Introduction

In recent years, cardio-metabolic disorders have changed into a major public health issue worldwide. Some underlying risk factors of the cardio metabolic diseases include inflammation, hyperglycemia, dyslipidemia, and endothelial dysfunction. These rampant and burgeoning health issues affect about 20–25% of the world’s adult population [1, 2]. Since chemical drugs have several adverse effects, no single treatment has ever been identified for management of the cardio-metabolic disorders [3, 4]. Several studies reported the effective role of nutritional supplements such as antioxidants in prevention and treatment of the cardio-metabolic disorders [5, 6]. A growing number of researchers have been trying to find natural compounds to improve cardio-metabolic treatments [7]. Recently, herbal supplementation was found to be effective in treating chronic disorders [8].

Coffee, one of the most popular beverages, contains caffeine [9]. Coffee plays an important role in individuals’ diet and health due to its high consumption rate around the world [10]. Coffee is rich in phenolic compounds, which are protective agents against chronic degenerative diseases [11]. Green coffee consists of several bioactive components including chlorogenic acid (CGA), trigonelline, and polyphenols. Chlorogenic acids, the ester of quinic acid and caffeic acid, are the main components of polyphenols with antioxidant abilities [12]. Previous studies demonstrated the anti-diabetic, anti-carcinogenic, anti-inflammatory, and anti-obesity effects of CGA [13,14,15,16]. Several animal studies also reported that CGA had anti-diabetes [17], anti-obesity [18], and anti-lipidaemic properties [18, 19] with beneficial effects on insulin resistance [20]. A recent meta-analysis also indicated that green coffee supplements could decrease blood pressure [21]. Despite the confirmed positive effects of green coffee supplementation on cardio-metabolic factors in animal studies, the results of human studies are still inconsistent. Although positive effects of green coffee were reported on cardio-metabolic factors, such as lipid profile in the literature [22], some studies did not find any significant association in this regard [23, 24].

So, this systematic review and meta-analysis aimed to review the available randomized controlled trials (RCTs) to assess the effect of green coffee extract (GCE) supplementation on fasting blood sugar (FBS), insulin, homeostatic model assessment for insulin resistance (HOMA-IR), triglyceride (TG), low-density lipoprotein (LDL), high-density lipoprotein (HDL) and C - reactive protein (CRP) concentrations.

Materials and methods

Data sources and search strategy

The current systematic review was conducted based on the guidelines provided by PRISMA [25]. Online databases including PubMed, Embase, Scopus, and Web of Science were systematically searched to find the related articles over the effect of green coffee supplementation on cardio metabolic risk factors up to October 2019 with no language and time restrictions. In addition, Google Scholar, Cochrane Library, and reference lists of the related papers were manually searched to find any other possible references. The keywords used in this systematic search included: (“green coffee” or “green coffee extract” or “chlorogenic acid”) AND (“cardiometabolic” OR “cardio metabolic” OR “blood sugar” OR “fasting blood sugar” OR “FBS” OR insulin OR glycaemia OR “insulin resistance” OR “insulin sensitivity” OR “HOMAIR” OR “HOMA-IR” OR “triglyceride” OR “TG” or “lipid profile” OR “Low-density lipoprotein” OR LDL OR HDL OR “High – density lipoprotein” OR cholesterol OR “Total cholesterol” OR inflammation OR “CRP” OR “C-reactive protein”) AND (“trial” OR “randomi*” OR “control” OR “clinical” OR “intervention” OR “randomized” OR “placebo” OR “blind”).

Inclusion and exclusion criteria

Articles were included in our meta-analysis in the case that they: [1] were clinical trials, [2] investigated adults over 18 years of age [3], reported sufficient data on the baseline and final tests of FBS, TG, HDL, LDL, cholesterol, insulin, HOMA-IR, or/and CRP in both green coffee and control groups [4], and conducted the intervention using any green coffee species. Studies were excluded in the case that they lasted less than 2 weeks, were nonrandomized clinical trials, were animal or cross-sectional studies, conducted on children, and examined the effects of green coffee along with other interventions. Grey literature such as conference papers, dissertations, and patents were also removed from the analysis.

