Self-reported alcohol consumption, carbohydrate deficient transferrin and risk of cardiovascular disease: The PREVEND prospective cohort study.

• Alcohol consumption categories are continually associated with %CDT values. • Light to moderate alcohol consumption is associated with reduced CVD risk. • Percent CDT within the normal reference range is not associated with CVD risk. Self-reported alcohol consumption is an established risk factor for cardiovascular disease (CVD). Carbohydrate deficient transferrin (CDT) is an established objective marker of excessive alcohol consumption, but data on its prospective association with CVD are lacking. We aimed to evaluate the associations of self-reported alcohol consumption and CDT (expressed as %CDT, a more reliable marker than absolute CDT levels) with CVD risk. In the PREVEND prospective study of 5,206 participants (mean age, 53 years; 47.7% males), alcohol consumption by self-reports, absolute CDT measured using the Siemens nephelometric assay and %CDT calculated as the percentage of total transferrin concentrations, were assessed at baseline. Alcohol consumption was classified into 5 categories: abstention (reference), light, light–moderate, moderate and heavy alcohol consumption. Hazard ratios (HRs) (95% confidence intervals [CI]) for first CVD events were estimated. Mean (SD) of %CDT was 1.59 (0.54) %. During a median follow-up of 8.3 years, 326 first CVD events were recorded. Compared with abstainers, the multivariable-adjusted HRs (95% CIs) of CVD for light, light–moderate, moderate and heavy alcohol consumption were 0.66 (0.46–0.95), 0.83 (0.62–1.11), 0.83 (0.61–1.14) and 0.80 (0.48–1.36), respectively. Light alcohol consumption was associated with reduced coronary heart disease risk 0.62 (0.40–0.96), whereas light-moderate alcohol consumption was associated with reduced stroke risk 0.45 (0.24–0.83). The association of %CDT with CVD risk was not significant. Our findings confirm the established association between self-reported light to moderate alcohol consumption and reduced CVD risk. However, %CDT within the normal reference range may not be a risk indicator for CVD. [ABSTRACT FROM AUTHOR]