 11 d to 16 weeks among RCTs.
Table 2. Demographic characteristics of the included studies
RCT, randomised controlled trial; HCA, hydroxycitric acid; ALT, alanine aminotransferase; AST, aspartate aminotransferase; FBS, fasting blood sugar; NAFLD, non-alcoholic fatty liver disease.
Risk of bias assessment
As shown in Table 1, in terms of random sequence generation, all studies(Reference Al-Kuraishy and Al-Gareeb10–Reference Arefhosseini, Tutunchi, Nomi-Golzar, Mahboob, Pouretedal and Ebrahimi-Mameghani18) were scored with low risk since they mentioned random sequence generation methods. Additionally, five studies were evaluated as low risk for allocation concealment(Reference Kim, Jeon and Park13–Reference Chong, Beah, Grube and Riede16,Reference Arefhosseini, Tutunchi, Nomi-Golzar, Mahboob, Pouretedal and Ebrahimi-Mameghani18) ; however, four were rated with unclear risk(Reference Al-Kuraishy and Al-Gareeb10–Reference Kovacs and Westerterp-Plantenga12,Reference Lu, Yang and Wu17) . Researchers blinding was performed in no studies and was, thus, identified as high risk. Except for two studies that were considered as having a high risk of bias in the participants blinding(Reference Al-Kuraishy and Al-Gareeb10,Reference Arefhosseini, Tutunchi, Nomi-Golzar, Mahboob, Pouretedal and Ebrahimi-Mameghani18) , others reported that the participants were blinded and were given a low risk score. All studies were considered as having a low risk of bias in selective reporting. Eight RCTs explicitly mentioned incomplete outcome data receiving a low risk of bias(Reference Al-Kuraishy and Al-Gareeb10–Reference Kim, Jeon and Park13,Reference Hayamizu, Ismi and Kaneko15–Reference Arefhosseini, Tutunchi, Nomi-Golzar, Mahboob, Pouretedal and Ebrahimi-Mameghani18) . Since all studies were rated as having a high risk of bias in one domain (blinding of outcome assessment), they were considered as low quality.
Primary outcomes
Effect of Garcinia cambogia on FBS
As an end measure, the effects of Garcinia cambogia administration on FBS levels were investigated in seven trials(Reference Al-Kuraishy and Al-Gareeb10,Reference Kovacs and Westerterp-Plantenga12,Reference Hayamizu, Tomi, Kaneko, Shen, Soni and Yoshino14–Reference Arefhosseini, Tutunchi, Nomi-Golzar, Mahboob, Pouretedal and Ebrahimi-Mameghani18) . The random-effects model’s overall findings showed that taking Garcinia cambogia supplements did not significantly alter FBS levels (WMD: 1.02 mg/dl, 95% CI: −1.29, 3.33, P = 0.378) (I 2 = 52.4%, P = 0.050) (Fig. 2). According to Egger’s regression test, we did not find any evidence of publication bias for FBS concentrations (P = 0.335). To determine the impact of a single study on the total effect size, the sensitivity analysis was conducted by excluding each trial one at a time. Sensitivity analysis indicated that the overall effect of Garcinia cambogia supplementation on FBS was not dependent on any single study.
Figure 2. Forest plot detailing weighted mean difference and 95% confidence intervals (CIs) for the effect of Garcinia cambogia on glucose.
Due to significant heterogeneity among studies, subgroup analysis was also performed in which we stratified RCTs based on the intervention duration (≤ 8 or > 8 weeks), age (< 40 or ≥ 40 years), BMI (< 25, 25–29.9, or ≥ 30 kg/m2), and sex (both, male, or female). The heterogeneity was reduced when subgroup analysis was performed based on intervention duration (I 2 = 0%, P = 0.779), age (I 2 = 0%, P = 0.779), and sex (I 2 = 0%, P = 0.948). When individuals were < 40 years and the duration of intervention was less than 8 weeks, Garcinia cambogia supplementation reduced FBS levels; however, it was not statistically significant (WMD: −0.38 mg/dl, 95% CI: −2.3, 1.54, P = 0.134), and (WMD: −0.38 mg/dl, 95% CI: −2.3, 1.54, P = 0.134), respectively. In studies that mean BMI of participants was ≥ 30 kg/m2, FBS increased following Garcinia cambogia consumption (WMD: 3.57 mg/dl, 95% CI: 0.34, 6.80, P = 0.030). Similarly, female participants experienced increased levels of FBS following Garcinia cambogia supplementation (WMD: 5.41 mg/dl, 95% CI: 1.21, 9.61, P = 0.012) (Table 3).
