Serum concentrations of retinol-binding protein 4 in women with and without gestational diabetes

Aims/hypothesis Pregnancy is characterised by temporarily increased insulin resistance. Gestational diabetes occurs when pancreatic beta cell function is unable to compensate for this insulin resistance. Retinol-binding protein 4 (RBP4) could be related to insulin resistance. We hypothesised that RBP4 is elevated in gestational diabetes. Methods Serum RBP4, transthyretin and retinol were cross-sectionally measured in 42 women with gestational diabetes and 45 pregnant controls. Of these, 20 women with and 22 without gestational diabetes were included in an additional longitudinal study. RBP4 was determined by enzyme immunometric assay (EIA) and western blot. Results Women with gestational diabetes had lower RBP4 EIA and western blot levels than controls (median 6.8 [interquartile range, 3.9-14.3] vs 11.3 [7.8-19.9] mu g/ml, p < 0.001 and 25.1 [21.7-29.6] vs 26.6 [23.5-32.2] mu g/ml, p=0.026). Transthyretin and the RBP4:transthyretin molar ratio were comparable between the groups. Serum retinol was lower (p < 0.001) and the RBP4 Western blot level: retinol molar ratio was higher in women with gestational diabetes (p=0.044). RBP4 was not associated with the glucose or homeostasis model assessment of insulin resistance (HOMA-IR), but in gestational diabetes the RBP4:retinol molar ratio correlated with blood glucose and negatively with 2 h post-load insulin. The RBP4:transthyretin ratio correlated with HOMA-IR and fasting insulin in controls. In women with gestational diabetes RBP4 EIA and western blot levels increased after delivery. Retinol increased in both groups, while transthyretin and the RBP4:transthyretin ratio were not altered after parturition. Conclusions/interpretation RBP4 measured by two different techniques is not elevated, but the RBP4:retinol molar ratio is higher and correlates with fasting blood glucose in women with gestational diabetes. Thus, the RBP4:retinol ratio and the RBP4:transthyretin ratio are more informative than RBP4 levels alone when assessing insulin-glucose homeostasis during pregnancy.