Fat-free mass at admission predicts 28-day mortality in intensive care unit patients: the international prospective observational study Phase Angle Project.

Abstract

Phase angle as measured by bioelectrical impedance analysis reflects fat-free mass. Fat-free mass loss relates to worse prognosis in chronic diseases. Primary aim of this study was: to determine the association between fat-free mass at intensive care unit admission and 28-day mortality. Ten centres in nine countries participated in this multicentre prospective observational study. The inclusion criteria were age >18 years; expected length of stay >48 h; absence of pacemaker, heart defibrillator implant, pregnancy and lactation. Fat-free mass was assessed by measurement of the 50-kHz phase angle at admission. The primary endpoint was 28-day mortality. The area under the receiver operating characteristic curve (AUC) was used to assess prediction of 28-day mortality by fat-free mass at ICU admission. The variables associated with 28-day mortality were analysed by means of multivariable logistic regression. Of the 3605 patients screened, 931 were analysed: age 61 ± 16 years, male 60 %, APACHE II 19 ± 9, body mass index 26 ± 6, day 1 phase angle 4.5° ± 1.9°. Day 1 phase angle was lower in patients who eventually died than in survivors (4.1° ± 2.0° vs. 4.6° ± 1.8°, P = 0.001). The day 1 phase angle AUC for 28-day mortality was 0.63 [0.58-0.67]. In multivariable analysis, the following were independently associated with 28-day mortality: age (adjusted odds ratio (aOR) 1.014 [95 % confidence interval 1.002-1.027], P = 0.03), day 1 phase angle (aOR 0.86 [0.78-0.96], P = 0.008), APACHE II (aOR 1.08 [1.06-1.11], P < 0.001), surgical patient (aOR 0.51 [0.33-0.79], P = 0.002), and admission for other diagnosis (aOR 0.39 [0.21-0.72], P = 0.003). A multivariable combined score improved the predictability of 28-day mortality: AUC = 0.79 [0.75-0.82]. Low fat-free mass at ICU admission is associated with 28-day mortality. A combined score improves mortality predictability. NCT01907347 ( http://www.clinicaltrials.gov ).