Table 3. Subgroup analysis of included randomised controlled trials in meta-analysis of the effect of Garcinia cambogia on glycaemic control and liver enzymes
FBS, fasting blood sugar; ALT, alanine aminotransferase; AST, aspartate aminotransferase; WMD, weight mean difference.
Secondary outcomes
Effect of Garcinia cambogia on insulin
The effect of Garcinia cambogia supplementation on insulin concentrations was considered in six RCTs(Reference Vasques, Schneider, Klein, Falavigna, Piazza and Rossetto11,Reference Kovacs and Westerterp-Plantenga12,Reference Hayamizu, Tomi, Kaneko, Shen, Soni and Yoshino14,Reference Hayamizu, Ismi and Kaneko15,Reference Lu, Yang and Wu17,Reference Arefhosseini, Tutunchi, Nomi-Golzar, Mahboob, Pouretedal and Ebrahimi-Mameghani18) . Combined results using the random-effects model (Fig. 3) showed that there was no significant effect of Garcinia cambogia consumption on insulin levels (WMD: −0.12 mU/L, 95% CI: −1.50, 1.25, P = 0.861) with significant between-study heterogeneity (I 2 = 64.2%, P = 0.016). There was no evidence of publication bias (P = 0.719). The results of the sensitivity analysis showed that none of the individual studies had a significant effect on the overall effect size of findings.
Figure 3. Forest plot detailing weighted mean difference and 95% confidence intervals (CIs) for the effect of Garcinia cambogia on insulin.
When the BMI (I 2 = 47.3%, P = 0.150) and sex (I 2 = 47.3%, P = 0.150) subgroups were analysed, the heterogeneity vanished. Most reduction in insulin was found among participants of normal range (BMI = 18.5-24.9) when compared to participants with overweight (BMI = 25–29.9) following intervention (WMD: −1.05 mU/L, 95% CI: −2.68, 0.58, P = 0.207 vs. WMD: −0.64 mU/L, 95% CI: −1.78, 0.51, P = 0.276). Likewise, the reducing effect of Garcinia cambogia on insulin was greater in males (WMD: −1.05 mU/L, 95% CI: −2.68, 0.58, P = 0.207 compared by both genders (WMD: −0.64 mU/L, 95% CI: −1.78, 0.51, P = 0.276). These results also demonstrated that insulin levels significantly increased in trials conducted on females (WMD: 1.81 mU/L, 95% CI: 0.28, 3.34, P = 0.021) and participants with mean BMI ≥ 30 kg/m2 (WMD: 1.81 mU/L, 95% CI: 0.28, 3.34, P = 0.021) (Table 3).
Effect of Garcinia cambogia on ALT
A total of six studies(Reference Kim, Jeon and Park13–Reference Arefhosseini, Tutunchi, Nomi-Golzar, Mahboob, Pouretedal and Ebrahimi-Mameghani18) provided changes in ALT levels as an outcome measure. The results of the pooled analysis (Fig. 4) revealed that Garcinia cambogia supplementation did not alter ALT concentrations significantly (Hedges’ g: 0.27, 95% CI: −0.20, 0.73, P = 0.264) (I 2 = 76.9%, P = 0.001). Egger’s regression test found no significant publication bias among selected trials (P = 0.330). The findings were unchanged when any RCT was eliminated and no significant changes to our results occurred.
Figure 4. Forest plot detailing weighted mean difference and 95% confidence intervals (CIs) for the effect of Garcinia cambogia on alanine transaminase.
Due to the significant heterogeneity among RCTs, stratified analysis was also carried out by trial duration (I 2 = 26.1%, P = 0.258) and age (I 2 = 0%, P = 0.492), which decreased heterogeneity (Table 3). Non-significant changes in ALT levels were observed when studies were classified based on trial duration and age.
