Prognostic Value of Multifrequency Bioelectrical Impedance Analysis in Chronic Obstructive Pulmonary Disease: Systematic Review.

Background and Objectives: Chronic obstructive pulmonary disease (COPD) is a systemic condition in which muscle wasting, malnutrition, and altered fluid balance strongly influence prognosis. While spirometry remains essential for diagnosis and staging, it often fails to reflect the heterogeneity of outcomes. Multifrequency bioelectrical impedance analysis (MF-BIA) enables the assessment of body composition and fluid distribution, offering additional prognostic information. This systematic review aimed to evaluate the prognostic significance of MF-BIA in COPD, with emphasis on outcomes such as mortality, exacerbations, and hospital admissions. Materials and Methods: We systematically searched PubMed, Web of Science and Scopus from inception to 29 April 2025. The earliest record retrieved was published in 1996 but was excluded during screening. Studies including COPD patients in whom MF-BIA-derived parameters were related to clinical outcomes were eligible. Risk of bias was assessed using the Newcastle-Ottawa Scale. Data on design, population, methodology, and endpoints were extracted and narratively synthesized due to heterogeneity. The review protocol was not registered. Results: Eight studies were included. Phase angle (PhA) consistently showed prognostic value, being inversely related to mortality and rehospitalizations. Fat-free mass index (FFMI) was integrated into multidimensional models, but its independent role was inconsistent. Parameters describing fluid distribution, such as Extracellular Water/Total Body Water ratio, also appeared relevant, though interpretation was often limited by the absence of consistent consideration of underlying cardiac disease. Conclusions: MF-BIA provides useful prognostic insights in COPD patients, particularly through PhA. It may refine risk stratification beyond spirometry, yet further prospective studies with standardized methods are needed to confirm its independent value. Heterogeneity of methods and small sample sizes remain important limitations.