Abstract
Objective
In the past decade, increased research and media attention has been given to the weight loss effects of green coffee bean extract (GCBE). The purpose of our study was to perform a review and meta-analysis of available randomized clinical trials examining the effects of GCBE supplementation on weight loss in both normal weight and overweight/obese individuals.
Methods
Electronic and nonelectronic searches were conducted to identify all relevant randomized clinical trials. Reference lists of all retrieved articles were searched, and experts and manufacturers were contacted for unpublished trials.
Results
Six randomized trials were included in our review. The meta-analysis revealed a significantly larger weight loss effect for GCBE compared to placebo control (M ES=0.55). Moderator analysis revealed a large effect size for GCBE with overweight/obese individuals (M ES=1.00) on weight loss compared to a null effect for normal weight individuals with mild hypertension (M ES=0.01). There was considerable heterogeneity amongst the studies.
Conclusion
GCBE supplementation may be effective in promoting weight loss in overweight/obese subjects, but no effects were observed in normal weight subjects with mild hypertension. The paucity of data, inconsistent methodology, and low quality of currently available studies limited the conclusiveness of this analysis. Studies are needed to determine a mechanism of action for GCBE, and larger multisite clinical trials of high quality and consistent design are needed to elucidate the long-term effects and safety of GCBE supplementation on weight loss in normal weight, overweight, and obese subjects.
Introduction
Overweight and obesity are serious health concerns that have reached epidemic proportions. Increases in unhealthy food options and decreases in physical activity, in part, have led to increased obesity rates over the past 4 decades. Since 1980, worldwide obesity has increased by almost 200%. By 2015, the World Health Organization predicts that 2.3 billion adults will be overweight and 700 million of them will be obese.1 Long-term obesity has been linked to heart disease, hypertension, diabetes, and other life-threatening conditions.2
Concurrent with the rise in obesity is the rise in nutraceuticals that may aid weight loss. Green coffee bean extract (GCBE) is a nutraceutical that has recently received both media and research attention as a weight loss supplement. GCBE is present in green or raw coffee bean. GCBE contains large amounts of chlorogenic acid, a polyphenol that has antioxidant properties and influences glucose, fat, and brain energy metabolism.3,4 Although GCBE contains other components including caffeine and theophylline, it has been clinically studied for its chlorogenic content and for its lipolytic and weight-loss properties.5 In a meta-analysis of 3 randomized controlled trials Onakpoya et al found that GCBE extract resulted in significant weight loss in overweight and obese individuals compared to placebo controls.6
Although a larger portion of overweight and obese individuals are trying to lose weight, about 30% of normal weight adults are also attempting to lose weight.7,8 Thus, it is important to examine GCBE efficacy on weight loss in normal weight individuals. Due to the increased media popularity and research attention of GCBE as a weight-loss supplement, an updated meta-analysis examining the weight loss effects of GCBE is warranted to advance our understanding of its weight loss effects in both normal weight and overweight/obese individuals. The primary purpose of our study was to perform a review and meta-analysis of the available clinical research examining the effects of GCBE supplementation on weight loss in both normal and overweight/obese populations. The secondary purpose of our review was to discuss methodological limitations in the current literature and explore promising directions for future research.
Methods
To ensure the rigor of our meta-analysis, to the extent possible, we designed and reported our findings using a checklist of items in accordance with the “Preferred Reporting Items for Systemic Review and Meta-Analyses (PRISMA)” statement.9 Briefly, we used the following PRISMA checklist section/topics of title, introduction, methods, results, discussion, and funding).
Search Strategy and Inclusion Criteria
We used the following 5 search strategies to avoid bias retrieval of searching only major journals and to obtain grey literature (e.g., abstracts, unpublished studies).10 First, the following electronic databases were searched: AMED, CINAHL, The Cochrane Library, EMBASE, MEDLINE, PubMed, and Web of Science. Each database was searched from inception up until February 2013. The electronic databases were searched for title and abstract using the key terms outlined in Onakpoya et al.6 Second, ancestry searches (ie, treeing backward) were conducted using the reference lists of all located clinical studies. Third, we contacted active researchers in the field to retrieve either current or unpublished research. Fourth, computerized author database searches were conducted on all authors of retrieved studies meeting the inclusion criteria. Finally, manufacturers of commercial GCBE products were contacted to identify published and unpublished material. Only articles published in English were reviewed.
