Impaired metabolism of retinol has been shown to occur in alcohol-induced liver disease (ALD). The purpose of the present study was to investigate the saturation of retinol-binding protein (RBP) in 6 patients with different stages of ALD. Hospitalized alcohol consumers (n=118) with different stages of ALD (ALD1: mild stage of liver damage; ALD2: moderately severe changes of the liver with signs of hepatic inflammation; ALD3: severely impaired liver function) and 45 healthy control subjects were nutritionally assessed, and retinol and RBP content was measured in plasma by high-performance liquid chromatography and enzyme-linked immunosorbent assay methods, respectively. No differences were noted in daily retinol intake, but subjects with ALD had significantly lower concentrations of retinol in plasma (ALD1: 1.81+/-0.17 micromol/l [mean+/-S.E.M.]; ALD2: 1.95+/-0.24 micromol/l; ALD3: 0.67+/-0.13 micromol/l) compared to controls (2.76+/-0.19 micromol/l). Subjects of group ALD2 had significantly higher plasma RBP levels than controls (P<.05) and patients with ALD1 (P<.05) and ALD3 (P<.001). The relative saturation of RBP with retinol decreased with severity of ALD (controls: 76.8+/-5.0%; ALD1: 55.8+/-6.5%; ALD2: 43.5+/-6.2%; ALD3: 29.0+/-5.1%). The present study indicates that plasma concentrations of retinol and RBP per se do not correlate to severity of ALD, but rather that the retinol/RBP ratio links to the severity of alcohol-induced liver damage. From these results, a reduced availability of retinol in the periphery due to an altered saturation of RBP can be concluded.