Aim: Patients with type 2 diabetes (T2DM) have more risk of bone fractures. However, areal bone mineral density (aBMD) by conventional dual-energy x-ray absorptiometry (DXA) is not useful for identifying this risk. This study aims to evaluate 3D-DXA parameters determining the cortical and trabecular compartments in patients with T2DM compared to non-diabetic subjects and to identify their determinants.
Materials and methods: Case-control study in 111 T2DM patients (65.4 ± 7.6 years old) and 134 non-diabetic controls (64.7 ± 8.6-year-old). DXA, 3D-DXA modelling via 3D-Shaper software and trabecular bone score (TBS) were used to obtain aBMD, cortical and trabecular parameters, and lumbar spine microarchitecture, respectively. In addition, biochemical markers as 25-hydroxyvitamin d, type I procollagen N-terminal propeptide (P1NP), C-terminal telopeptide of type I collagen (CTX), and glycated haemoglobin (HbA1c) were analysed.
Results: Mean-adjusted values showed higher aBMD (5.4%-7.7%, ES: 0.33-0.53) and 3D-DXA parameters (4.1%-10.3%, ES: 0.42-0.68) in the T2DM group compared with the control group. However, TBS was lower in the T2DM group compared to the control group (-14.7%, ES: 1.18). In addition, sex (β = 0.272 to 0.316) and body mass index (BMI) (β = 0.236 to 0.455) were the most consistent and positive predictors of aBMD (p ≤ 0.01). BMI and P1NP were negative predictors of TBS (β = -0.530 and -0.254, respectively, p ≤ 0.01), while CTX was a positive one (β = 0.226, p=0.02). Finally, BMI was consistently the strongest positive predictor of 3D-DXA parameters (β = 0.240 to 0.442, p<0.05).
Conclusion: Patients with T2DM present higher bone mass measured both by conventional DXA and 3D-DXA, suggesting that 3D-DXA technology is not capable of identifying alterations in bone structure in this population. Moreover, BMI was the most consistent determinant in all bone outcomes.