Relationships Between Serum Retinol Binding Protein4 (RBP-4) And Bone Metabolism In Patients With Type 2 Diabetes Mellitus

Background and aims Type 2 diabetes mellitus (T2DM) is a pandemic metabolic disease with substantial morbility and mortality. Bone and mineral abnormalities in patients with diabetes mellitus may be caused by many mechanisms, including adipokine productions. Retinol binding protein 4 (RBP-4) is secreted by adipocytes, besides the liver. RBP-4 may be the potential contributor of bone mass density or bone mineral metabolism in patients with T2DM. The aim of the present study was to investigate the relationships between serum concentration of RBP-4 and bone mass density, markers of bone metabolism in patients with T2DM.Materials and methods 82 patients with newly diagnosed T2DM and 46 subjects with normal glucose tolerance (NGT) participated in the cross-sectional study. Additionally, subset analyses were performed, dividing subjects on the basis of gender into M-T2DM, F-T2DM, M-NGT and F-NGT. Serum concentrions of RBP-4, osteocalcin(OC) and C-terminal telopeptide of collagen type l(CTX) were measured with ELISA, then Napierian logarithmic transformation (ln) were made before statistical analysis. BMD value were measured using dual-energy-x-ray absorptiometry(DXA) with a Hologic QDR4500A device (Hologic, Waltam, MA, USA) at the anteroposterior spine (AP), femeral neck (FN), total hip (T-hip) and total body BMD (TBMD); Body composition including lean tissue mass (LTM) and fat tissue mass (FTM) were measured by DXA.Results (1) BMD measured at any site in T2DM had no significant difference compared with that in NGT. (2) Both lnOC and lnCTX had no significant difference between T2DM and NGT. (3) In the group of NGT, lnRBP-4 had no correlation with lnOC or lnCTX. No regression model described the change in lnRBP-4 with lnOC or lnCTX well. In a multivariate analysis, lnRBP-4 was not the independent determiner of lnOC or lnCTX. (4) In the group of T2DM, regardless of gender, lnRBP-4 showed a positive relationship with lnCTX (M-T2DM, r= 0.564, P=0.000; F-T2DM, r= 0.386, P=0.018), but had no association with lnOC, After adjusting for age, smoking, creatinine clearance rate (Ccr) and waist/hip ratio (WHR), lnRBP-4 still showed a strong association with lnCTX in the groups of T2DM (r'=0.442, P=0.000) and M-T2DM (r'=0.536, P=0.000), while the relationship disappeared in F-T2DM (r'=0.317, P=0.072) after adjusting for WHR . In the total subjects , the logarithm regression model described the best change in lnRBP-4 with lnCTX (R~2=0.144, P= 0.000), so did in the group of T2DM (R~2=0.260, P=0.000 ). The linear regression model described the best change in lnRBP-4 with lnCTX in the group of M-T2DM (R~2=0.318, P=0.000 ), but in the group of F-T2DM the best model was the inverse (R~2=0.255, P=0.001). In a multivariate analysis, lnRBP-4 and gender were the independent determiners of lnCTX, explaining 28.4% of the variance of lnCTX. As in the group of NGT, lnRBP-4 was not the independent determiner of lnOC in the group of T2DM. (5) lnRBP-4 showed no association with BMDs in the group of both NGT and T2DM. Additionally, lnRBP-4 was not the independent determiner of BMDs in both groups in a multivariate analysis.Conclusion (1) There were no significant difference in BMD measured at any site between groups of newly diagnosed T2DM and NGT. (2) There were no significant difference in Both lnOC and lnCTX between the two groups. (3) In the group of newly diagnosed T2DM, lnRBP-4 showed a positive relationship with lnCTX and was the independent determiner of lnCTX, but had no relationship with lnOC. In the group of NGT, lnRBP-4 showed no correlation with lnCTX or lnOC. (4) In both groups of newly diagnosed T2DM and NGT, lnRBP-4 showed no correlation with BMD measured at any site in the two groups.