Use this URL to cite or link to this record in EThOS:
Title: Body composition analysis : evaluation of methods in adolescents of varying fatness
Author: Radley, Duncan
ISNI:       0000 0001 2419 2527
Awarding Body: Leeds Metropolitan University
Current Institution: Leeds Beckett University
Date of Award: 2007
Availability of Full Text:
Access from EThOS:
Introduction The objectives of the present study were twofold: 1) to investigate the cross-sectional and longitudinal accuracy of percentage body fat (%fat) estimates from laboratory methods and field methods in a sample of mainly overweight and obese adolescents; and 2) to examine the use of each method to assess the effectiveness of the Carnegie International Weight Loss Camp (CIC) intervention. ABSTRACT / / Methods The laboratory methods: dual-energy X-ray absorptiometry (OXA), air displacement plethysmography (AOPSiri. AOPLoh), total body water (TBW73, TBWLoh) , threecompartment mineral model (3CMin) and three-compartment total body water model (3CrBw), and the field methods: skinfold thicknesses (SKFs), bie-electrical impedance analysis (BIA) and four-compartment bie-electrical impedance model (4CB1A) were evaluated against a criterion four-compartment model (4CLoh). 76 adolescents participated in the cross-sectional study, age (mean ± SO) 14.0 ± 1.6 y, body mass index 30.0 ±6.7 kg.m-2 and %fat (4C model) 36.9 ± 11.9%. 13 children attending the CIC (campers), 14 overweight comparison group children (OWCG) and 8 normal weight comparison group children (NWCG) participated in the longitudinal and intervention evaluation study. Results Cross-sectional analysis of laboratory methods, in all subjects, revealed mean percentage body fat (%fat) determined by 3CrBW (37.0 ± 11.8%) and 3CMin (36.8 ± 12.9%) were within ± 0.5% of that determined by 4CLoh (36.9 ± 11.9%). %Fat determined by AOPsiri (38.0 ± 12.4%), AOPLoh (36.0 ± 12.8%), TBW73 (35.8 ± 11.7) and TBWLoh (37.6 ± 11.4%) were within between ± 0.7% and ± 1.2% of that determined by 4CLoh, whilst OXA overestimated %fat by 3.6%. Considering individual agreement 3CrBW revealed the lowest 95% limits of agreement (± 1.1%) followed by values between ± 3.6% and ±7.2% for all other methods. Cross-sectional analysis of field methods, in all subjects, revealed mean %fat determined by 4CBIA (36.2 ± 12.8%) and 1 BIA prediction equation were within ± 0.5% of that determined by 4CLoh %fat (35.9 ± 12.5%). %Fat determined by 3 BIA prediction equations were within between ± 0.8% and ± 1.8% of that determined by 4CLoh , whilst all other methods differed by between ± 2.4% and ± 6.0%. Considering individual agreement 4CB1A revealed the lowest 95% limits of agreement (± 3.2%) and SKFs the greatest (± 15.6%). The 95% limits of agreement for all other BIA prediction equations ranged from ± 7.9% to ± 12.8%. Longitudinal analysis of laboratory methods, by group, revealed a less than ± 1% mean difference to 4CLoh %fat delta values in all cases except by 3CMin, TBW73 and TSWLoh in Campers, 3CMin in the OWCG and DXA in the NWCG. Considering individual agreement the lowest 95% limits of agreement were produced by 3CrBW.(± 0.6% to ±1.0%) and the highest by DXA (%fat ±4.2% to ±5.9). Longitudinal analysis of field methods, by group, revealed a less than ± 1% mean difference to 4CLoh %fat delta values in all cases except 2 BIA prediction equations and SKFs in Campers, but a greater than ± 1% mean difference in all cases in the OWCG and NWCG with the exception of SKFs in the NWCG. In all cases the lowest 95% limits of agreement were produced by 4CB1A (%fat ± 1.8% to ± 6.2%). Evaluation of the CIC intervention revealed that 3CrBW was the only method to show agreement with the 4C model findings of a significant difference between each group. Discussionl Conclusion The present study demonstrates that a number of methods were able to provide accurate mean but not individual cross-sectional and longitudinal body composition estimates compared to a criterion 4C 'model. In general laboratory methods provided more accurate estimates than field methods. In addition, the study highlights 3CTBW as an accurate alternative to 4C' model analysis in lean, overweight and obese adolescents when very accurate individual estimates are required. Similarly, when evaluating changes in an intervention compared to control groups using only a small number of participants the 3CrBW model is the most accurate altemative method. Findings from the present study also highlight caution is warranted when interpreting data, both cross-sectionally and longitudinally, without careful attention to and understanding of t~e theoretical underpinnings of the method being used.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available