Nutritional assessment of serum and hepatic vitamin A levels in patients with cirrhosis.

BACKGROUND: Serum vitamin A (retinol) levels may not correlate with hepatic vitamin A stores in patients with cirrhosis; thus, supplementation of vitamin A based on serum levels may have a detrimental effect. Our aim was to determine whether serum levels correlate with hepatic stores in cirrhotic patients.
METHODS: A prospective study of patients with cirrhosis undergoing orthotopic liver transplantation was completed. Serum and hepatic levels of vitamin A were measured by high-performance liquid chromatography. Statistical analysis was performed using rank sum tests and Spearman rank correlation coefficients.
RESULTS: Fifty cirrhotic patients (33 men and 17 women, mean age 53 years) were compared with a control group (25 men and 25 women, mean age 47 years) of liver donors. Median serum levels of retinol were 259 microg/L in controls and 166 microg/L in cirrhotic patients (p < .001). Median hepatic levels of retinol were 25 microg/g in controls and 27.5 +/- g/g in cirrhotic patients (p not significant). Total hepatic vitamin A levels (retinol plus retinyl esters) were 471 microg/g in controls and 244 microg/g in cirrhotic patients (p = .028). Serum retinol did not correlate with total hepatic vitamin A stores in cirrhotic patients (rs = .10, p = .332).
CONCLUSIONS: Serum retinol and total hepatic vitamin A stores are lower in cirrhotic patients than in controls. However, because levels of serum retinol do not correlate with hepatic vitamin A levels, the decision to prescribe vitamin A replacement for patients with cirrhosis should not be made solely on the basis of serum retinol levels.