Serum retinyl esters are not associated with biochemical markers of liver dysfunction in adult participants in the third National Health and Nutrition Examination Survey (NHANES III), 1988–1994

Abstract

Serum retinyl ester concentrations are elevated in hypervitaminosis A. It was suggested that retinyl esters >10% of total serum vitamin A indicate potential hypervitaminosis, but this cutoff was derived from small clinical samples that may not be representative of the general population. We sought to examine the distribution of serum retinyl ester concentrations and associations between retinyl ester concentrations and biochemical markers of liver dysfunction in a nationally representative sample. We assessed the associations between serum retinyl ester concentrations and 5 biochemical indexes of liver dysfunction by using multivariate linear and multiple logistic regression techniques and controlling for age, sex, use of supplements containing vitamin A, alcohol consumption, smoking status, and use of exogenous estrogens in 6547 adults aged > or =18 y in the third National Health and Nutrition Examination Survey (NHANES III), 1988--1994. Thirty-seven percent of the sample had serum retinyl ester concentrations >10% of total serum vitamin A and 10% of the sample had serum retinyl esters >15% of total vitamin A. We found no associations between serum retinyl ester concentrations and 1) concentrations of any biochemical variable (multiple linear regression) or 2) risk of having biochemical variables above the reference range (multiple logistic regression). We did not find a serum retinyl ester value with statistically significant sensitivity and specificity for predicting increases in biochemical indexes of liver dysfunction. The prevalence of serum retinyl ester concentrations >10% of the total vitamin A concentration in the NHANES III sample was substantially higher than expected but elevated retinyl ester concentrations were not associated with abnormal liver function.