IMPORTANCE Murine studies reveal that sympathetic nervous system activation leads to decreased bone mass. Stimulant medications used to treat attention-deficit/hyperactivity disorder (ADHD) increase sympathetic tone and may affect bone remodeling. Because bone mass accrual is completed by young adulthood, assessing stimulant effects on bone density in growing children is of critical importance. OBJECTIVE To investigate associations between stimulant use and bone mass in children and adolescents. DESIGN, SETTING, AND PARTICIPANTS This cross-sectional analysis used data collected from January 1, 2005, to December 31, 2010, from the National Health and Nutrition Examination Survey (NHANES) database. NHANES is a series of cross-sectional, nationally representative health and nutrition surveys of the US population. All children, adolescents, and young adults aged 8 to 20 years with dual-energy x-ray absorptiometry (DXA), anthropometric, demographic, and prescription medication use data were eligible for participation. Of the 6489 respondents included in the multivariable linear regression analysis, 159 were stimulant users and 6330 were nonusers. Data were analyzed from October 8, 2015, to December 31, 2016. EXPOSURES Stimulant use, determined by questionnaires administered via interview. MAIN OUTCOMES AND MEASURES The association between stimulant use and total femur, femoral neck, and lumbar spine bone mineral content (BMC) and bone mineral density (BMD) was assessed using DXA. RESULTS Study participants included 6489 NHANES participants with a mean (SD) age of 13.6 (3.6) years. Stimulant use was associated with lower bone mass after adjustment for covariates. Mean lumbar spine BMC was significantly lower in stimulant users vs nonusers (12.76 g; 95% CI, 12.28-13.27 g vs 13.38 g; 95% CI, 13.26-13.51 g; P = .02), as was mean lumbar spine BMD (0.90 g/cm 2 ; 95% CI, 0.87-0.94 g/cm 2 vs 0.94 g/cm 2 ; 95% CI, 0.94-0.94 g/cm 2 ; P = .03) and mean femoral neck BMC (4.34 g; 95% CI, 4.13-4.57 g vs 4.59 g; 95% CI, 4.56-4.62 g; P = .03). Mean BMD of the femoral neck (0.88 g/cm 2 ; 95% CI, 0.84-0.91 g/cm 2 vs 0.91 g/cm 2 ; 95% CI, 0.90-0.91 g/cm 2 ; P = .08) and total femur (0.94 g/cm 2 ; 95% CI, 0.90-0.99 g/cm 2 vs 0.99 g/cm 2 ; 95% CI, 0.98-0.99 g/cm 2 ; P = .05) were also lower in stimulant users vs nonusers. Participants treated with stimulants for 3 months or longer had significantly lower lumbar spine BMD (0.89 g/cm 2 ; 95% CI, 0.85-0.93 g/cm 2 vs 0.94 g/cm 2 ; 95% CI, 0.94-0.94 g/cm 2 ; P = .02) and BMC (12.71 g; 95% CI, 12.14-13.32 g vs 13.38 g; 95% CI, 13.25-13.51 g; P = .03) and femoral neck BMD (0.87 g/cm 2 ; 95% CI, 0.74-0.83 g/cm 2 vs 0.91 g/cm 2 ; 95% CI, 0.83-0.84 g/cm 2 ; P = .048) than nonusers. CONCLUSIONS AND RELEVANCE Children and adolescents reporting stimulant use had lower DXA measurements of the lumbar spine and femur compared with nonusers. These findings support the need for future prospective studies to examine the effects of stimulant use on bone mass in children.

No abstract

particular model of accelerometer overestimated step count by 5265 steps per day in free-living condition. 2 However, the model of accelerometer used in this study was not reported nor was the compliance rate (eg, mean wear/nonwear time per day), which may affect the validity of measurement.Second, step count may not sufficiently reflect the true physical activity level of patients following concussion. Patients with concussion may have fewer walking-based activities (1815 steps per day as reported in this study), and their physical activity may have been restricted to low-intensity, home-based, multidimensional body movements. In this connection, using an accelerometer to capture multidimensional body movements and express these as "activity intensity" would better reflect the true physical activity level. However, "activity intensity" was not reported in this study, despite using an accelerometer.Third, the patients were aged 11 to 19 years and were adolescents and young adults. The body of evidence shows that the physical activity level of adolescents and adults are different. 3 The recovery rate following concussion in adolescents and adults is also different. 4 Therefore, it calls into question whether it is appropriate to include both adolescents and adults in the data analyses.