Longitudinal changes in anthropometrics and impact on self-reported physical function after traumatic brain injury.

BACKGROUND AND AIMS Patients admitted to the ICU with a traumatic brain injury (TBI) are at risk of muscle wasting but this has not been quantified. Our aims were to describe longitudinal changes in anthropometrical data, compare the accuracy of non-invasive methodologies to the validated dual-energy x-ray absorptiometry (DXA), and assess the relationships between anthropometrical data and self-reported physical function. METHODS In a prospective observational study, we recruited patients admitted to the ICU with a moderate-to-severe TBI over 12 months. Anthropometric measurements included the subjective global assessment (SGA), bodyweight and ultrasoundderived quadriceps muscle layer thickness (QMLT), which we performed weekly in hospital and 3 months after admission. We assessed total body composition using DXA within 7 days of ICU discharge, and compared the total lean muscle mass with ultrasound-derived QMLT taken within 5 days of the DXA measurement. We assessed functional outcomes at 3 months using the physical component score of the Short Form-36 (SF- 36) and the Extended Glasgow Outcome Scale (GOS-E). RESULTS Thirty-seven patients were included, with a mean age of 45 years (SD, 16 years), and 87% were men. Participants were admitted to the ICU for a mean of 13 days (IQR, 6-18 days) and to hospital for a mean of 38 days (IQR, 19-52 days). They had significant weight loss in hospital (mean, 4.9% [SD, 7.7%]; P = 0.001). Malnutrition, measured with the SGA, was twice as prevalent at hospital discharge than at admission (P = 0.005). A reduction in QMLT occurred in the ICU but stabilised after ICU discharge. DXA-derived total lean mass taken within 7 days of ICU discharge strongly correlated with ultrasound-derived QMLT taken within 5 days of DXA measurements (ρ = 0.74, P = 0.037). Improvements in self-reported physical function, using the SF- 36 and GOS-E at 3 months, were associated with a greater QMLT at hospital discharge (SF-36: ρ = 0.536, P = 0.010; GOS-E: ρ = 0.595, P = 0.003, n = 23) and at 3 months (SF-36: ρ = 0.658, P = 0.020; GOS-E: ρ = 0.642, P = 0.025, n = 12). CONCLUSIONS Patients with a TBI lose muscle thickness while in the ICU but the trajectory of loss stabilises after ICU discharge. Ultrasound-derived QMLT is related to total lean mass and physical function after discharge. Further studies are needed to confirm that ultrasound measurement of QMLT is a useful surrogate measure of muscle mass and functional outcomes after trauma and critical illness.