Determinants of bone mass in 10- to 26-year-old females: A twin study

Abstract

This cross‐sectional twin study aimed to quantify the roles of constitutional and lifestyle factors on bone mass in adolescent and young adult women. Areal bone density (BMD) at the lumbar spine, femoral neck, Ward's triangle, and total hip, total body bone mineral content (BMC), and lean mass and fat mass were measured using dual energy X‐ray absorptiometry (DXA) in 215 female volunteer twin pairs (122 monozygotic [MZ], 93 dizygotic [DZ]) aged 10 to 26 years. Height, weight, menarchial history, dietary calcium intake, physical activity, current tobacco use, and alcohol consumption were determined by questionnaire. Mean BMD increased with age to around 16 years, when it reached a plateau. Within‐pair differences in BMD at the lumbar spine (expressed as a percentage of the pair mean BMD) were univariately associated with pair differences in menarchial status (14 ± 3%), height (0.7 ± 0.1% per cm), weight (0.4 ± 0.1% per kg), lean mass (1.0 ± 0.1% per kg), and fat mass (0.5 ± 0.1% per kg). Only menarchial status, height, and lean mass, however, were independent predictors. At the proximal femoral sites, within‐pair BMD differences were associated with within‐pair lean mass differences (1.0 to 1.1 ± 0.2%/kg), and no other factor was significant. The same conclusions applied to within‐pair differences in BMD/height. Total body BMC was independently associated with menarchial status, height, lean mass, and fat mass; the effects of the latter two variables were stronger in pairs both premenarchial. After adjusting for constitutional factors, no lifestyle factor was independently predictive. By reducing collinearity, the cotwin method clearly identified that lean mass, not fat mass, was the major independent determinant of bone mass at the hip, both pre‐ and postmenarche.