Related Experiment Videos
Inter-observer variability in the measurement of body composition
N J Fuller1, S A Jebb, G R Goldberg
1MRC Dunn Clinical Nutrition Centre, Cambridge.
European Journal of Clinical Nutrition
|January 1, 1991
Summary
Inter-observer variability in body composition assessment is significant, with skinfold thickness and near-infrared interactance showing greater variation than resistance methods or simple weight and height measurements.
Area of Science:
- Anthropometry
- Body Composition Analysis
- Clinical Measurement Science
Background:
- Accurate body composition assessment is crucial for clinical practice and research.
- Simple bedside assessments are convenient but their reliability requires evaluation.
- Inter-observer variability can significantly impact the validity of measurement data.
Purpose of the Study:
- To evaluate the inter-observer variation in simple bedside body composition assessment methods.
- To compare the reliability of different techniques including skinfold thickness, bioelectrical impedance analysis (resistance), and near-infrared interactance.
- To identify which methods exhibit the least systematic bias and variability.
Main Methods:
- Twelve healthy adults (6 male, 6 female) were assessed by six independent observers.
- Measurements included height, weight, skinfold thickness, whole-body and forearm resistance, and near-infrared interactance.
- Systematic bias and residual coefficients of variation (rCV) were calculated for each measurement and derived body composition estimates.
Main Results:
- Systematic bias was evident in most measurements, except height.
- Skinfold thickness showed the largest variability (rCV 11-18% for individual, 9% for sum).
- Weight and height measurements exhibited the lowest variability (rCV < 0.4%).
- Body fat estimation variability: skinfold (4.6%), near-infrared (4.2%), resistance (2.6%), weight/height alone (1.1%).
- Fat-free mass variability followed a similar trend to body fat, with lower rCVs.
Conclusions:
- Simple bedside body composition assessments, particularly skinfold thickness and near-infrared interactance, demonstrate considerable inter-observer variability.
- Methods relying solely on weight and height, or bioelectrical impedance analysis (resistance), offer greater reliability.
- Standardization of techniques and observer training are essential to minimize bias in body composition assessments.