Data selection

The data extraction process was completed by two independent reviewers (MM and MR) according to the eligibility criteria (Table 1). Extracted information included: first author’s name, paper publication year, sample size, trial design, green coffee extract (GCE) dose, study duration, participants’ mean age, research quality, and study outcomes. Any disagreements were resolved by the third reviewer (ER) at this stage.

Table 1 Inclusion and exclusion criteria following the PICOS approach
Full size table

Quality assessment

Quality of the papers was evaluated by the Cochrane Collaboration’s risk of bias assessment tool [26] according to the following domains: random sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), and selective reporting (reporting bias). Each domain was judged to have low, high, or unclear risk of bias. Finally, based on these results, the overall quality of studies was defined as good (low risk in more than three domains), fair (low risk in three domains), and poor (low risk in less than three domains).

Statistical analysis

Statistical analysis was performed by Stata 12. Weighted mean difference (WMD) and 95% confidence interval (CI) were used to evaluate the effect of GCE supplementation on cardio metabolic risk factors. In the case that the changes were not reported in studies, the baseline and final mean values and standard deviations (SDs) were extracted and SD of the mean changes was calculated using the correlation coefficient of 0.5. If the heterogeneity was high, fixed- or random-effects models were performed. Significant and high heterogeneity were assessed using p value <0.05 and I2 ≥ 50%, respectively. Sensitivity analysis was conducted to examine the effect of each study on the pooled results. Egger weighted regression test was also performed to identify the potential publication bias. P value <0.05 was considered as statistically significant.

Results

The database searches identified 465 articles. In the first screening phase, after reading the studies’ titles and abstracts, 325 articles were removed and 140 studies were retrieved for full-text review. Finally, 27 RCTs met the eligibility criteria [22,23,24, 27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50]. The study flow diagram is depicted in Fig. 1.

Fig. 1
figure 1

Flow diagram of literature search according to the PRISMA statement

Full size image

Study characteristics

Table 2 includes the articles examined in this meta-analysis. Of 27 articles that met the eligibility criteria, five had cross-over design and others had a parallel design. Studies were conducted in Spain, Japan, China, Korea, France, Netherlands, Colombia, Mexico, and Iran. The mean dose of GCE or its active compounds varied from 200 to 2000 mg/day and the duration of intervention ranged from 2 to 24 weeks. All evaluated studies were carried out among adults. A total number of 992 participants (565 participants in the intervention groups and 427 participants in the control groups) were included in the final analysis.

Table 2 Characteristics of included studies in the systematic-review
Full size table

Quality assessment

The Cochrane risk-of-bias tool was applied to evaluate quality of the studies. According to the results, 11 articles had good quality, seven had fair quality, and nine had poor quality.

Meta-analysis results over the effect of GCE on glycemic controls

The FBS was investigated and reported in 14 studies among a total of 512 participants. The random-effects’ model findings revealed that GCE supplementation caused a significant reduction in the FBS concentration (WMD = −2.28, 95% CI: −4.49 to −0.07, P = 0.043) with a significant heterogeneity (I2 = 86.3%, P < 0.001) (Fig. 2). Based on the subgroup analysis, the GCE could decrease the FBS concentration significantly only in studies with good quality (WMD = −3.76, 95% CI: −6.59 to −0.93, P = 0.009) (Table 3). The leave-one-out sensitivity analysis revealed that the pooled effect size was not dependent on a single study (Table 1S). No publication bias was found based o9n the funnel plots analysis (Egger’s test P = 0.864; Begg’s test P = 0.274) (Fig. 3).

Fig. 2
figure 2

Forest plot summarizing the association between green coffee extract supplementation on circulating FBS concentrations

Full size image
Table 3 Results of subgroup-analysis for effect of green coffee extract on cardio metabolic risk factors
Full size table
Fig. 3
figure 3

Funnel plots detailing publication bias in the selected studies of the relation between intakes of green coffee extract and FBS

Full size image

The effect of GCE on insulin levels was examined in 10 articles. According to the meta-analysis, insulin levels decreased significantly after the GCE supplementation (WMD = −0.53, 95% CI: −0.93 to −0.14, P = 0.008) (Fig. 4). Heterogeneity was significant between these studies (I2 = 53.8, P = 0.02). Considering the subgroup analysis, GCE supplementation decreased the insulin levels in studies with poor and fair quality levels (WMD = −0.53, 95% CI: −0.93 to −0.14, P = 0.008) and in studies that used higher doses of GCE (≥400 mg/day) (WMD = −0.53, 95% CI: −0.93 to −0.14, P = 0.004) (Table 3). Sensitivity analysis did not reveal alterations in the findings (Table 2S). According to the funnel plot and Egger’s statistics, no evidence of publication bias was found among trials (Egger’s test P = 0.485; Begg’s test P = 0.929) (Fig. 1S).

