Bioelectrical impedance-derived total body water for predicting exertional rhabdomyolysis-induced acute kidney injury during military training: a prospective cohort study
摘要
Exertional rhabdomyolysis-induced acute kidney injury (AKI) is a serious complication among military recruits undergoing intensive physical training in hot climates. Current screening parameters, including body mass index and urine specific gravity, have limited predictive accuracy. We aimed to evaluate whether baseline body composition, particularly total body water (TBW) measured by dual-frequency bioelectrical impedance analysis (BIA), can predict exertional rhabdomyolysis-induced AKI, and to characterize dynamic body composition changes during intensive training in military recruits.
MethodsA prospective cohort study was conducted among 301 healthy male conscripts from four military camps in Kanchanaburi, Thailand, during May–July 2017. Baseline body composition (TBW, muscle mass, and fat mass) was measured using a portable dual-frequency BIA device and reassessed biweekly throughout an 8-week intensive training program. The primary outcome was rhabdomyolysis-induced AKI, defined by clinical manifestations of rhabdomyolysis, creatine phosphokinase ≥ 5,000 U/L, and AKI according to KDIGO criteria. Logistic regression and receiver operating characteristic (ROC) analysis were performed to evaluate predictive performance.
ResultsFifty-eight participants (19.3%) developed rhabdomyolysis-induced AKI. The AKI group had significantly lower baseline TBW than the non-AKI group (45.08 ± 2.95% vs. 54.31 ± 5.21%, p < 0.001) and demonstrated greater losses in muscle mass and fat mass during training. On multivariate logistic regression, lower baseline TBW was the strongest independent predictor (adjusted OR 1.56 per 1% decrease, 95% CI 1.26–1.92, p < 0.001), together with lower plasma potassium (adjusted OR 10.88 per 1 mEq/L decrease, 95% CI 1.87–63.16, p = 0.008) and lower exercise frequency (adjusted OR 12.80 per one fewer session/week, 95% CI 4.73–34.65, p < 0.001). Baseline TBW demonstrated an area under the ROC curve of 0.938 (95% CI 0.909–0.964). At a cut-off of TBW < 50%, sensitivity was 98.3%, specificity 74.9%, positive predictive value 48.3%, and negative predictive value 99.5%.
ConclusionBaseline TBW measured by portable BIA is a strong predictor of exertional rhabdomyolysis-induced AKI in military recruits undergoing intensive training. With a negative predictive value of 99.5% at TBW < 50%, this measure is particularly suited as a field-deployable rule-out screening tool. Multicenter validation is warranted before implementation in large-scale screening programs.
Clinical trial registrationNot applicable.