Validity of multi-frequency bioelectric impedance methods to measure body composition in obese patients: a systematic review

Objective Excessive lean tissue loss following bariatric surgery may pose serious metabolic consequences. Accurate methods to assess body composition following bariatric surgery are required. This review aimed to investigate if multi-frequency bioelectric impedance (MF-BI) is a valid tool to determine body composition in obese patients. Methods MEDLINE, EMBASE, CINAHL and CENTRAL databases were searched until March 2017. Included studies were published in English with obese (body mass index (BMI) >= 30 kg/m(2)) adults measuring body composition with MF-BI methods in comparison with reference methods. Exclusions were pregnancy, animal studies, non-English language studies, single frequency BI. A total of 6395 studies were retrieved. Results Sixteen studies were eligible for inclusion. Sample sizes ranged from 15 to 157, with BMI 26-48 kg/m(2). MF-BI underestimated fat mass (FM) in 11 studies and overestimated fat-free mass (FFM) in nine studies in comparison with reference methods. Correlations of absolute values from MF-BI and reference methods for FM and FFM were high, however, agreement was lower at an individual level. When adjustments for BMI were made to machine algorithms, measurement accuracy improved. Significant heterogeneity was evident among included studies. Conclusions This review found that MF-BI is reliable for use at a group level. Obese-specific adjustment of algorithms for MF-BI machines increases the accuracy of absolute measures of body composition in obese individuals, improving their utility in the clinical setting. Multiple variables contributed a lack of consistency among studies included, highlighting the need for more robust studies that control confounding variables to establish clear validity assessment.