Professional Documents
Culture Documents
REVIEW
The effect of chlorogenic acid on blood pressure: a systematic
review and meta-analysis of randomized clinical trials
IJ Onakpoya1, EA Spencer1, MJ Thompson1,2 and CJ Heneghan1
Several dietary supplements are currently marketed for management of hypertension, but the evidence for effectiveness is
conflicting. Our objective was to critically appraise and evaluate the evidence for the effectiveness of chlorogenic acids (CGAs) on
blood pressure, using data from published randomized clinical trials (RCTs). Electronic searches were conducted in Medline,
Embase, Amed, Cinahl and The Cochrane Library. We also hand-searched the bibliographies of all retrieved articles. Two reviewers
independently determined the eligibility of studies and extracted the data. The reporting quality of all included studies was
assessed by the use of a quality assessment checklist adapted from the Consolidated Standard of Reporting Trials Statement.
Disagreements were resolved through discussion. Seven eligible studies were identified, and five including 364 participants were
included. There were variations in the reporting quality of the included RCTs. Meta-analysis revealed a statistically significant
reduction in systolic blood pressure in favour of CGA (mean difference (MD): 4.31 mm Hg; 95% confidence interval (CI): 5.60 to
3.01; I2 ¼ 65%; Po0.00001). Meta-analysis also showed a significant reduction in diastolic blood pressure favouring CGA (MD:
3.68 mm Hg; 95% CI: 3.91 to 3.45; I2 ¼ 97%; Po0.00001). All studies reported no adverse events. In conclusion, the evidence
from published RCTs suggests that CGA intake causes statistically significant reductions in systolic and diastolic blood pressures.
The size of the effect is moderate. Few clinical trials have been conducted; they vary in design and methodology and are confined
to Asian populations and funded by CGA manufacturers. Large independent trials evaluating the effects of CGA on blood pressure
are warranted.
1
University of Oxford, Nuffield Department of Primary Care Health Sciences, New Radcliffe House, Radcliffe Observatory Quarter, Oxford, UK and 2Department of Family Medicine,
University of Washington, Seattle, WA, USA. Correspondence: Dr IJ Onakpoya, University of Oxford, Nuffield Department of Primary Care Health Sciences, New Radcliffe House,
Radcliffe Observatory Quarter, Oxford, Oxfordshire OX2 6GG, UK.
E-mail: igho.onakpoya@phc.ox.ac.uk
Received 7 February 2014; revised 15 April 2014; accepted 12 May 2014
Onakpoya et al./chlorogenic acids and blood pressure
IJ Onakpoya et al
2
pressure, green coffee, green coffee extract, CGA, caffeic acid, ferulic acid, (Figure 2). No RCT adequately reported randomization and
and derivatives of these (a Medline search strategy has been included as a allocation techniques, but three RCTs (60%) reported adequate
web supplement; Supplementary Appendix 1). We also searched Google blinding of study investigators and participants. No RCT clearly
Scholar for relevant conference proceedings using the search terms reported how outcome assessments were blinded. There was no
‘chlorogenic acid’ and ‘conference’, and hand-searched the bibliographies evidence of selective outcome reporting across the RCTs, but only
of all retrieved articles. No age, gender or language restrictions were
imposed. two RCTs (40%) showed low risk of other biases. All RCTs were
Only double-blinded RCTs were included in this review. To be conducted in two Asian countries (four in Japan, and one in India;
considered for inclusion, RCTs had to test the effectiveness of orally Table 1). Participants in three RCTs26,28,29 had mild hypertension,
administered CGA supplement against placebos or identical controls for whereas those in two18,27 were reported to have normal blood
blood pressure reduction in participants with or without hypertension. pressure; one of which18 included participants with reduced
Studies had to report blood pressure as outcome measures, and must have vasodilatory responses. The age of participants in these studies
had at least four weeks of intervention because the study findings may ranged between 22 and 65 years, and the study duration ranged
have implications for public health.21 Studies were included irrespective of from 4 to 26 weeks, with daily dosages of CGA varying from 25 to
lifestyle modification incorporated into the trial. Studies in which CGA was
1050 mg (Table 1).
combined with other supplements for blood pressure management were
excluded from the review.