Fig. 4
figure 4

Forest plot summarizing the association between green coffee extract supplementation on circulating insulin concentrations

Full size image

Meta-analysis of HOMA-IR data showed that GCE supplementation had no significant effect on insulin sensitivity (WMD = −0.2, 95% CI: −0.47 to 0.08, P = 0.168) (Fig. 5), but the heterogeneity test was significant (I2 = 52.5%, P = 0.04). According to the subgroup analysis, GCE supplementation could reduce HOMA-IR significantly in studies with good quality (WMD = −0.32, 95% CI: −0.61 to −0.04, P = 0.03) (Table 3). Sensitivity analysis did not show any significant differences. Analysis of funnel plots and Egger’s statistics did not report any publication bias in the studies (Egger’s test P = 0.920; Begg’s test P = 0.621) (Fig. 2S).

Fig. 5
figure 5

Forest plot summarizing the association between green coffee extract supplementation on HOMA-IR

Full size image

Meta-analysis over the effect of GCE on lipid profile

After combining the findings from 19 published papers, we found that GCE supplementation had a significant effect on TG levels (WMD = −9.28, 95% CI: −14.93 to - 3.63, P = 0.001) (Fig. 6). Considering the high heterogeneity between trials (I2 = 67.6%, P < 0.001), we stratified trials based on the supplementation dose (>400 mg/day vs. ≤400 mg/day), intervention duration (>4 vs. ≤ 4 weeks), and study quality (good quality vs. poor or fair quality). The GCE supplementation significantly reduced TG concentration in studies with good quality (WMD = −10.18, 95% CI: −17.77 to - 2.58, P = 0.009) and longer duration (WMD = −10.81, 95% CI: −17.73 to - 3.89, P = 0.002) (Table 3). Sensitivity analysis indicated that removal of each study from the meta-analysis did not overthrow the result of the present pooled analysis. Moreover, no evidences of publication bias were found in studies (Egger’s test P = 0.920; Begg’s test P = 0.621) (Fig. 3S).

Fig. 6
figure 6

Forest plot summarizing the association between green coffee extract supplementation on TG

Full size image

Findings of 19 studies showed that GCE supplementation caused a non-significant reduction in the total cholesterol concentration (WMD = −2.87, 95% CI: −6.74 to 1, P = 0.146) (Fig. 7), with a high heterogeneity among studies (I2 = 58.5%, P < 0.001). Based on the subgroup analysis, GCE supplementation significantly reduced the cholesterol levels in studies with good quality (WMD = -8.98, 95% CI: −14.62 to −3.34, P = 0.005). In addition, longer intervention period (>8 weeks) reduced the cholesterol concentration significantly (WMD = -12.44, 95% CI: −18.44 to −5.86, P = 0.002). Subgroup analysis did not show any significant association between GCE supplementation and cholesterol levels (Table 3).

Fig. 7
figure 7

Forest plot summarizing the association between green coffee extract supplementation on Total cholesterol

Full size image

No evidence of publication bias was observed among the studies (Egger’s test P = 0.599; Begg’s test P = 0.328) (Fig. 4S).

Based on the results collected from 17 studies, GCE supplementation did not have any significant effect on the LDL cholesterol level (WMD = −0.8, 95% CI: −4.61 to 3, P = 0.215) (Fig. 8). According to the subgroup analysis, GCE supplementation could decrease LDL concentration significantly in the studies with good quality (WMD = −6.01, 95% CI: −11.30 to −0.72, P = 0.032) (Table 3). However, no evidence of publication bias was found in the studies.

