Mortality after hip fracture in Japan: the role of nutritional status.

PURPOSE To assess factors that influence 4-year mortality following hip fracture surgery in a Japanese population. METHODS Records of 129 hips in 24 men and 103 women aged 50 to 103 (mean, 79) years who underwent surgery for femoral neck or trochanteric fractures were reviewed. Clinical data reviewed included age, gender, body mass index (BMI), side of fracture, fracture type, fracture stability, surgery type, interval from admission to surgery, length of hospital stay, number of pre-fracture comorbidities, pre-fracture ambulatory level, pre-fracture place of residence, preoperative dementia, preoperative skeletal traction, blood haemoglobin level, serum albumin level, number of postoperative complications, and postoperative delirium. Univariate and multiple logistic regression analyses were performed to identify the relative contribution of the variables to mortality. Receiver operating characteristic (ROC) curves were used to identify optimal cut-off levels. RESULTS The 4-year mortality was 48%. Multiple logistic regression analysis showed that serum albumin level (p = 0.0004, odds ratio [OR] = 5.8541) and BMI (p = 0.0192, OR = 1.1693) significantly influenced mortality; the cut-off points were 36 g/l and 18.9 kg/m square, respectively, based on the ROC curves. Kaplan-Meier curves showed that survival rates were significantly worse in patients with values below these cut-off points. CONCLUSION Serum albumin level and BMI on admission are predictive of mortality after hip fracture surgery.