Drinking tea (Camellia sinensis, Theaceae) is associated with a reduced risk of cardiovascular disease, partly due to beneficial effects on the vasculature. These effects can be studied by measuring flow-mediated dilation (FMD), the extent to which the brachial (upper arm) artery relaxes as a result of its endothelium (inner lining of arteries) responding to a signal by nitric oxide (NO). This response is induced when a blood pressure cuff is released after prolonged tightness. Several human intervention studies have assessed the effect of tea on FMD and have shown a positive effect. This paper reports a meta-analysis of these studies.
Studies were gleaned from the Medline, Embase, Chemical Abstracts, and Biosis databases. Studies using freshly brewed green or black tea or tea powder from freshly brewed green or black tea were included; those using purified or isolated tea components (e.g., the tea catechin EGCG) were not. Also excluded were those studies missing data on FMD, with no measures of variability of FMD reported, no suitable control, or no full text available.
Quality of the studies was assessed using a tool specially designed for this analysis, based on the Delphi Consensus, which took into account randomization, similarity of baseline characteristics between treatment groups, specification of eligibility criteria, blinding procedure, proper execution of FMD measurements, and recording of compliance.
A total of 478 studies were discovered in the databases, of which 470 were excluded for a variety of reasons. One additional study was found after the search, bringing the total number of studies included in the meta-analysis to 9. Seven studies had a crossover design, and 2 had a parallel design. The studies included 213 subjects with the mean age ranging from 30.0-62.1 years and body mass index (BMI) ranging from 22.1 to 29.7 kg/m2. Subjects were healthy or mildly hypercholesterolemic (high cholesterol) in 5 of the studies; the other studies included patients with kidney disease or transplant, or coronary artery disease. One study used only men, 2 studies used only women, and the rest included both (1 study did not report the sex of its subjects).
Mean baseline FMD ranged from 4.3-7.8%. Six studies looked at the acute effects of FMD, and 3 studied long-term effects (1 month).
Black tea was used in 7 studies and green or green and black tea in 3 studies. Tea was brewed in hot water or tea powder dissolved in hot water using defined amounts of tea, water, and brewing time. The median dose of tea was 500 ml per day, roughly equal to 2-3 normal cups.
The meta-analysis showed that tea increased FMD by 2.6% (95% confidence interval [CI]: 1.8-3.3%; P<0.001), which was 40% better than the average FMD measured with placebo (in crossover studies) or at baseline (in parallel studies). In post-hoc analysis, the exclusion of 1 study with 4 arms because of its disproportionate contribution to the total, and 1 study in a very specialized disease population (renal transplant recipients) did not change the outcome of the analysis.
The quality scores (as measured by the Delphi Consensus tool) correlated significantly with the net FMD response (P<0.002), but this correlation was lost when other variables were included in the analysis. The type of tea, type of placebo, health status, study duration, age, and baseline FMD did not correlate significantly with net FMD.
Among predefined subgroups, there was a correlation of study quality score, with higher quality studies showing smaller improvements in FMD (P=0.005). There were significant overall FMD effects for subgroups with diseased and healthy subjects, young and old subjects, high and low baseline FMD values, different amounts of tea prescribed, different study quality scores, black tea as intervention, when using hot water as control, and with acute intake of tea.
In post-hoc analysis, the position of the cuff (whether proximal or distal to the area of ultrasound measurement) correlated significantly with FMD response; smaller responses were observed with distal occlusion (P=0.017; R2=0.36).
Between-study heterogeneity was significant (62.1, P<0.001, with accompanying I2 statistic 75.8%). A funnel plot showed an absence of publication bias (intercept: P=0.176). When the most extreme outlier was excluded, only the symmetry of the plot was improved (intercept: P=0.401).
While the active ingredients in tea are not completely known, the flavonoids (especially the catechins) have received wide interest and support. It is possible that the flavonoids may act by increasing NO, but other mechanisms may also be involved. The authors were not able to show a dose-dependent effect on FMD. This may be because a given volume of tea could contain varying amounts of active ingredients across studies, making comparisons difficult.
Limitations of the meta-analysis include methodological differences across studies (difficulty of reproducing FMD measurements and differences in cuff positioning); the fact that a majority of the studies were acute and the clinical relevance of such studies is unclear; and the inclusion of 2 studies that compared more than one active ingredient to the same control.
The authors concluded that, “Moderate consumption of tea substantially enhances FMD,” and that the results seem robust because 8 out of 9 studies showed a difference using various tea types and various populations, and there were no indications of systematic publication bias. Whether improved FMD is correlated with decreased risk of cardiovascular disease is still unclear.
—Risa Schulman, PhD