Assessment of body composition: Intrinsic methodological limitations and statistical pitfalls

Area of Science:

• Nutrition science
• Biomedical engineering
• Clinical research methodology

Background:

• Body composition (BC) evaluation is essential for assessing nutritional status, especially in various pathological conditions.
• Reliable and feasible BC measurement methods are needed for clinical practice, research, and epidemiological studies.
• Existing BC techniques yield inconsistent results due to differing principles, intrinsic biases, and varied population/clinical settings.

Purpose of the Study:

• To review common body composition measurement techniques used in clinical practice.
• To provide guidance on the appropriate statistical interpretation of BC data.
• To identify and discuss the limitations and potential biases associated with BC assessment methods.

Main Methods:

• Review of scientific literature on body composition measurement techniques.
• Analysis of statistical methods commonly employed for comparing BC assessment tools (e.g., Pearson's correlation, ICC, Lin's CCC, Bland-Altman analysis).
• Identification of factors influencing BC measurement variability.

Main Results:

• Various techniques exist for body composition measurement, but results can be non-equivalent.
• Statistical methods like Pearson's correlation, ICC, Lin's CCC, and Bland-Altman are used for comparison, but interpretation can be biased.
• Factors such as ethnicity, age, and disease type can affect BC measurement outcomes.

Conclusions:

• Understanding the principles and limitations of different BC measurement techniques is crucial.
• Careful application and interpretation of statistical methods are necessary to avoid bias in BC data analysis.
• Standardized approaches and awareness of potential pitfalls are needed for reliable BC assessment across diverse settings.