Background - Flow-mediated dilation (FMD) is an accepted technique to quantify endothelial (dys)function. The predictive value of FMD for future cardiovascular disease (CVD) risk has been investigated in several prospective studies. The purpose of the present study was to systematically review these studies and to quantify the relationship between FMD and CVD risk. It was hypothesized that the risk of experiencing a cardiovascular event later in life would be smaller when having a higher FMD value at baseline. Methods - Potentially relevant prospective studies were searched in several databases and by hand searching, through March 2011. Studies were selected based on predefined selection criteria. The statistical analysis was performed separately for continuous (per 1% increase in FMD) and categorical (when having a high vs. low FMD) risk estimates for CVD. The pooled overall risk estimate was calculated based on a random-effects model while weighting each study by the inverse of its variance. Potential sources of heterogeneity were investigated as well as presence of publication bias. Results - In total, 23 studies including 14,753 subjects were eligible for inclusion in the present meta-analysis. Subjects were followed for an average duration of 42.9 months (range: 12.1-94.6 months). The average age was 60.1 years (range: 46.0-78.6 years). In 8 out of 23 studies, asymptomatic subjects were included whereas in 15 studies, subjects were diagnosed as having various forms of diseases. In studies reporting continuous risk estimates, the pooled overall risk estimate for CVD expressed per 1 % increase in FMD was 0.90 (95%CI: 0.86, 0.94). One study, that was the only study including exclusively renal patients, reported results that significantly differed from this estimate. Excluding this study from the analysis did not materially affect the pooled overall risk estimate (RR = 0.92 (95%CI: 0.88; 0.95). Based on studies with categorical risk estimates, the pooled overall risk estimate for CVD when having a high FMD vs. a low FMD was 0.49 (95%CI: 0.38, 0.63). Covariate analysis revealed that health status significantly affected the association between FMD and CVD risk with a stronger association in diseased populations. Presence of publication bias could not be excluded. Conclusions - Our findings show that baseline FMD is inversely associated with CVD risk, particularly in diseased populations. Further studies should address the impact of underlying health status on the relationship between FMD and CVD.
|Publication status||Published - 2012|