Prediction equation for body fat in spinal cord injured patients

Abstract

Body mass index (BMI) is the one method to determine cardiovascular risk which is the common cause of death in spinal cord injury. Previous study reported that patient with spinal cord injury had lower total lean mass and higher total fat mass than the normal person with the same body mass index. Therefore, this study aimed to develop and validate equation for predicting percentage body fat in spinal cord injured patient. Cross-sectional descriptive study was conducted among sixty paraplegic patients for development group and 31 paraplegic patients for validation group. Demographic data including body weight, height and body mass index (BMI) were retrieved. All patients were evaluated percentage of body fat (%BF) using dual energy X-ray absorptiometry (DXA). Regression equation for %BF was created from the data in development group. After that, the equation was evaluated for accuracy in the validation group. Only BMI showed significantly correlated with %BF ( P < 0.001). The suggested equation by stepwise multiple regression analysis was: %BF = (1.477 × BMI) −0.091 (R 2 = 0.396, P < 0.001). When the equation was validated in the validated data group, the correlation coefficient of the predicted and DXA-measured %BF was r = 0.704 ( P = 0.003). Bland and Altman analysis showed mean differences −1.49% with the limits of agreement ranged from −18.6% to 15.6%. This study suggested the simple equation which might be helpful for evaluation of percentage of body fat in spinal cord injured patients.