Impaired Trabecular Microarchitecture at the Distal Radius and Tibia Is Associated with Lower Strength Estimates in Adolescents with Anorexia Nervosa

Presentation Number: MON 332
Date of Presentation: April 3rd, 2017

Vibha Singhal*1, Shreya Tulsiani2, Meghan Slattery2, Madhusmita Misra1 and Anne Klibanski3
1Massachusetts General Hospital/Harvard Medical School, Boston, MA, 2Massachusetts General Hospital, Boston, MA, 3Massachusetts General Hospital and Harvard Medical School, Boston, MA


Background: We have shown using standard analysis of high resolution peripheral quantitative CT (HRpQCT) data that trabecular bone at the radius is adversely affected in adolescents with anorexia nervosa (AN), and impaired trabecular parameters are associated with lower strength estimates. Individual trabeculae segmentation (ITS) is a method used to further characterize trabecular morphology, orientation and connectivity. There are no data regarding the effect of AN on trabecular microarchitecture at the non-weight bearing radius and weight bearing tibia using ITS.

Objective: To compare trabecular microarchitecture at the distal radius and tibia using ITS in adolescent girls with AN and normal weight controls (C). We hypothesized that in AN, ITS measures of trabecular microarchitecture would be affected adversely, and impact strength estimates.

Methods: 85 subjects, 46 with AN (per DSM-5 criteria) and 39 C 14-22 years old were enrolled. All participants underwent distal radius and tibia HRpQCT, ITS and finite element analysis.

Results: Groups did not differ for age, bone/menarchal/gynecologic age or height. AN had lower BMI, fat and lean mass than C. At the radius, AN had lower stiffness [65.1 (56.7-77.3) vs. 77.3 (68.4-87.7) N/mm, p=0.0008] than C. AN also had lower trabecular BV/TV (0.13±0.03 vs. 0.14±0.03, p=0.04), and greater trabecular separation [0.45 (0.41-0.49) vs. 0.42 (0.38-0.48) mm, p=0.03]. On ITS, plate trabecular thickness [0.201±0.009 vs. 0.205±0.008 mm, p=0.04], plate-rod junction density (3.8±0.9 vs. 4.5±1.2 1/mm3, p=0.008), and plate-plate junction density (1.9±0.6 vs. 2.2±0.7 1/mm3, p=0.01) were lower in AN; rod parameters did not differ. Within AN, plate BV/TV and thickness correlated with stiffness (ρ= 0.76 and 0.81, respectively, p<.0001 for both). Correlations of plate thickness with stiffness were stronger in AN than C (ρ= 0.81 vs. 0.58; p= 0.04).

At the tibia, AN had lower stiffness (197.2±33.1 vs. 214.3±35.0 N/mm, p=0.02) than C. AN also had lower trabecular number (1.8±0.23 vs. 1.9±0.21 1/mm, p=0.002) and greater trabecular separation [0.48 (0.44-0.52) vs. (0.43 (0.40-0.46) mm, p=0.001]. On ITS, AN had lower BV/TV (0.29±0.03 vs. 0.31±0.04, p=0.04), plate-rod junction density (3.93±0.72 vs. 4.34±0.83 1/mm3, p=0.01), and plate-plate junction density [2.52 (2.11-2.73) vs. 2.72 (2.24-3.01) 1/mm3, p=0.03]; rod parameters did not differ. Plate number and plate-plate connectivity correlated positively with stiffness (ρ= 0.79 and 0.78, respectively, p<.0001 for both).

Conclusions: Adolescents/young adults with AN have altered trabecular morphology and connectivity at the distal radius and tibia. Trabecular plate parameters are stronger determinants of strength estimates than rod parameters. This further delineates the site of trabecular bone that is more affected in AN and may help in targeting and monitoring the effect of therapy.


Nothing to Disclose: VS, ST, MS, MM, AK