Open Access Open Badges Research article

Bone mass and quality in patients with juvenile idiopathic arthritis: longitudinal evaluation of bone-mass determinants by using dual-energy x-ray absorptiometry, peripheral quantitative computed tomography, and quantitative ultrasonography

Stefano Stagi1, Loredana Cavalli2, Carla Signorini2, Federico Bertini3, Marco Matucci Cerinic3, Maria Luisa Brandi2 and Fernanda Falcini3*

  • * Corresponding author: Fernanda Falcini

  • † Equal contributors

Author Affiliations

1 Health’s Sciences Department, University of Florence, Florence, Italy

2 Department of Internal Medicine, Endocrinology Unit, University of Florence, Florence, Italy

3 Department of Internal Medicine, Section of Rheumatology, Transition Clinic, University of Florence, Viale Pieraccini 18, 50139 Florence, Italy

For all author emails, please log on.

Arthritis Research & Therapy 2014, 16:R83  doi:10.1186/ar4525

Published: 31 March 2014



Our objective was to evaluate longitudinally the main bone-mass and quality predictors in young juvenile idiopathic arthritis (JIA) patients by using lumbar spine dual-energy X-ray absorptiometry (DXA) scan, radius peripheral quantitative computed tomography (pQCT), and phalangeal quantitative ultrasonography (QUS) at the same time.


In total, 245 patients (172 females, 73 males; median age, 15.6 years: 148 oligoarticular, 55 polyarticular, 20 systemic, and 22 enthesitis-related-arthritis (ERA) onset) entered the study. Of these, 166 patients were evaluated longitudinally. Data were compared with two age- and sex-matched control groups.


In comparison with controls, JIA patients, but not with ERA, had a reduced spine bone-mineral apparent density (BMAD) standard deviation score (P < 0.001) and musculoskeletal deficits, with significantly lower levels of trabecular bone mineral density (TrabBMD) (P < 0.0001), muscle cross-sectional area (CSA) (P < 0.005), and density-weighted polar section modulus (SSIp) (P < 0.05). In contrast, JIA showed fat CSA significantly higher than controls (P < 0.0001). Finally, JIA patients had a significant reduced amplitude-dependent speed of sound (AD-SoS) (P < 0.001), and QUS z score (P < 0.005).

Longitudinally, we did not find any difference in all JIA patients in comparison with baseline, except for the SSIp value that normalized. Analyzing the treatments, a significant negative correlation among spine BMAD values, TrabBMD, AD-SoS, and systemic and/or intraarticular corticosteroids, and a positive correlation among TNF-α-blocking agents and spine BMAD, TrabBMD, and AD-SoS were observed.


JIA patients have a low bone mass that, after a first increase due to the therapy, does not reach the normal condition over time. The pronounced bone deficits in JIA are greater than would be expected because of reduction in muscle cross-sectional area. Thus, bone alterations in JIA likely represent a mixed defect of bone accrual and lower muscle forces.