Meta-analysis of the five RCTs (n ¼ 507; Figure 3) showed
Two reviewers (IJO and EAS) independently determined the eligibility of a statistically significant reduction in systolic blood pressure in
studies. Data extracted by two reviewers (IJO and EAS) included patient the CGA group compared with placebo or controls (MD:
characteristics, interventions and results. The reporting quality of all 4.31 mm Hg; 95% CI: 5.60 to 3.01; I2 ¼ 65%; Po0.00001).
included studies was assessed by the use of a quality assessment checklist A dose-effect plot showed a significant correlation between CGA
(Supplementary Appendix 2) of the Consolidated Standard of Reporting dose and reduction in systolic blood pressure (P ¼ 0.005; Figure 4).
Trials Statement.22 Disagreements were resolved through discussion. Meta-analysis of three RCTs (n ¼ 415) with good evidence of
The data were presented as means with standard deviations. Mean blinding of participants and personnel showed a significant
changes in systolic and diastolic blood pressures were used as primary reduction in systolic blood pressure in favour of CGA (MD:
endpoints to assess the differences between the CGA and placebo or
control groups. Using standard meta-analysis software (RevMan 5.0),23 we 3.20 mm Hg; 95% CI: 4.01 to 2.36; I2 ¼ 0%; Po0.00001).
generated a risk of bias graph, and computed mean differences (MDs) and Meta-analysis excluding the crossover study and participants with
95% confidence intervals (CIs) for studies with sufficient data for statistical normal blood pressure (n ¼ 423) also revealed a significant
pooling. The random-effects model was used for meta-analyses.24 reduction in systolic blood pressure favouring CGA (MD:
Sensitivity analyses (analysing trials based on study design) were used to 3.71 mm Hg; 95% CI: 5.06 to 2.36; I2 ¼ 63%; Po0.00001).
test the robustness of overall analyses using the I2 statistic; values of 25, 50 Meta-analysis of the two RCTs with larger sample sizes (n ¼ 395)
and 75% indicated low, medium and high statistical heterogeneity, showed a statistically significant reduction in systolic blood
respectively. Dose-effect correlations were used to examine the pressure favouring CGA over placebo/controls (MD: 3.12; 95%
relationship between the dosage of CGA and changes in blood pressure.
CI: 3.98 to 2.27; I2 ¼ 0%; Po0.00001).
Meta-analysis of the five RCTs (n ¼ 507; Figure 5) showed a
significant reduction in diastolic blood pressure favouring CGA
RESULTS over placebo or controls (MD: 3.68 mm Hg; 95% CI: 3.91 to
Our searches identified 83 non-duplicate citations from which 3.45; I2 ¼ 97%; Po0.00001). A dose-effect plot excluding two
seven eligible studies were identified (Figure 1). Two trials were outliers failed to show a significant association between CGA dose
excluded because their duration was less than 4 weeks.19,25 Five and diastolic blood pressure reduction (r ¼ 0.6; P ¼ 0.1). Meta-
trials18,26–29 including a total of 364 participants were included in analysis of the RCTs excluding participants with normal blood
the review. pressure (n ¼ 487) revealed a significant reduction in diastolic
Four trials were of parallel design, whereas one27 was crossover. blood pressure in favour of CGA (MD: 2.45 mm Hg; 95% CI:
There were variations in the reporting quality of the included RCTs 2.72 to 2.17; I2 ¼ 51%; Po0.00001). Meta-analysis of three
RCTs with parallel group design and excluding participants with
normal blood pressure (n ¼ 423) revealed a significant reduction in
# of records # of additional records diastolic blood pressure favouring CGA over placebo or controls
identified through identified through other (MD: 2.51 mm Hg; 95% CI: 2.79 to 2.23; I2 ¼ 44%;
database sources (n=0)
searching (n=107) Po0.00001). Meta-analysis of the RCTs with larger sample sizes
(n ¼ 395) also showed a significant reduction in diastolic blood
pressure in favour of CGA (MD: 2.18 mm Hg; 95% CI: 2.54 to
# of records after duplicates
removed (n= 83)
1.82; I2 ¼ 0%; Po0.00001).
