BPK 142 Lecture Notes - Lecture 4: Bioelectrical Impedance Analysis, Subcutaneous Tissue, Iliac Crest
27 Apr 2018
School
Department
Course
Professor
Unit 4: Prediction of Percentage Body Fat
Introduction
- 1940s-early 1990s: underwater weighing used as criterion standard for body fat estimation
- Not practical field technique
- Weight for height tables have been used to measure obesity. However, not all people heavy for their height
are obese (ex. Body builders)
- Method of choice: prediction of body fat from anthropometric measures.
- Based on regression eqtns for the prediction of % of body fat or body density
- Ex. Siri eqtn, Brozek eqtn
- Eqtns produced by determining indirect body densities in a group of subjects using underwater weighing or
volume measurements and then carrying out a regression analysis on the body density (anthropometric
variables)
- Result: prediction eqtn containing anthropometric variables that predict body density (or its derivative %
body fat) in the given sample
- Predictive ability of eqtns indicated by correlation coefficient (R) and a standard error of estimate (SEE) of
the predictive measure.
- Estimation of total body fat and fat distribution non-invasively (CT scan, MRI scan, DXA)
- DXA = dual-energy x-ray absorptiometry
- Criterion measurement/reference standard for body composition assessment in clinics + research facilities
- 2 component model of body composition analysis: fat and fat-free mass (FFM). assumes each has fixed
density
- Does not take into account age, sex, race/ethnic differences
- Multi-component models incl. Measures of constituents of FFM [fat + fat-free mass] (usually protein, mineral,
water). More accurate.
- Older % fat prediction eqtns that used underwater weighing as criterion standard based on 2 component
model of body composition.
- Skinfold analysis = based on principle that amount of subcutaneous fat (fat located immediately under skin)
is directly proportional to total amount of body fat
- Proportion varies w/ age, sex, ethnicity, etc.
- Multiple regression eqtns to predict body density + fat from skinfold measurements
- In a sample upon which eqtn was developed, the following were contant
- Skinfold patterning, skinfold compressibility, tissue densities (particularly bone), ration of external/internal
adipose tissue, fat (lipid) content of adipose tissue
- Eqtns used in today’s lab: Yuhasz (Carter 1982), Durnin and Womersley (2011), Jackson and Pollack (1985,
2010), Peterson (2003), Wang (1994)
- SEE of predictive eqtns: when eqtn is used to predict body fat, prediction is ⅔ times within +/- 3.7% body fat
of correct value.
- Eqtns different for sexes, may be different for ethnicities
- Yuhasz (Carter, 1982):
- Eqtns derived on Ontario university males + females in early 1970s, a Caucasian population
- Sum of 6 skinfold measurements
- Don’t need calc of body density
- Durnin and Womersley (2011):
- Originally developed in early 1970s on a Caucasian sample of British subject 16-70 years old
- 2011, developed % fat eqtns for diff ethnic groups based on data collected from 1675 subjects in NY from
1985-2006 (Caucasian, Asian, African-American, Hispanic)
- Eqtns developed using 3 component model DXA procedure as reference standard.
- Jackson and Pollack (1978, 1980):
- Regarded as most appropriate eqtns for individuals of various ages
Document Summary
1940s-early 1990s: underwater weighing used as criterion standard for body fat estimation. Weight for height tables have been used to measure obesity. However, not all people heavy for their height are obese (ex. Method of choice: prediction of body fat from anthropometric measures. Based on regression eqtns for the prediction of % of body fat or body density. Eqtns produced by determining indirect body densities in a group of subjects using underwater weighing or volume measurements and then carrying out a regression analysis on the body density (anthropometric variables) Result: prediction eqtn containing anthropometric variables that predict body density (or its derivative % body fat) in the given sample. Predictive ability of eqtns indicated by correlation coefficient (r) and a standard error of estimate (see) of the predictive measure. Estimation of total body fat and fat distribution non-invasively (ct scan, mri scan, dxa) Criterion measurement/reference standard for body composition assessment in clinics + research facilities.