Fig. 8
figure 8

Forest plot summarizing the association between green coffee extract supplementation on LDL

Full size image

In the final analysis, 18 studies examined the effect of GCE supplementation on HDL concentration. The results indicated that green coffee supplementation significantly increased the HDL levels (WMD = 1.33, 95% CI: 0.08 to 2.58, P = 0.037) (Fig. 9), with a significant heterogeneity (I2 = 65.6% P < 0.001). According to the subgroup analysis, GCE supplementation could increase the HDL concentration in the studies that administered higher doses (>400 mg/day) (WMD = 2.42, 95% CI: 1.26 to 3.58, P < 0.001), had fair and poor quality (WMD = 1.98, 95% CI: 0.36 to 3.61, P < 0.001), and lasted <4 weeks (WMD = 3.26, 95% CI: 0.55 to 5.96, P = 0.018) (Table 3). Sensitivity and publication bias analyses did not show any changes in the results (Fig. 5S).

Fig. 9
figure 9

Forest plot summarizing the association between green coffee extract supplementation on HDL

Full size image

Meta-analysis over the effect of green coffee on CRP level

Considering the results of 6 studies, green coffee supplementation did not have any significant effect on CRP concentration (WMD = −0.03, 95% CI: - 0.12 to 0.06, P = 0.486) (Fig. 10), with a high heterogeneity in studies (I2 = 82.3%, P < 0.001). Subgroup analysis and sensitivity analysis did not show any alteration in the findings. In addition, no evidence of publication bias was found in the studies (Egger’s test P = 0.430; Begg’s test P = 0.851) (Fig. 6S).

Fig. 10
figure 10

Forest plot summarizing the association between green coffee extract supplementation on CRP

Full size image

Discussion

Recently, many researchers focused on the effects of supplementation with green coffee and its ingredients on different aspects of human health. The present meta-analysis was conducted on eligible RCTs over the effect of GCE supplementation on serum CRP, lipid, and glycemic profile.

According to the results, GCE supplementation had a considerable effect on the reduction of insulin and FBS concentrations, but it had no significant effect on insulin resistance (measured by HOMA-IR). Based on the subgroup analysis, a significant decrease was observed in HOMA-IR among studies with high quality. Many recent pieces of evidence suggested a beneficial effect of green coffee on blood glucose. Green coffee is a complex beverage containing a wide variety of biologically active compounds [36], such as chlorogenic acid, which is the main bioactive component in coffee and is responsible for potential health effects of green coffee [49]. Inconsistent with our results, some previous studies reported a lowering effect of coffee on serum FBS [51, 52] and insulin [53, 54]. Based on a new review, GCE supplementation reduced FBG, but had no effect on insulin levels and the HOMA-IR status [52]. Mechanisms proposed to explain the glycemic lowering effects of green coffee and chlorogenic acid include: stimulating insulin secretion by increasing glucagon-like peptide-1 [55] and glucose-dependent insulinotropic peptide [56]; enhancing glucose uptake by skeletal muscles using adenosine 5′ monophosphate- protein kinase activation [57]; attenuating adipogenesis and proinflammatory cytokines accompanied with increased translocation of glucose transporter type 4 in muscle and adipose tissue [54]; decreasing intestinal glucose absorption by inhibiting digestive enzymes related to carbohydrates such as α-amylase, α-glucosidase, and pancreatic amylase isoenzymes I and II [58, 59]; up-regulating expression of the hepatic proliferation-activated receptor-α (PPAR-α) that promotes clearing lipids by liver and intensifies insulin sensitivity [60]; inhibiting hepatic glucose-6-phosphatase activity and subsequently less insulin secretion [61]; and destructing the Na + electrochemical gradient and accumulating glucose in enterocytes [62].

Some investigations reported no special effect of the GCE on glycemic parameters [63, 64]. A part of such discrepancies can be due to the short study duration, small sample size, differences in coffee preparing and administrating methods, as well as GCE doses.

