OBJECTIVES: To test the hypothesis that because of errors associated with growth and development, osteoporosis is frequently overdiagnosed in children when using dual-energy x-ray absorptiometry (DXA). This study compared bone density values obtained by DXA with those from computed tomography (CT), which is not influenced by body or skeletal size. STUDY DESIGN: Vertebral bone density was measured by using both DXA and CT in 400 children (100 each, healthy and sick boys and girls). Regression analysis was used to compare DXA and CT Z scores, and the agreement between DXA and CT classifications of Z scores below -2.0 was examined. RESULTS: DXA and CT Z scores were moderately related (r2 = 0.55 after accounting for age and anthropometric measures). DXA Z scores predicted CT Z scores below -2.0 with reasonable sensitivity (72%), specificity (85%), and negative predictive value (98%), but positive predictive value was low (24%). Many more subjects were classified as having bone density lower by DXA (76/400) than by CT (25/400), particularly subjects below the 5 th percentile of height and/or weight for age. CONCLUSIONS: The inability of DXA to account for the large variability in skeletal size and body composition in growing children greatly diminishes the accuracy of this projection technique for assessing bone acquisition and diagnosing osteoporosis in pediatric populations.
OBJECTIVES: To test the hypothesis that because of errors associated with growth and development, osteoporosis is frequently overdiagnosed in children when using dual-energy x-ray absorptiometry (DXA). This study compared bone density values obtained by DXA with those from computed tomography (CT), which is not influenced by body or skeletal size. STUDY DESIGN: Vertebral bone density was measured by using both DXA and CT in 400 children (100 each, healthy and sick boys and girls). Regression analysis was used to compare DXA and CT Z scores, and the agreement between DXA and CT classifications of Z scores below -2.0 was examined. RESULTS: DXA and CT Z scores were moderately related (r2 = 0.55 after accounting for age and anthropometric measures). DXA Z scores predicted CT Z scores below -2.0 with reasonable sensitivity (72%), specificity (85%), and negative predictive value (98%), but positive predictive value was low (24%). Many more subjects were classified as having bone density lower by DXA (76/400) than by CT (25/400), particularly subjects below the 5 th percentile of height and/or weight for age. CONCLUSIONS: The inability of DXA to account for the large variability in skeletal size and body composition in growing children greatly diminishes the accuracy of this projection technique for assessing bone acquisition and diagnosing osteoporosis in pediatric populations.
Authors: A Tsampalieros; M K Berkenstock; B S Zemel; L Griffin; J Shults; J M Burnham; R N Baldassano; M B Leonard Journal: Osteoporos Int Date: 2014-04-24 Impact factor: 4.507
Authors: Jason M Nagata; Jennifer L Carlson; Neville H Golden; Jin Long; Stuart B Murray; Rebecka Peebles Journal: Int J Eat Disord Date: 2019-02-16 Impact factor: 4.861
Authors: Tishya Al Wren; David C Lee; Robert M Kay; Frederick J Dorey; Vicente Gilsanz Journal: Dev Med Child Neurol Date: 2010-12-17 Impact factor: 5.449
Authors: E Brannon Morris; John Shelso; Matthew P Smeltzer; Nicole A Thomas; E Jane Karimova; Chin-Shang Li; Thomas Merchant; Amar Gajjar; Sue C Kaste Journal: Pediatr Radiol Date: 2008-09-04