Trabecular Bone Score has Poor Association With pQCT Derived Trabecular Bone Density in Indian Children With Type 1 Diabetes and Healthy Controls

Background: In children with type 1 diabetes mellitus (T1DM), low trabecular volumetric bone mineral density (Trab vBMD) has been reported. However, studies using the trabecular bone score (TBS) are scarce. The objective of our study was to assess areal bone mineral density at the lumbar spine (LS aBMD), the TBS and Trab vBMD in children with type 1 diabetes in comparison with healthy controls and to assess the relationship of Trab vBMD with TBS. Methods: A total of 205 children were assessed for their LS bone mineral content (BMC) and LS aBMD by dual energy x-ray absorptiometry (DXA) and Trab vBMD at distal radius by peripheral quantitative computed tomography (pQCT). Machine generated Z-scores for both LS aBMD and Trab vBMD were used. The retrospective DXA LS scans in children with T1DM (n=137, age 13.1 +/- 3.2 years) and controls (n = 68, age 13.0 +/- 2.7 years) were analysed with a research trial version of TBS iNsight software (Medimaps Group). The established TBS cut-offs were used to categorize TBS. Results: The mean LS BMC, LS aBMD, TBS and Trab vBMDs were lower in children with T1DM. TBS was positively correlated with LS aBMD but not with Trab vBMD in both groups. Distribution of T1DM and control children was similar in the TBS categories. Over a fourth of the T1DM children with low Trab vBMD (below-2 Z score) had normal TBS, while, in children with LS aBMD Z-score >-2 from both groups, >50% had degraded or partially degraded TBS. Degraded TBS was seen in half the control children although none of them had low Trab vBMD. Conclusion: We found poor correlation between TBS and Trab vBMD in paediatric diabetic and healthy population. Our results also suggest establishing paediatric TBS cut offs in improving the classification of children having degraded trabecular bone.