Effect of Garcinia cambogia on AST
The pooled mean differences of six trials(Reference Kim, Jeon and Park13–Reference Arefhosseini, Tutunchi, Nomi-Golzar, Mahboob, Pouretedal and Ebrahimi-Mameghani18) illustrated a non-significant effect on AST levels compared to the control group following Garcinia cambogia intervention (Hedges’ g: −0.08, 95% CI: −0.43, 0.26, P = 0.632) (Fig. 5). We found no evidence of publication bias for AST levels (P = 0.638). According to the results of the sensitivity analysis, none of the studies exerted significant effects on the combined effect size.
Figure 5. Forest plot detailing weighted mean difference and 95% confidence intervals (CIs) for the effect of Garcinia cambogia on aspartate transaminase.
Although there was a lot of heterogeneities among the articles, it was reduced when trials were grouped by age (I 2 = 5.7%, P = 0.303) and the duration (I 2 = 34.5%, P = 0.217) of the intervention (Table 3). The non-significant effect of Garcinia cambogia on AST levels remained after classifying articles by aforementioned variables with a higher reducing effect among subjects aged > 40 years (Hedges’ g: −1.01, 95% CI: −0.37, 0.34, P = 0.951).
Discussion
The primary outcome of the current systematic review and meta-analysis was that there was no significant effect on FBS. The secondary outcomes indicated consumption of Garcinia cambogia has no significant influence on insulin, AST, or ALT. Although subgroup analysis showed Garcinia cambogia significantly increased FBS and insulin levels in women with a BMI ≥30, it significantly reduced insulin levels when it was used for more than 8 weeks of administration.
Primary outcomes
Effect of Garcinia cambogia on FBS
Our results are in agreement with previous studies. In animal studies, Hayamizue et al. demonstrated that the administration of garcinia daily to mice for a duration of 28 d had no effect on the serum glucose level of the mice(Reference Hayamizu, Hirakawa and Oikawa28). Moreover, Shetty et al found no hypoglycaemic effect on diabetic rat models treated with Garcinia cambogia for 21 d(Reference Shetty, Rai, Ravindran, Gopalakrishna, Pai and Kalal29). Another study in healthy obese men has indicated that treatment with 166 mg/d Garcinia cambogia for 3 months did not affect glucose levels(Reference Al-Kuraishy and Al-Gareeb10). According to Chonge et al.’s study, administering Garcinia cambogia extract with 50% HCA (1950 mg) twice daily for a duration of 12 weeks did not result in any significant alteration in glucose levels in healthy overweight adults(Reference Chong, Beah, Grube and Riede16). In an animal study, oral administration of HCA at a dose of 310 mg/kg/bw reduced glucose levels in rats(Reference Wielinga, Wachters-Hagedoorn and Bouter30). In another animal study, Kirana et al. showed the aqueous extract of Garcinia cambogia rind at 100 and 200 mg/kg/bw decreased glucose in diabetic rat models(Reference Kirana and Srinivasan31). Differences in studies may be attributed to differences in population, intervention duration, doses, and formulation of Garcinia cambogia, which influence the bioavailability of Garcinia cambogia. The phytochemical makeup of a plant extract also influences its biological capabilities. The phytochemical composition can vary depending on geographical location, climate, and soil type(Reference Hayat, Bundschuh and Jan32,Reference Shahinfar, Amini and Payandeh33) . Moreover, the common significant effect of Garcinia cambogia on the reduction of glucose levels was seen in subjects with a high baseline range of glucose levels(Reference Buchholz and Melzig34).
Secondary outcomes
Effect of Garcinia cambogia on insulin
Furthermore, according to the results of our study, Vasque et al. reported that insulin levels did not change after receiving 2.4 g of Garcinia cambogia extract with 50% HCA for 60 d(Reference Vasques, Schneider, Klein, Falavigna, Piazza and Rossetto11). In patients with type 2 diabetes, small intestinal infusions of 2800 mg HCA did not affect blood glucose or plasma insulin(Reference Thazhath, Wu and Bound35). In contrast, HCA supplementation in exercised individuals at 500 mg/d for 7 d enhanced post-meal insulin sensitivity(Reference Cheng, Huang and Lu36). Differences among studies may be for various populations, intervention duration, Garcinia cambogia doses, and formulation of Garcinia cambogia, which may influence the bioavailability of Garcinia cambogia.
