Search

Aims: To characterise the distribution of peak bone mass at the total hip, lumbar spine (LS) and total body (TB) sites of women from one European Caucasian (EC) and two South Asian populations, namely Pakistani Muslims (PM) and Gujarati Hindus (GH), residing in the UK, and to explore whether variations of lifestyle and other non-genetic putative risk factor characteristics may explain any variation in peak bone mass between the EC and South Asian groups. Methods: Women, aged 18 years to 36 years, were recruited from primary care age-sex registers from practices located within Greater Manchester. Women were mailed a screening questionnaire concerning various lifestyle factors. Those women who replied, were invited to attend for a more detailed interview-assisted lifestyle questionnaire, assessments of height, weight and waist-to-hip ratio, bone density (BMD) at the total hip, LS and TB, fat mass (FM) and lean mass (LM) and blood tests for serum calcium, parathyroid hormone (PTH) and 25 hydroxyvitamin D (25 OHD). Results: The participation rates for BMD screening, were 19%, 9% and 14% for EC, PM and GH women respectively. EC and GH women who completed the screening questionnaire were slightly older than those who did not. Based on a study comparing responses to the screening questionnaire, differences in characteristics between those who attended for BMD screening and those who did not, for all three ethnic groups , were generally small and the direction inconsistent in terms of benefit or detriment to BMD. BMD data was available on 118 EC, 98 PM and 28 GH women. Compared to PM, EC had significantly higher BMD at the hip (0.962 g/cm2 vs 0.916 g/cm2 ), LS (1.031 g/cm2 vs 0.984 g/cm2) and TB (1.096 g/cm2 vs 1.058 g/cm2). Although BMD at all three sites were higher among EC compared with GH, differences were small and non-significant. There were no significant differences in LS BMAD (calculated by dividing areal BMD by subject height), between EC and either South Asian group. When comparing EC and PM women, adjusting for age and separately for height, LM, maximum non-pregnant weight, serum 25 OHD, PTH, dietary calcium consumption or vitamin D consumption, eliminated significant BMD differences at all three sites, adjusting for weight eliminated differences at the hip and LS, while adjusting for hormonal contraception use eliminated differences at the hip only. In multivariate models, at the hip, after adjusting for age, weight, height, 25 OHD, PTH, dietary calcium consumption and hormonal contraception use, at the LS, after adjusting for these characteristics (other than hormonal contraception use) and at the TB, after adjusting for age, height, LM, 25 OHD, PTH and dietary calcium consumption, BMD was significantly higher in PM women (b coefficient for PM with EC as referent = 0.063, 0.064 at the hip, LS and TB respectively).