Plasma alpha1-acid glycoprotein can be used to adjust inflammation-induced hyporetinolemia in vitamin A-sufficient, but not vitamin A-deficient or -supplemented rats.

We examined the association between alpha(1)-acid glycoprotein (AGP), all-trans-retinol (retinol), and albumin concentrations in a longitudinal animal model of IL-6-induced inflammation. Vitamin A-sufficient (VAS) male Sprague-Dawley rats were administered recombinant human IL-6 [n = 4, 65 mug/(kg.d)] or PBS (n = 4) continuously for 7 d via osmotic minipumps. Plasma samples were obtained daily and concentrations of retinol, AGP, albumin, and total protein were measured. Compared with both baseline and controls, retinol and albumin decreased (P < 0.05), AGP increased (P < 0.05), and total protein concentrations were unaffected in IL-6-treated rats. In vitamin A-deficient (VAD) rats, AGP concentrations were significantly lower at all time points and increased only to one-third of that in VAS rats. The AGP cut-off value indicative of inflammation was 0.11 g/L (i.e., 95% upper limit of baseline concentrations). After 20.5 h, there was an inverse linear correlation between AGP concentrations and the relative change in retinol to baseline (y = -0.18x + 0.48, r = -0.84, P < 0.001). However, changes in AGP and albumin were not correlated (P = 0.94). The application of this function to retinol concentrations in rats from separate experiments showed that hyporetinolemia cannot be adjusted using plasma AGP in VAD or vitamin A-supplemented rats. In conclusion, correcting inflammation-induced hyporetinolemia using an acute-phase protein requires longitudinally derived data, knowledge of vitamin A status, and a common underlying mechanism of change.