Share this article on:


Milliken, L A.1; Faigenbaum, A D.1; Loud, R Larosa1; Westcott, W L.1

Medicine & Science in Sports & Exercise: May 2003 - Volume 35 - Issue 5 - p S81
B-13L Free Communication/Poster Bone and Connective Tissue

1University of Massachusetts Boston, Boston, MA and the South Shore YMCA, Quincy. MA

Dpd has become routine in estimating bone resorption. In children, where bone is rapidly changing, it is important to determine the reliability of Dpd measurements because the ability to detect long-term changes in bone metabolism depends on the day-to-day variability of Dpd measurements.

Back to Top | Article Outline


To determine the reliability of Dpd measurements in children ages 7 to 12 years.

Back to Top | Article Outline


The present study included subjects taking part in an 8-week strength training study. Fasting first morning urine samples were collected from 14 girls and 26 boys twice within a 3 day period both at baseline and after 8 weeks. Eight girls and 7 boys did not collect a urine sample on at least one of the four data collection days. The present analyses include 65 observations because the baseline and follow-up samples were combined. Dpd, corrected for creatinine excretion (Cr), was measured by immunoassay using commercially avaliable kits (Quidel Corp., Mountain View, CA). Dpd on day 1 versus day 2 was compared using regression analysis and paired t-tests. Bland-Altman plots (day 2 minus day 1 regressed against the mean Dpd) were also performed to determine whether day 1 and 2 differences were affected by the magnitude of Dpd excretion.

Back to Top | Article Outline


Bone resorption [mean(SD)] for day 1 and day 2 was 25.9 (7.6) and 25.2 (7.3) nmol Dpd/mmol Cr, respectively. The mean difference was 0.63 (7.2) nmol Dpd/mmol Cr, which was not significant (p = 0.49). The day-to-day differences ranged from −15.5 to 14.3 nmol Dpd/mmol Cr. The correlation was 0.53 (p = 0.00) and the standard error of estimate was 6.5 nmol Dpd/mmol Cr. The Bland-Altman plot regression coefficient (r = 0.05) was not significant.

Back to Top | Article Outline


Though no significant mean differences existed between days and the differences were not affected by the magnitude of Dpd, the relationship between days 1 and 2 was only moderate. Reliability coefficients should be > 0.90 (vs 0.53). Therefore, a single urine sample may not provide a stable measurement of Dpd in children.

©2003The American College of Sports Medicine