Error reduction as a calibration strategy for body composition measurements: A comparison between bioelectrical impedance analysis and dual-energy X-ray absorption.

BIA demonstrates systematic measurement errors compared to DEXA across body composition variables.

• 61 participants (33 men, 28 women; aged 24-63 years) underwent assessments using both BIA and DEXA
• Both height, weight, BMI, and body composition variables were assessed and compared between methods
• A mirrored dataset was created using both methods to enable direct comparison
• BIA values differed systematically from DEXA values across multiple anthropometric and body composition variables

After calibration, most body composition variables showed the highest predictive accuracy, while trunk and limb body fat mass (BFM) variables performed the worst.

• A linear-percentile calibration and OLS approach was used to align empirical quantiles, adjusting BIA measurements to DEXA standards
• Trunk FFM (fat-free mass) showed improvement only by the OLS method, not by percentile calibration
• Trunk BFM and limb BFM variables were identified as the poorest performing variables after calibration
• The regression models showed significant improvement in agreement between methods within the sample

Calibration reduced errors in most BMI and sex subgroups, except for trunk FFM and leg FFM, for which calibration showed no benefit.

• Sex-specific and BMI-specific analyses were conducted to identify subgroup differences in calibration effectiveness
• Trunk FFM and leg FFM did not benefit from the calibration approach in these subgroups
• Calibration was effective for most other variables across BMI and sex groups
• The study included both male and female participants across a BMI range broad enough to permit BMI-specific subgroup analyses

A percentile-based calibration combined with ordinary least squares (OLS) linear regression was proposed as a method to reduce measurement error between BIA and DEXA.

• The calibration method aligned empirical quantiles of BIA measurements to DEXA standards
• OLS linear regression was applied in addition to percentile alignment to further improve agreement
• The approach was described as 'a practical strategy to enhance the accuracy of BIA-based assessments in clinical and research contexts'
• The method was developed and validated within a sample of 61 participants aged 24-63 years

DEXA is identified as the clinical gold standard for body composition assessment, while BIA scales are widely used due to their lower cost despite producing differing values.

• DEXA was described as costly, leading to widespread use of BIA scales as an alternative
• BIA scales 'often yield values that differ from those obtained with DEXA'
• The study aimed to quantify the error between anthropometric variables measured by BIA and DEXA and to identify variables with the highest and lowest errors
• Accurate diagnosis of obesity was cited as the primary clinical motivation for valid body composition assessment