To conduct a comprehensive review of human clinical research we reviewed all randomized clinical trials that examined the effects of GCBE supplementation on any health condition in both healthy and diseased populations. Those studies that, based on their abstract, appeared to meet the criteria were independently considered for inclusion by the 2 authors. To be considered for inclusion, studies had to report the efficacy of GCBE for weight reduction and be a randomized controlled study. No weight, age, or gender restrictions were imposed for inclusion of studies. Studies that involved the use of GCBE as part of a combination treatment were excluded. Disagreements between the authors were resolved through discussion. A study was included if both reviewers agreed that it met the inclusion criteria.
Our search identified 10 clinical trials examining the health effects of GCBE. One trial was excluded because it failed to report sufficient information on weight to compute an effect size.11 Another trial was excluded because it did not measure weight as an outcome.12 A paper by Saito et al was excluded because it was published in Japanese.13 Finally, the Blum et al paper was excluded because it was not a randomized trial.14 A total of 6 randomized clinical trials (N=303 participants) met our inclusion criteria and were included in our meta-analysis.12,15-19
Meta-analytic Procedures
Using random effects modeling procedures, we calculated weighted mean effect sizes (ES) for the GCBE supplementation groups compared to the control groups and performed corrections for sample-size bias to estimate d. We used Hedges and Olkin’s procedures to correct for sample-size biases.20 When reported we used the M/SD scores of weight (kg) or BMI of the pre-post scores to compute ES. If not available, we used the M differences between pre and post scores. To derive ES for within-subject studies, one needs the correlation (r) between posttest and pretest measures. Unfortunately, it is very rare to find reported values of r in studies when the primary research studies do not investigate relationships between measures, and none of the studies in this synthesis reported r. Thus, we used a conservative value of r=0.50 to estimate the correlation between pretest and posttest values on measures of weight. A positive ES represented a positive effect for GCBE supplementation compared to the control group for weight loss.
Along with the weighted average ESs, we computed the 95% confidence intervals (CI). If the confidence interval does not include zero, then the M ES is statistically significant at the P< 0.05 level. The I2 statistic was used to assess for statistical heterogeneity amongst studies. We also graphed a forest plot, which is a graph of each study as a point estimate bounded by its confidence intervals. Finally, to assess publication bias, we calculated Rosenthal's Fail Safe N (Nfs) or file drawer analysis procedures.21 The Nfs indicates the number of missing studies (with M effect of 0) that would need to be added to the analysis before the combined effect would no longer be statistically significant. According to Rosenthal's conservative guidelines, Nfs should exceed 5k (k = number of studies) + 10 to effectively overcome the file drawer problem. The methodological quality of all included studies was assessed by the use of a quality assessment checklist adapted from the consolidated standard of reporting trials (CONSORT) guidelines.22
Results
Based on the search strategies described above, 6 randomized controlled trials were identified that examined the effects of GCBE on weight loss and met our inclusion criteria (see Table 1 for a summary of key details of these studies). Trial length ranged from 4 to 22 weeks (with the length of GCBE supplementation ranging from 4 to 12 weeks), and the dose of GCBE ranged from 46 mg/d to 1,050 mg/d.
GCBE supplementation resulted in significant weight loss for overweight/obese individuals but not for normal weight individuals with mild hypertension.
Our meta-analysis revealed a significantly larger weight loss effect for the GCBE group compared to the control group (M ES=0.55, 95% CI: 0.05, 1.05, P=0.03, n=9, Z=2.15. I2=83.24, Nfs=51; See Figure 1). The Nfs of 51 reveals that we would need to locate 50 ‘null’ studies for the combined 2-tailed P-value to exceed 0.05. That is, there would need to be 5.7 missing studies for every observed study for the effect to be nullified. The I2 statistic of 83.24% suggests that there is considerable heterogeneity amongst the studies.
Because of the large heterogeneity, we examined the ES by weight status/health (ie, overweight/obese vs normal weight individuals with mild hypertension). A significantly larger ES was evidenced for overweight/obese individuals (M ES=1.00, CI: 0.28, 1.71, n=5, P=0.006, I2=82.06) on weight loss compared to a null effect for normal weight individuals with mild hypertension (M ES=0.01, CI: -0.26, 0.29, P=0.91, I2=0) [QB(1)=6.46, P=0.01]; see Figure 2). In other words, GCBE supplementation resulted in significant weight loss for overweight/obese individuals but not for normal weight individuals with mild hypertension. None of the trials reported any severe adverse events associated with the use of GCBE supplementation.
