Bone Accrual in Males with Autism Spectrum Disorder

Abstract

Objective: To test the hypothesis that bone accrual over a 4-year period is reduced in boys with autism spectrum disorder (ASD) compared with typically developing controls.

Study design: Twenty-five boys with ASD and 24 controls were assessed for bone outcomes. Fourteen boys with ASD and 11 controls were assessed both at baseline and after 4 years. The mean subject age was 11.0 ± 1.6 years at study initiation and 14.9 ± 1.6 years at follow-up. Bone mineral density (BMD) was measured at the spine, hip, and whole body using dual-energy X-ray absorptiometry and normalized for age, race, and sex (BMD z-scores). Height adjustments were performed as well. We assessed medical history, physical activity using questionnaires, vitamin D and calcium intake using food records, and serum calcium, phosphorus, 25(OH)-vitamin D, and pubertal hormone levels.

Results: Boys with ASD had lower spine, hip, and whole body BMD z-scores compared with controls. In those subjects assessed both at baseline and after 4 years, bone accrual rates did not differ between the 2 groups; however, spine and hip BMD z-scores remained lower in the boys with ASD than in controls at follow-up. Notably, the ASD group was less physically active at both time points.

Conclusion: Although pubertal bone accrual was similar to that in controls, BMD in children with ASD remained low over a 4-year follow-up period, suggesting that low BMD is a consequence of prepubertal factors, such as low physical activity. Studies are needed to investigate the causes and consequences of decreased BMD, to assess BMD in females and adults with ASD, and to evaluate therapeutic interventions.

Keywords: autism spectrum disorder; bone mineral density.

Figure 1

Cross-sectional BMD z-scores at the spine, femoral neck, and whole body less the head in boys with ASD vs controls. For the lumbar spine and femoral neck, BMD z-scores from participants’ baseline visits (2011 where applicable; 2015 for newly enrolled participants) are summarized. For the whole body less the head, BMD z-scores from the 2015 visit are summarized. BMD z-scores at all sites were significantly lower in boys with ASD compared with controls.

Figure 2

BMD z-scores at the spine, femoral neck, and whole body less the head by age in boys with ASD vs controls. Individual trajectories from baseline to follow-up are respresented by colored lines. An ordinary least squares regression line is presented in black. The boys with ASD had lower BMD z-scores compared with controls consistently across all ages.