Regional and total body bioelectrical impedance analysis compared with DXA in Icelandic elderly

Background/Objectives: The aims were (1) to compare fat free mass (FFM) estimates from regional hand-held bioelectrical impedance analysis (HHBIA) with conventional BIA (CBIA) and dual energy X-ray absorptiometry (DXA) and (2) to develop a population specific equation for FFM prediction in Icelandic elderly. Subjects/Methods: DXA, CBIA and HHBIA data were available for 98 free-living Icelandic elderly (age=73.0±5.6 years, body mass index=28.8±5.2 kg/m 2 ). Participants were randomized into a development block ( n =50) and validation block ( n =48). A population specific equation for FFM prediction was calculated using CBIA-derived resistance and anthropometric data from the development block and then compared with other BIA equations (Deurenberg, Segal, company-specific equations) and DXA estimates using the validation block. Results: The correlations between BIA methods and DXA were very high, that is, >0.9; however, mean differences compared with DXA were quite variable, ranging from −5.0 (Deurenberg) to +2.5 (Segal, HHBIA) and +3.3 kg (CBIA). Mean difference of the population-specific equation was below 0.1 kg. The standard deviations of the differences ranged from 2.6 to 3.3 kg. The limits of agreement of the BIA methods were similar and between 9.9 and 12.9 kg. Conclusions: In Icelandic elderly, HHBIA and CBIA produce similar FFM estimates when using company-specific prediction equations. CBIA provides the additional possibility to use a population-specific prediction equation, which yields best results. However, limits of agreement were wide and similar of all employed BIA methods, which indicates principal limitations of BIA analysis in the determination of FFM.