Body size and pubertal development explain ethnic differences in structural geometry at the femur in Asian, Hispanic, and white early adolescent girls living in the U.S.

Variation in structural geometry is present in adulthood, but when this variation arises and what influences this variation prior to adulthood remains poorly understood. Ethnicity is commonly the focus of research of skeletal integrity and appears to explain some of the variation in quantification of bone tissue. However, why ethnicity explains variation in skeletal integrity is unclear. Methods: Here we examine predictors of bone cross sectional area (CSA) and section modulus (Z), measured using dual-energy X-ray absorptiometry (DXA) and the Advanced Hip Analysis (AHA) program at the narrow neck of the femur in adolescent (9-14 years) girls (n = 479) living in the United States who were classified as Asian, Hispanic, or white if the subject was 75% of a given group based on parental reported ethnicity. Protocols for measuring height and weight follow standardized procedures. Total body lean mass (LM) and total body fat mass (FM) were quantified in kilograms using DXA. Total dietary and total dairy calcium intakes from the previous month were estimated by the use of an electronic semi-quantitative food frequency questionnaire (eFFQ). Physical activity was estimated for the previous year by a validated self-administered modifiable activity questionnaire for adolescents with energy expenditure calculated from the metabolic equivalent (MET) values from the Compendium of Physical Activities. Multiple regression models were developed to predict CSA and Z. Results: Age, time from menarche, total body lean mass (LM), total body fat mass (FM), height, total calcium, and total dairy calcium all shared a significant (p<0.05), positive relationship with CSA. Age, time from menarche, LM, FM, and height shared significant (p<0.05), positive relationships with Z. For both CSA and Z, LM was the most important covariate. Physical activity was not a significant predictor of geometry at the femoral neck (p >= 0.339), even after removing LM as a covariate. After adjusting for covariates, ethnicity was not a significant predictor in regression models for CSA and Z. Conclusion: Variability in bone geometry at the narrow neck of the femur is best explained by body size and pubertal maturation. After controlling for these covariates there were no differences in bone geometry between ethnic groups. Published by Elsevier Inc.