Fracture risk in diabetes: insights from a study of bone microarchitecture

Previous literature suggested an increase in the risk of fractures in diabetes. The risk seemed to be higher in T1D than in T2D but the effect of several features such as fracture site, gender, age, BMI and diabetes-related features such as DM duration, insulin use and the presence of complications has not been fully explored. This thesis investigated the risk of fractures in diabetes. The first meta-analysis (chapter 3) investigated the risk of hip and non-vertebral fractures in diabetes and how this risk was affected by several features associated with the patients and the disease. A significant increase in the risk of fracture in diabetes was found both for hip (RR 1.52, 95% CI 1.42-1.63) and for non-vertebral fracture (RR 1.20, 1.14-1.27). The increase in the risk was greater for insulin users and longer diabetes duration, at both sites. At the hip, the risk was higher in the younger population, women, and those with T1D.
The second meta-analysis (chapter 4) investigated the risk of peripheral fractures in diabetes, since the wrist and ankle are the sites assessed by HR-pQCT. There was a discordant pattern and while at the wrist the risk of fractures was decreased (RR 0.85 95% CI 0.77 – 0.95) at the ankle the risk was increased (RR 1.30 95%CI 1.15 – 1.48). The sample included mainly T2D participants and the pattern was similar to the risk pattern observed in obesity.
Finally, a clinical study was conducted to assess axial DXA and peripheral microarchitecture in patients with type 1 diabetes with and without neuropathy. HR-pQCT was used to evaluate the standard site and also a less distal (14% limb length) site. There was no difference in DXA at lumbar spine or proximal femur between the groups. On HR-pQCT, the 14% site showed preserved trabecular structure particularly in the group without neuropathy and no abnormalities in the cortical compartment in the diabetic groups. At the standard site, cortical porosity was increased in the group with diabetes and neuropathy at the tibia. However, there were no differences in bone strength estimated by finite element analysis. Since bone turnover markers are decreased in diabetes, bone turnover is suppressed. This could suggest that the bone turnover suppression could prevent bone loss and preserve trabecular microarchitecture. Conversely, cortical porosity was increased only at the tibia in the group with neuropathy. This finding suggested that vascular and/or neural integrity might also be important to bone remodelling and consequently, bone microarchitecture.
In summary, there was an increase in the risk of hip, non-vertebral and ankle fractures and a decrease in the risk of wrist fracture in diabetes. Our findings suggested that bone microarchitecture is not the main determinant of this increase in the risk of fractures.