Cohort study shows that peripheral dual-energy X-ray absorptiometry is of limited epidemiologic use in prepubertal children

• Autor(a) principal: Martins, Ana
• Data de Publicação: 2017
• Outros Autores: Monjardino, Teresa, Canhão, Helena, Lucas, Raquel
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10362/151785
• topic: Birth cohort; Bone mineral density; Diagnostic accuracy; Fracture; Peripheral densitometry; Pediatrics, Perinatology, and Child Health

Aim: Peripheral methods are increasingly used to assess bone health, despite little evidence on their predictive ability. We aimed to evaluate the usefulness of forearm dual-energy X-ray absorptiometry in prepubertal children, by estimating the agreement between peripheral and central measures and the ability to predict fracture history. Methods: In 2012/2014, we assessed 1177 seven-year-old children from the Generation XXI cohort who were recruited at birth in all five public hospitals with maternity wards in Porto, Portugal. Subtotal and lumbar spine bone mineral density (BMD) and content, left-arm BMD and peripheral forearm BMD were measured. Parents reported the child's lifetime fracture history. We estimated agreement using Bland–Altman's method and Cohen's kappa. Fracture prediction ability was calculated using area under the receiver operator characteristic curve (ROC-AUC). Results: The limits of agreement were very wide, ranging from −2.20/2.20 to −1.87/1.87 standard deviations for the comparison between peripheral and central measures. Categorical agreement was also poor, with all kappa values below 0.40. In addition, none of the measures predicted fractures, because all the ROC-AUCs were close to 0.50. Conclusion: This study suggests that forearm BMD has limited use for bone health research or as a basis for clinical decisions in prepubertal children.