Discussion
Obesity and overweight are serious health problems that have reached epidemic proportions. Weight problems can have a negative impact on quality of life and, in the case of obesity, can even lead to a significant reduction in life expectancy. Due to safety concerns and side effects of many prescription weight loss drugs, herbal remedies are becoming increasingly popular as alternatives to prescription medications for weight loss. Of these, GCBE, which contains high concentrations of chlorogenic acids, has emerged as a promising herbal compound for potentially preventing and treating overweight and obesity.
The purpose of our study was to perform a review and meta-analysis of the available clinical research examining the weight loss effects of GCBE supplementation in both normal-weight and overweight/obese populations. Of importance the normal-weight subjects included in our review also had mild hypertension. We found: (a) that GCBE supplementation resulted in significant weight loss compared to the control groups, (b) that GCBE supplementation resulted in a large weight loss ES for overweight/obese individuals compared to a null effect for normal weight individuals with mild hypertension, (c) considerable heterogeneity existed amongst the results, and (d) none of the trials reported any severe adverse events associated with the use of GCBE supplementation. Future research is needed to examine the effects of GCBE supplementation for weight loss on normal-weight people without a disease condition. More detailed review findings and directions for future research are briefly discussed below.
Consistent with the findings of Onakpoya et al, we found that GCBE supplementation resulted in significant weight loss compared to the control group.6 We found a large weight loss effect for overweight/obese individuals compared to a null effect for normal weight individuals with mild hypertension. The normal weight individuals had a BMI in the upper range of normal (ie, M BMI=24.40 and 24.67), had mild hypertension, and were 88.27% men. Because a significant number of normal weight individuals who may be approaching an overweight BMI and/or exhibiting a comorbidity of overweight such as hypertension may attempt to lose weight, further research is needed to examine the potential weight loss effects of GCBE in this population.
The studies reviewed included several design limitations. Specifically, studies conducted to date are limited by the following characteristics: use of single-center trials, short-term supplement duration studies (4–12 weeks), small sample sizes (N=16–117), lack of follow-up, limited descriptive information on the participants (eg, socioeconomic status, marital status, education level), and a lack of moderator analysis. The small sample sizes most likely precluded the examination of important moderator variables such as gender, age, and weight status (normal weight, overweight, and obese). Future trials involving GCBE supplementation should be conducted in line with the CONSORT guidelines.22 This will ensure the validity and applicability of study results. The paucity and often low quality of the primary studies reviewed limits the conclusiveness of our findings.
No adverse events were reported in any of the trials. This suggests that GCBE is safe for human consumption. However, given the short duration of the clinical trials to date, longer-term effects are currently unknown. It remains to be determined how safe this nutritional supplement is over the intermediate and long term. Future studies are also encouraged to examine a dose-response. In the studies reviewed, varying doses of GCBE were used ranging from 46 mg/d to 1,050 mg/d. As well, the GCBE was administered as both an oral capsule supplementation and a functional food added to soup, coffee, and fruit drinks. All six of the trials identified that GCBE supplement was enriched with chlorogenic acids. This information is important to accurately determine the health effects of GCBE. For example, Svetol® a commercially available green, unroasted decaffeinated coffee extract is the most studied form of GCBE. It contains about 40–45% chlorogenic acids, of which 5-caffeoylquinic acid represents about 15%.16
GCBE may exert its weight loss effects due to an increase in the consumption of fatty deposits, as shown by a change in the fat mass percentage, and the prevention of fatty deposits being accumulated.23 Chlorogenic acid might act by inhibiting glucose absorption in the small intestine. Further research is needed to examine the variety of molecules present in GCBE and the safety issues of caffeine consumption. In addition, preclinical trials reveal that the inhibition of the activity of glucose-6-phosphatase would limit the release of glucose into the general circulation and, therefore, limit insulinemia.24 This would lead to fewer fatty deposits in the adipose tissue through harder access into the adipose cells owing to reduced insulin activity and the consumption of fat reserves, due to the reduced availability of glucose as an energy source. Although the proposed mechanisms are supported, in part, by animal studies, future research is needed to elucidate the mechanisms of action in human research.
In summary, the results of our meta-analysis reveal that GCBE may be an effective nutraceutical in reducing weight in overweight/obese adults. However, further high-quality studies investigating the mechanism of action and efficacy of GCBE on weight loss are needed. While preliminary findings of weight loss combined with limited data demonstrating short-term safety point to a promising future, there is no way to know if these findings will translate into long-term health benefits until well-controlled studies are performed. Thus, we conclude that larger multi-site trials are needed to elucidate the potential mechanisms of action and role of GCBE in both the treatment and prevention of overweight/obesity.