No adverse events were observed in all the included studies,
and the compliance rate was at least 90%. In total, six drop-outs
# of records # of records
screened (n=83) excluded (n=76) were reported. Three studies26,28,29 were wholly funded by
manufacturers; one was part-funded by the manufacturer,27
whereas the funding source in one study was not specified.18
# of full-text articles
excluded. with reasons
Participants in three RCTs18,26,29 did not change their lifestyle,
(n=2)
# of full-text articles Studies with duration
assessed for eligibility
less than four weeks: 2 Random sequence generation (selection bias)
(n=7)
Allocation concealment (selection bias)
Blinding of participants and personnel (performance bias)
# of studies included in Blinding of outcome assessment (detection bias)
qualitative synthesis (n=5) Incomplete outcome data (attrition bias)
Selective reporting (reporting bias)
Other bias
# of studies included in quantitative
synthesis (meta-analysis) (n=5) 0% 25% 50% 75% 100%
Study Year Relevant Type of subjects Age Study Intervention Comparator Baseline BP Adverse Lifestyle Funding
Country outcome and sample size (years) duration and daily dose (mm Hg): events modification source
(weeks) of CGA CGA/control
Kozuma et al.b Blood Mildly 30–50 4 25, 50 and Soy soup SBP: 146±5/ None No change Manufacturer
(2005)26 pressure hypertensive 100 mg in soy 145.4±5.5 in dietary sponsored
Japan (117) soup DBP: lifestyle
92.5±2.7/
91.7±2.5
Ochiai et al. Blood Normotensive 34.8–37.2 26 125 ml drink Placebo SBP: None No change Not specified
(2004)18 pressure males (20) containing drink free of 115.1±15.5/ in dietary
Japan CGA 140 mg GCE 118.9±6.9 lifestyle
DBP:
69.5±7.9/
66.9±6.6
Vinson et al.c Blood Normotensive 22–46 22 700 and 700 mg of SBP: None Dietary and Part
(2012)27 pressure and 1050 mg green inert capsule 125.38±5.10 exercise sponsored by
India overweight coffee with DBP: counselling manufacturer
(16) CGA 45.9% 81.88±2.68
Watanabe Blood Mild essential 43–63 12 125 ml fruit 125 ml fruit SBP: 145±6/ None Weight Manufacturer
et al. (2006)28 pressure hypertension and vegetable and 147±5 reduction, sponsored
Japan (28) juice vegetable DBP: 91±2/ sodium
containing juice 90±4 reduction,
CGA 140 mg physical
activity
Yamaguchi Blood Mild essential 30–65 4 Test coffee Test coffee SBP: None No change Manufacturer
et al.b (2008)29 pressure hypertension containing 82, 143.6±8.3/ in dietary sponsored
Japan (183) 172 and 143.6±8.7 lifestyle
299 mg CGA DBP:
91.0±2.8/
91.6±2.8
Abbreviations: CGA, chlorogenic acid; DBP, diastolic blood pressure; GCE, green coffee extract; RCT, randomized clinical trial; SBP, systolic blood pressure.
a
Unless otherwise stated, blood pressure values have been reported as means with s.d. bStudies had three CGA intervention groups. cStudy had two CGA
intervention groups.
whereas those in two RCTs27,28 had dietary and exercise analyses and the variation in the study designs and reporting
modifications. quality of the included studies. To our knowledge, this is the first
systematic review evaluating the effects of CGA on blood
pressure.
CGAs are involved in the suppression of macrophage infiltra-
DISCUSSION tion, and thus may be involved in blood vessel remodelling.30 A
Main findings short-term RCT in humans demonstrated that acute ingestion of
The results of our meta-analyses reveal that supplementation CGA-enriched coffee resulted in significant reductions in both
with CGA results in a statistically significant reduction in systolic systolic and diastolic blood pressures.25 There was evidence of
blood pressure. Our results also show that CGA administration greater flow-mediated dilation and increase in plasma nitric oxide
causes small, statistically significant reductions in diastolic concentration in the CGA group compared with controls, but
blood pressure. The sizes of the effects appear moderate, and these did not reach statistical significance. Although it is unclear if
the clinical relevance is modest at best. The results of the meta- the effects could have reached significant levels with longer-term
analysis should be interpreted with caution because of the consumption, our results corroborate the study findings on the
moderate-to-large statistical heterogeneity observed in the overall effects of CGA on blood pressure.
6
5 Research implications
4 Well-designed, independently-funded and adequately powered
3 RCTs and including participants of non-Asian descent evaluating
the effects of CGA on systolic and diastolic blood pressures are
2 needed. These trials should be of longer-term duration to allow for
1 monitoring of any adverse events associated with the ingestion of
CGA. Future investigators should adhere to standardized guide-
0
0 100 200 300 400 500 600 lines when reporting clinical trials results so as to minimize the risk
of bias in their studies. To better understand the biologic
Daily dose of CGA (mg)
mechanisms of CGA on blood pressure, well-designed studies
Figure 4. Scatter plot showing the relationship between daily CGA with longer duration examining the effects of CGA on flow-
dose and systolic blood pressure reduction. mediated dilatation and nitric oxide status in blood vessels are
Supplementary Information accompanies this paper on the Journal of Human Hypertension website (http://www.nature.com/jhh)