One possible mechanism that explains the link between Garcinia cambogia and glycaemic factors is a competition of HCA with citric acid for citrate lyase, which leads to reduced decomposition of citric acid and increased cellular citric acid that cause inhabitation of glycolysis and facilitated glycogenesis and resulted in reduction of glucose concentrations(Reference Soni, Burdock, Preuss, Stohs, Ohia and Bagchi37,Reference Asghar, Monjok, Kouamou, Ohia, Bagchi and Lokhandwala38) . As a result, the amount of insulin that should be secreted to regulate blood glucose is reduced(Reference Asghar, Monjok, Kouamou, Ohia, Bagchi and Lokhandwala38). The HCA isomer (2s,3r) also inhibited intestinal and pancreatic glucosidase, which reduced glucose absorption and carbohydrate metabolism and ultimately lowered blood sugar levels(Reference Buchholz and Melzig34). Moreover, Garcinia cambogia may decrease intestinal uptake of glucose by releasing serotonin(Reference Chuah, Ho, Beh and Yeap39).
Effect of Garcinia cambogia on ALT and AST
In this study, we systematically reviewed the effect of Garcinia cambogia on AST and ALT levels. Consistent with our findings, an animal study by Clouatre and his colleagues indicated no change in AST and ALT in rats who received HCA for 16 weeks(Reference Clouatre and Preuss40). Besides, a clinical study showed AST and ALT levels did not change in overweight adults who received Garcinia cambogia (containing 1000 mg of HCA per day) for 12 weeks(Reference Hayamizu, Tomi, Kaneko, Shen, Soni and Yoshino14). However, Stohs et al. found that in participants who received 4600 mg of HCA for 8 weeks, the AST and ALT levels decreased significantly(Reference Stohs, Preuss and Ohia41). A recent study has reported six cases that had a rise in AST and ALT levels associated with weight loss with HCA(Reference Lobb42). In an animal study, Sanchez et al. reported that mice supplemented with Garcinia cambogia showed increased plasma AST and ALT levels(Reference Sánchez-Valle, C Chavez-Tapia, Uribe and Méndez-Sánchez43). Discrepancies among studies may partly be due to differences in age, gender, health conditions, sample size, dose, and duration of interventions.
Consistent with our results, evidence has shown that Antichol, which contains Garcinia cambogia (8% w/w), inhibits cholesterol-induced fatty degeneration of the liver and changes in the liver enzymes(Reference Pamidiboina, Chikkaboraiah, Razdan and Krishnamurthy44). Another study revealed the antioxidant property of Garcinia cambogia could keep the AST and ALT at normal levels(Reference Mahendran and Shyamala45).
In the sub-group analysis, Garcinia cambogia increased glucose and insulin levels in the females with BMI ≥ 30; however, it was still in the standard range of healthy humans. In fact, the level of increasing glucose and insulin was statistically significant, but this was not clinically significant. Furthermore, the sub-group analysis based on the duration of the study showed that the administration of Garcinia cambogia for more than 8 weeks decreased insulin levels.
Strengths and limitations
Our understanding allows us to say that the current meta-analysis is the first meta-analysis of the effects of Garcinia cambogia on glycaemic control and liver enzymes. Moreover, we assessed publication bias by the results of Egger’s test, and no evidence of publication bias was indicated which caused more reliability of our results. However, our results should be explained cautiously considering some of the limitations in the present meta-analysis. Most importantly, few studies were eligible, and the majority of them had small sample sizes. In addition, the significant differences in methodology, sample sizes, ages of patients, health conditions, and countries of participants indicated significant heterogeneity among the studies.
Conclusion
The findings of this meta-analysis demonstrated that Garcinia cambogia could be effective in reducing insulin levels when administered for more than 2 months. However, further well-designed clinical trials with long-term intervention and different doses of Garcinia cambogia, especially in pre-diabetic and diabetic subjects and patients with liver diseases, are advised to confirm our results.
Authorship
MRA created the research. Data screening and literature searches were carried out by ST and MS. MRA independently extracted data and evaluated its quality. The text was written by ST, MS, RR, MAS, and MA after data interpretation. The study was headed by AH. The final manuscript was read and approved by all writers.
Financial support
This study is related to the project no. 1401/59706 from Student Research Committee, Shahid Beheshti University of Medical Sciences, Tehran, Iran.
We also appreciate the ‘Student Research Committee’ and ‘Research and Technology Chancellor’ in Shahid Beheshti University of Medical Sciences for their financial support of this study.
Competing interests
The authors declared no conflicts of interest.
Open access
