Searchable abstracts of presentations at key conferences on calcified tissues
Bone Abstracts (2013) 2 P18 | DOI: 10.1530/boneabs.2.P18

ICCBH2013 Poster Presentations (1) (201 abstracts)

Influence of anthropometric parameters on assessment of paediatric bone mineral density and bone mineral content

Thomas N Hangartner 1 , David F Short 1 , Vicente Gilsanz 2 , Heidi J Kalkwarf 3 , Joan M Lappe 4 , Sharon Oberfield 5 , John A Shepherd 6 , Babette S Zemel 7 & Karen Winer 8


1Wright State University, Dayton, Ohio, USA; 2Children’s Hospital, Los Angeles, California, USA; 3Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio, USA; 4Creighton University, Omaha, Nebraska, USA; 5Columbia University, New York, New York, USA; 6University of California, San Francisco, California, USA; 7Children’s Hospital, Philadelphia, Pennsylvania, USA; 8Eunice Kennedy Shriver National Institute of Child Health and Human Development, Bethesda, Maryland, USA.

Objectives: Creation of reference curves for areal bone mineral density (aBMD) and bone mineral content (BMC) with consideration of relevant anthropometric variables.

Methods: Analysis of the dual-energy X-ray absorptiometry (DXA) data collected as part of the Bone Mineral Density in Childhood Study1, including 2012 boys and girls, 5–22 years old, with a total of 10 525 visits, resulting in aBMD and BMC observations at the lumbar spine, hip (neck and total), forearm and whole body (total and sub-cranial). Multivariate statistics were used to rank order the influence of the independent variables age, gender, race (black/non-black), height, weight, percent body fat (%fat), and sexual maturity. Two different models were created for each aBMD and BMC parameter, the practical model containing age, gender, race, height, and weight as well as the full model adding %fat. We compared the number of cases that fell below 2 S.D.s in our models with those below the same limit of the currently standard LMS model2, which is based on age, gender and race, and of the height adjusted Z-scores3.

Results: For the six aBMD parameters, age, gender, weight and %fat were the most influential predictors, whereas height, race and maturity added little improvement to the models. In contrast, for the six BMC parameters, age, weight and %fat were the top predictors, but not gender. In comparing the overlap of subjects identified as below the normal limit of −2 S.D.s, between 56 and 84% of subjects identified as below normal based on the LMS model are not classified as being below normal in our practical model. Using the full model, the misclassification increases for all aBMD and BMC parameters, ranging from 61 to 92%. The height-adjusted Z-scores reduced the misclassifications to 41–59% in comparison to the practical model and to 55–73% in comparison to the full model.

Conclusion: The traditional comparison of paediatric BMD and BMC data against age-, gender-, and race-matched controls can be refined if anthropometric parameters are taken into account.

References: 1. Sponsored by the National Institute for Child Health and Human Development.

2. Zemel BS et al. J Clin Endocrinol Metab 2011 96 3160–3169.

3. Zemel BS et al. J Clin Endocrinol Metab 2011 95 1265–1273.

Volume 2

6th International Conference on Children's Bone Health

Rotterdam, The Netherlands
22 Jun 2013 - 25 Jun 2013


Browse other volumes

Article tools

My recent searches

No recent searches.