Accuracy in Real-World Conditions Isn’t So Accurate

At GHBCI, we emphasize the importance of following standardized procedures when performing body composition assessments. Even accurate methods will produce less reliable results when testing protocols aren’t followed. A recent study by Potter et al. (2025) published in the European Journal of Clinical Nutrition illustrates just how much error can be introduced when those procedures are not followed using a multi-frequency bioelectrical impedance analysis (MFBIA).

Researchers tested 1,000 active-duty U.S. Marines (667 men, 333 women; average age 28; BMI 27) with no instructions regarding fasting, hydration, caffeine, or recent exercise. Participants were told to report “whenever available.” This meant their hydration status and other physiological factors – which heavily influence BIA – were uncontrolled. MFBIA results were compared with DXA measurements taken during the same session.

Findings: MFBIA underestimated body fat by 4% in men and nearly 3% in women. Despite high correlations between MFBIA and DXA, the error rates were large—5.2% for men and 3.8% for women, both exceeding the acceptable 3.5% error rate threshold. Practically, a woman measured at 25% fat could truly be between 21.2% and 28.8%, and a man at 25% could be between 19.8% and 30.2%.

For comparison, McLester et al. (2020) evaluated the same model of MFBIA under proper standardized protocols. Although MFBIA still slightly underestimated DXA (3.5% in men, 3.3% in women), the actual errors were noticeably smaller—3.0% in men and 2.6% in women. These both fall below the recommended 3.5% threshold, meaning the results were much more dependable.

Note: The Marine sample may actually show less variability than the general population due to more consistent routines and hydration patterns of active duty military personnel. In everyday settings, with the general public, errors could be even larger.

Bottom line: When using MFBIA—or any body composition method—following standardized procedures is essential. Proper preparation dramatically reduces error and improves the reliability of results.

References:

Potter AW et al. (2025) Real-world assessment of Multi-Frequency Bioelectrical Impedance Analysis (MFBIA) for Measuring Body Composition in Healthy Physically Active Populations. European Journal of Clinical Nutrition; https://doi.org/10.1038/s41430-025-01664-4

McLester CN et al. (2020) Reliability and Agreement of Various InBody Body Composition Analyzers as Compared to Dual-Energy X-Ray Absorptiometry in Healthy Men and Women. Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health, 23: 3, 443-450.