J Clin Densitom. 2020 Oct 24;S1094-6950(20)30134-7. doi: 10.1016/j.jocd.2020.10.008.Online ahead of print.

Yael Levy-Shraga ¹, Ophir Megnazi ², Dalit Modan-Moses ³, Liana Tripto-Shkolnik ⁴, Noah Gruber ³, Yael Haberman ⁵, Dror S Shouval ⁵, Batia Weiss ⁵

Author information

• ¹Pediatric Endocrinology and Diabetes Unit, The Edmond and Lily Safra Children's Hospital, Sheba Medical Center, Tel-Hashomer, Israel; The Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel. Electronic address: yael.levy.shraga@gmail.com.
• ²The Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel.
• ³Pediatric Endocrinology and Diabetes Unit, The Edmond and Lily Safra Children's Hospital, Sheba Medical Center, Tel-Hashomer, Israel; The Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel.
• ⁴The Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel; Division of Endocrinology, Diabetes and Metabolism, Sheba Medical Center, Tel-Hashomer, Israel.
• ⁵The Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel; Division of Pediatric Gastroenterology and Nutrition, Edmond and Lily Safra Children's Hospital, Sheba Medical Center, Tel Hashomer, Israel.

Abstract

Introduction: Trabecular bone score (TBS) is a textural index that evaluates bone microarchitecture of the lumbar spine. Our aim was to assess TBS in children with inflammatory bowel diseases and to evaluate correlations with clinical, laboratory and densitometric variables.

Methods: A retrospective study of TBS and areal bone mineral density measurements by dual-energy X-ray absorptiometry (DXA) of children with either Crohn's disease (CD) or ulcerative colitis (UC). Bone mineral apparent density was calculated for size adjustment. TBS Z-score for each child were calculated based on data from a healthy population of similar age and gender distribution. Variables significantly associated with TBS were included in stepwise linear regression models to examine independent predictors of TBS.

Results: Fifty patients (age at DXA scan 13.8 ± 3.0 years, 29 males) were included. No significant differences were observed between the patients with CD and UC, in age at diagnosis, age at DXA scan and disease duration. The mean TBS of patients with CD (n = 35) was lower than of patients with UC (n = 15): 1.340 ± 0.080 vs 1.395 ± 0.092, p = 0.040. The mean TBS Z-score of patients with CD, -0.443 ± 0.788, was significantly lower than expected in healthy children (p = 0.002), while the mean TBS Z-score of patients with UC, 0.231 ± 0.685, was similar to that of healthy children (p = 0.212). In the stepwise linear regression analysis, BMI Z-score at diagnosis, phosphorus level at diagnosis and age at the time of the DXA scan were significant independent predictors of TBS (r² = 0.604; β = 0.037, 95% confidence interval (CI) for β 0.022-0.051, p < 0.001; β = 0.045, 95% CI: 0.017-0.073, p = 0.002; and β = 0.031, 95% CI: 0.005-0.021, p < 0.002, respectively).

Conclusions: TBS is significantly reduced in pediatric patients with CD but not in patients with UC. This finding likely reflects the effect of CD on bone microarchitecture.