Effect of tea on FMD within and between different subgroups.

aFor continuous variables, studies were divided in subgroups based on their medians: 6.4% for baseline FMD, 55 years for age, 500 mL for dose of tea, and 6.0 points for quality.
bFor type of tea, type of placebo, study duration, and cuff placement, we performed sensitivity analysis due to limited number of studies (<5) in one of the subgroups.
cFor the subgroup with short-term duration, we excluded the long-term study arms (>2 weeks) and the study arms that investigated acute upon longer-term effects.
dPost-hoc analysis.