Significant bone microarchitecture impairment in premenopausal women with active celiac disease.

Patients with active celiac disease (CD) are more likely to have osteoporosis and increased risk of fractures. High-resolution peripheral quantitative computed tomography (HR-pQCT) permits three-dimensional exploration of bone microarchitectural characteristics measuring separately cortical and trabecular compartments, and giving a more profound insight into bone disease pathophysiology and fracture. We aimed to determine the volumetric and microarchitectural characteristics of peripheral bones-distal radius and tibia-in an adult premenopausal cohort with active CD assessed at diagnosis. We prospectively enrolled 31 consecutive premenopausal women with newly diagnosed CD (median age 29 years, range: 18-49) and 22 healthy women of similar age (median age 30 years, range 21-41) and body mass index. Compared with controls, peripheral bones of CD patients were significantly lower in terms of total volumetric density mg/cm(3) (mean ± SD: 274.7 ± 51.7 vs. 324.7 ± 45.8, p 0.0006 at the radius; 264.4 ± 48.7 vs. 307 ± 40.7, p 0.002 at the tibia), trabecular density mg/cm(3) (118.6 ± 31.5 vs. 161.9 ± 33.6, p<0.0001 at the radius; 127.9 ± 28.7 vs. 157.6 ± 15.6, p < 0.0001 at the tibia); bone volume/trabecular volume ratio % (9.9 ± 2.6 vs. 13.5 ± 2.8, p<0.0001 at the radius; 10.6 ± 2.4 vs. 13.1 ± 1.3, p < 0.0001 at the tibia); number of trabeculae 1/mm (1.69 ± 0.27 vs. 1.89 ± 0.26, p 0.009 at the radius; 1.53 ± 0.32 vs. 1.80 ± 0.26, p 0.002 at the tibia); and trabecular thickness mm (0.058 ± 0.010 vs. 0.071 ± 0.008, p < 0.0001 at the radius with no significant difference at the tibia). Cortical density was significantly lower in both regions (D comp mg/cm(3) 860 ± 57.2 vs. 893.9 ± 43, p 0.02; 902.7 ± 48.7 vs. 932.6 ± 32.6, p 0.01 in radius and tibia respectively). Although cortical thickness was lower in CD patients, it failed to show any significant inter-group difference (a-8% decay with p 0.11 in both bones). Patients with symptomatic CD (n = 22) had a greater bone microarchitectural deficit than those with subclinical CD. HR-pQCT was used to successfully identify significant deterioration in the microarchitecture of trabecular and cortical compartments of peripheral bones. Impairment was characterized by lower trabecular number and thickness-which increased trabecular network heterogeneity-and lower cortical density and thickness. In the prospective follow-up of this group of patients we expect to be able to assess whether bone microarchitecture recovers and to what extend after gluten-free diet.