Effect of alcohol consumption on systemic markers of inflammation.

BACKGROUND: Epidemiological studies suggest that light to moderate alcohol intake is associated with lower all-cause mortality than abstention or heavy alcohol intake, primarily through reduced risk of coronary heart disease. The underlying mechanisms are incompletely understood.
METHODS: We investigated the association between alcohol consumption (assessed by a 7-day food record) and concentrations of C-reactive protein (CRP), alpha1-globulins, alpha2-globulins, albumin, and transferrin, and leucocyte count in a sample of 2006 men and women aged 18-88 years participating in a national health survey carried out in former West Germany in 1987-88. Analyses were based on 781 men and 995 women with complete data.
FINDINGS: Among men, alcohol consumption showed a U-shaped association with mean values of CRP (p for linear term 0.65, for quadratic term 0.048), alpha1-globulins (p=0.20, 0.0006), alpha2-globulins (p=0.82, 0.31), and leucocyte count (p=0.51, 0.26) even after adjustment for age, smoking, body-mass index, HDL and LDL cholesterol, history of hypertension, education, and income. There were inverted U-shaped associations between the negative acute-phase reactants albumin (p=0.41, 0.006) and transferrin (p=0.14, 0.28) and alcohol intake. In women, the associations were less strong for CRP (p=0.35, 0.31), leucocyte count (p=0.28, 0.15), and transferrin (p=0.86, 0.83). Concentrations of alpha1-globulins and alpha2-globulins were inversely related to alcohol consumption, and albumin showed a positive association with increasing alcohol intake in women.
INTERPRETATION: Non-drinkers and heavy drinkers had higher CRP concentrations than moderate drinkers. In view of the robust association between markers of inflammation, especially CRP, and risk of coronary heart disease, an anti-inflammatory action of alcohol could contribute to the link between moderate consumption and lower cardiovascular mortality.