Bone properties in persons with type 1 diabetes and healthy controls – A cross-sectional study

Abstract

Background

The risk of fractures is increased in persons with type 1 diabetes (T1D) and assessment of bone health has been included in the 2024 updated Standards of Care by The American Diabetes Association (ADA). Previous studies have found that in T1D bone metabolism, mineral content, microstructure, and strength diverge from that of persons without diabetes. However, a clear description of a T1D bone phenotype has not yet been established. We investigated bone mechanical properties and microstructure in T1D compared with healthy controls. For the potential future introduction of additional bone measures in the clinical fracture risk assessment, we aimed to assess any potential associations between various measures related to bone indices in subjects with T1D.

Methods

We studied human bone indices in a clinical cross-sectional setup including 111 persons with early-onset T1D and 37 sex- and age-matched control persons. Participants underwent hip and spine DXA scans for bone mineral density (BMD) of the femoral neck (FN), total hip (TH), and lumbar spine (LS), and TBS evaluation, microindentation of the tibial shaft for Bone Material Strength index (BMSi), and high-resolution periphery quantitative computed tomography (HRpQCT) of the distal radius and tibia for volumetric BMD (vBMD) and structural measures of trabecular and cortical bone. Results are reported as means with (standard deviation) or (95% confidence intervals (CI)), medians with [interquartile range], and differences are reported with (95% CI).

Results

The study included 148 persons aged 20 to 75 years with a median age of 43.2 years. The T1D group who had all been diagnosed with T1D before the age of 18 years demonstrated values of HbA1c ranging from 39 to 107 mmol/mol and a median HbA1c of 57 mmol/mol. The BMD did not differ between groups (the mean difference in FN-BMD was 0.026 g/cm² (−0.026; 0.079), p = 0.319) and the median BMSi was comparable in the two groups (79.2 [73.6; 83.8] in the T1D group compared with 77.9 [70.5, 86.1] in the control group). Total and trabecular vBMD (Tb.vBMD), cortical thickness (Ct.Th), and trabecular thickness (Tb.Th) of both radius and tibia were lower in participants with T1D. The mean Tb.vBMD at the radius was 143.6 (38.5) mg/cm³ in the T1D group and 171.5 (37.7) mg/cm³ in the control group, p < 0.001. The mean Ct. Th^d of the radius was 0.739 mm (0.172) in the T1D group and 0.813 (0.188) in the control group, p = 0.044. Crude linear regressions revealed limited agreement between BMSi and Tb.vBMD (p = 0.010, r² = 0.040 at the radius and p = 0.008, r² = 0.040 at the tibia and between BMSi and the estimated failure load (FL) at the tibia (p < 0.001, r² = 0.090). There were no significant correlations between BMSi and Ct.Th. TBS correlated with Tb.vBMD at the radius (p = 0.008, r² = 0.044) and the tibia (p = 0.001, r² = 0.069), and with the estimated FL at the distal tibia (p = 0.038, r² = 0.026).

Conclusion

In this study, we examined the bones of persons with well-controlled, early-onset T1D. Compared with sex- and age-matched healthy control persons, we found reduced total and trabecular vBMD, as well as decreased trabecular and cortical thickness. These results suggest that a debut of T1D before reaching peak bone mass negatively impacts bone microarchitecture. No differences in areal BMD or BMSi were observed. Although the variations in total hip BMD reflect some variation in the vBMD, the reduction in trabecular bone mineral density was not captured by the DXA scan. Consequently, fracture risk may be underestimated when relying on DXA, and further research into fracture risk assessment in T1D is warranted.