Table 1
Summary of Clinical Trials Examining Effects of GCBE on Weight Loss
| First Author (year) [country] | Duration and Design | Participants | Treatment Group | Control / Comparison group | GCE Brand & Chlorogenic Acid Ratio | Control for Lifestyle Factors | Adverse Events |
| Ayton Res (2009) [England] | 4 wks, No further design information was explicit | N=62 overweight adults (BMI range = 25–30) | N=30 CGA enriched green coffee contained 180 mg/d GCE | N=32 Placebo control (instant coffee) | Coffee Shape 45-50% CA | Maintain same lifestyle habits | Not reported |
| Dellalibera (2007) [France] | 60 days randomized | N=50 men and women (age range = 19–75; BMI > 25) | N=30 Svetol GCBE (400 mg/d) | N=20 placebo | Svetol > 45% CA | Maintained same diet and exercise habits from T0-T60 | Not reported |
| Kozuma (2005) [Japan] | 28 d multicenter, randomized, double-blind, placebo-controlled, parallel grp | N=117 men with mild hypertension (age range = 30–50) | Soy sauce-flavored instant cup soup containing GCB a) 46 mg/d GCBE containing 25 mg CA b) 93 mg/d containing 50 mg CA c) 185 mg/d containing 100 mg CA | Placebo control (0 mg of CA) | T. Hasegawa Co., Ltd, Tokyo, Japan > 54% CA | Maintained pre-treatment dietary habits; breakfast was controlled as part of treatment | No adverse effects |
| Thom (2007) Study 2 [Norway] | 12 wk double-blind randomized placobe-controlled trial with 1 and 3 mnth follow-up | N=30 overweight adults (BMI range = 27. 5–32) M age = 24.2 | Coffee Slender® [five cups/d (11 g/day of coffee)] containing 200 mg of Svetol GCBE | Nescafé® Gold Norwegian blend instant coffee [five cups/day (11 g/day of coffee)] | Svetol > 45% CA | Maintained regular diet and exercise. No assessments. | No adverse effects |
| Vinson (2012) [India] | 22 wk randomized double-blind crossover | N=16 overweight men and women (M age = 33.19, BMI = 28.22) | a) high-dose (1,050 mg/d) for 6 wks b) low-dose for 6 wks (700 mg) | Placebo control for 6 wks | Applied Food Sciences (45.9% total chlorolgenic acids) (Austin, TX) GCATM 45.9% CA | Subjects were counseled on healthy dietary habits but no significant changes to diet were reported | No adverse effects |
| Watanabe (2006) [Japan] | 12 wk placebo-controlled, randomized clinical trial | N=28 adults with mild hypertension (M age=51.5) | 125 mL/d fruit and vegetable juice with 140 mg/d GCBE (n=14) | Placebo control (n=14) | Unknown GCE brand 30.9% CA | Unclear | No adverse effects |
Participants = number of participant who completed the trial
Table 2
Methodological Characteristics of Included Studies
| Author, Year, Country | Main Outcome | Main Diagnoses of Study Participants | Study Design | Gender Ratio M:F | M BMI | Randomized |
| Ayton Global Research, 2009, England | Body weight, Waist, hip and bust circumference | Overweight but otherwise healthy | Parallel | 62 total, unclear ratio | Range = 25 - 30 | yes |
| Dellalibera et al., 1998, France | Body weight, BMI | Overweight volunteers | Parallel | 50 total, unclear ratio | > 25 | ? |
| Kozuma et al., 2005, Japan | Body weight, Blood pressure, Hematological evaluation, Blood chemistry | Healthy with mild hypertension | Parallel | 117 M | 24.67 | yes |
| Thom, 2007, Norway | Body weight, BMI | Slight to moderately overweight subjects | Parallel | 30 total, unclear ratio | 27.5 - 32 | ? |
| Vinson et al., 2012, India | Body weight, BMI, percent body fat, Blood pressure, Heart rate | Overweight subjects | Crossover | 8 M 8 F | 28.22 | yes |
| Watanabe et al., 2006, Japan | BMI, Blood pressure, Pulse rate, Hematological evaluation | Patients with mild hypertension | ? | 11 M 17 F | 24.4 | yes |
