An assessment of the impact of South Asian ethnicity on body composition, dimensions and proportions in children and adolescents, and their potential for risk of metabolic syndrome

Shah, Mahjabeen (2015) An assessment of the impact of South Asian ethnicity on body composition, dimensions and proportions in children and adolescents, and their potential for risk of metabolic syndrome. Doctoral thesis, London Metropolitan University.


South Asians (SA) are a high-risk group for cardiometabolic disease, which is partly attributed, to their ‘thin-fat’ body composition (BC) phenotype. Generally, SAs have a higher % fat mass (FM) and less skeletal muscle mass (SMM), together with a more abdominal distribution of body fat, compared with white Europeans (WE) at equivalent body mass index (BMI) values. SAs also tend to have a shorter adult stature. Effective paediatric monitoring and clinical management requires improved tools for assessing body fatness and other components of BC. This is partly due to BMI being regarded as an inadequate indicator of adiposity and SMM, particularly for SAs, thus other field-based measures of BC have been investigated. In addition to BMI, several UK BC references for WE children and adolescents are available, including WC, %FM, and SMM; however, there are no similar references available for their SA counterparts.

This thesis is comprised of four key studies in which ethnic variations in BC (%FM and appendicular SMM (SMMa)), WC, and leg length (LL) in particular relative LL (RLL) between SA and WE children and adolescents (aged 5-18y) were investigated. The core aim was to develop age- and sex-specific percentile references for %FM, SMMa, and WC for the SA ethnicity. In study one, the BC418 bioelectrical impedance analyser (BIA) was validated against DXA for field-based BC assessment, in a sample of SA children (n= 53; 5-21y) to develop an ethnic-specific prediction equation for FM, FFM and SMMa determination, as prior studies have found BIA underestimates %FM in SAs. This equation was found to be only valid for children ≥9y, which was attributed to the opportunistic nature of recruitment. It was concluded that no single equation was valid across the whole child and adolescent population. This equation needs to be tested in an independent group to confirm its accuracy and functionality, prior to wider application. Due to discrepancies in DXA weight and scale weight, it was not possible to develop a prediction equation for SMMa, although the existing BIA output was considered acceptable due to the very small between-method relative differences.

In study two the new BIA prediction equation was applied to a large SA dataset of children (n =1,624) from low-income communities. Compared to UK90 (SDS) reference data, both SA girls and boys (9-14y) were significantly shorter, lighter, with a lower mean BMI and WC compared with their WE counterparts, with no significant differences in %FM. %FM and SMMa reference curves were constructed and comparisons were made between published WE (from affluent areas, WE1) reference curves, together with a low-income cohort (WE2). Comparisons in %FM at the 50th centile, between the SA and WE cohorts revealed that SAs had greater %FM overall, and this difference increased after application of the new equation. Similarly, across all age ranges SAs had significantly less relative SMMa than their WE counterparts.

The third study generated SA sex- and ethnic-specific WC centile curves. SDS comparisons with the WE2 cohort revealed SAs had a significantly lower mean WC than their WE2 counterparts. It was concluded that, as WC acts as a proxy for visceral fat, ethnic-specific cut-offs similar to those adopted for adults in India should be considered for children. The final study on LL revealed that SAs had a longer RLL than WE children, although as RLL data for WEs was derived, this would require further verification. The outcomes from these studies provide the evidence base and assessment tools to support the use of ethnic-specific references for children and youths in the UK from a SA background. The findings in this thesis demonstrate that overweight and obesity vary across different ethnic groups and this variation needs to be considered in the context of the clinical referral for individual children as well as for population surveillance. These are the first set of reference percentile charts for BC, proportions and dimensions in the UK SA paediatric population. Our findings support the use of these ethnic-specific references that go beyond BMI as an indicator of obesity-related metabolic health risk.

ShahMahjabeen - PhD Final Thesis.pdf - Published Version

Download (5MB) | Preview


Downloads per month over past year

Downloads each year

View Item View Item