Diagnostic performance of weight loss to predict body fatness improvement in cardiac rehabilitation patients

PURPOSE: To determine the diagnostic performance of weight loss to identify fat mass loss in cardiac rehabilitation (CR) patients. METHODS: We included consecutive patients enrolled in early outpatient CR who underwent air displacement plethysmography with measurements of height, weight, and waist circumference at initiation and completion of CR. We calculated the accuracy of >1 kg of weight loss to predict a >1 kg loss of fat mass. RESULTS: We analyzed data from 142 patients (mean age ± SD = 60 ± 12 years), 74% male, 94% non-Hispanic whites, and body mass index (BMI) 29.9 ± 5.1 kg/m. Following 87 ± 49 days and 22 ± 9 CR sessions, there was a small but significant change in weight (-1.3 ± 3.8 kg), BMI (-0.4 ± 1.2 kg/m), fat mass (-2.6 ± 3.9 kg), lean mass (+1.3 ± 1.9 kg), and waist circumference (-4.3 ± 5.1 cm), P < .001 for all. Overall, patients who lost weight consistently lost fat mass, positive predictive value 0.91 (95% CI: 0.83-0.96). However, the negative predictive value of lack of weight loss to exclude fat mass loss was poor, 0.59 (95% CI: 0.52-0.64). Among patients who did not lose weight, waist circumference reduction was modestly predictive of fat mass loss (r = 0.33, P = .004.) CONCLUSIONS: Although weight loss in CR is indicative of fat mass loss in most patients, absence of weight loss, or even weight gain, would not necessarily rule out fat loss in a significant number of patients attending CR. These findings speak to the importance of body fatness measurements beyond BMI in the CR setting.