The lipid-improving effects of GCE were confirmed by previous studies [60, 65, 66]. In the current study, we found that GCE had significant positive effect on HDL and TG levels, but had no significant influence on LDL and total cholesterol concentrations. Further subgroup analyses revealed significant reductions in the levels of TG, LDL, and total cholesterol in high-quality studies. A recent review reported that GCE supplementation increased HDL-C significantly, but reduced LDL-C and total cholesterol levels. However, the reduction in TG level was not significant [66]. In this regard, some studies indicated that CGA decreased TG, cholesterol, and LDL [60, 65, 67, 68], but increased HDL concentrations [63, 67]. Some mechanisms have been suggested for beneficial effects of green coffee, such as declining intestinal lipid absorption and changing hepatic lipid metabolism [37]. The inhibitory role of CGA was also suggested in cholesterol micelles formation and pancreatic lipase function that resulted in declining intestinal fat absorption [69]. Based on in vitro experiments, CGA had an inhibitory effect on some enzymes such as 3-hydroxy-3 -methylglutaryl coenzyme A reductase [69] and acyl-coenzyme A cholesterol acyltransferase, which reduced blood cholesterol [19]. As mentioned previously, CGA increased PPAR-α expression, which increased uptake and catabolism of fatty acids by the liver [19, 60]. However, some other studies found no substantial effect of CGA [67] or even indicated its adverse effects [70, 71] on lipid profile.

Coffee preparing method can be effective on its lipid reducing property; in other words, roasting in high temperatures results in degradation of GCE bioactive components, especially CGA [48], which reduces the health effects of green coffee. Coffee oil was also found to have hyperlipidemic effects due to its monodehydrated fatty acid esters, such as kahveol and cafestol. Coffee oil lipid-rising effect can be reduced by filtering [72].

Some RCTs over the association between green coffee consumption and inflammatory biomarkers, such as CRP showed positive effects of coffee on CRP concentration [73,74,75]. However, our findings did not show any significant effect of GCBE supplementation on CRP concentration. In our meta-analysis, three studies (of six) found that GCE supplementation had no significant effect on the CRP level [29, 42, 76]. Mansour et al. [29] reported no substantial decline in CRP levels after six months of coffee/ or chlorogenic acid supplementation in type 2 diabetes patients with non-alcoholic fatty liver disease. Nikpayam et al. [42] carried out a study among patients with metabolic syndrome and observed that GCE administration did not decrease oxidative stress, systematic, and vascular inflammation. Moreover, the acute and chronic effects of GCE and its bioactive components on endothelial function were examined in few studies. Cheong et al. investigated the effects of GCE administration on endothelial dysfunction in a mouse model of the metabolic syndrome and showed that GCE did not attenuate endothelial dysfunction induced by high-fat diet in animals [77]. In another study, Priftis et al. examined the protective effects of GCE in myoblasts and endothelial cells and showed that GCE caused a significant improvement in the cell redox status and endothelial function [78]. Kajikawa et al. evaluated the effects of consuming coffee with a high content of CGA on endothelial function among 37 patients with borderline or stage 1 hypertension. They reported a significant improvement in postprandial FMD after CGA supplementation [79]. The effects of GCE ingestion on endothelial function are quite controversial and more studies are required in this area [80]. Variation in participants’ characteristics might justify these discrepancies. In other words, some studies were carried out on participants with metabolic disease [36, 37, 42], while some others examined participants without this disease [36, 76]. The observed discrepancy in the findings can also be explained by different methods of GCE administration. For example, Martinez et al. [36] demonstrated that green/roasted coffee blend had beneficial effects on CRP level, while GCE had no substantial effects in other studies [42, 76]. Furthermore, long-term coffee consumption may have a potential effect on improving the systematic inflammation [63] . Most RCTs included in our meta-analysis were conducted in an 8-week duration, which may be not long enough to make substantial changes in the CRP level.

Our meta-analysis had some limitations. High heterogeneity was observed between the included studies. In addition, a limited number of studies were included regarding some variables, which could affect our results. Most of the included studies were conducted within ≤12 weeks with a small population size.

Conclusion

Based on the results, GCE supplementation improved the serum levels of FBS, insulin, TG, and HDL. As a result, consumption of GCE is recommended to improve the cardio-metabolic risk factors. Moreover, no adverse side effects were observed for GCE supplementation. However, further studies are needed to provide enough evidences on the effect of green coffee intake on preventing cardio metabolic disorders.