Discriminant models to estimate the body weight loss after a six-month long diet and exercise-based intervention

BACKGROUND: The main concern of the people who follow a weight loss program is the body weight loss, independently of the body composition. The aim of this study was to create a mathematical model able to discriminate the body weight change based on initial body composition variables. METHODS: The study included 239 overweight and obese participants (18-50 years; Body Mass Index (BMI)>25 and <34.9 kg/m(2)). Participants performed an endrance, strength or a combined endurance and strength training three times per week, or followed the American College of Sports Medicine physical activity guidelines for body weight loss, during twenty-four weeks while having 25-30% caloric restriction. Two multivariate discriminant models were performed taking into account the groups below and above the mean body weight change. RESULTS: The discriminant models obtained could discriminate the body weight change with a 65-70% of correct classification. BW, fat free mass (FFM), and fat mass (FM) were shown to be the most discriminant variables for the discriminant models. CONCLUSIONS: People having higher FM and FFM at the beginning of an intervention will lose a greater amount of weight until the end of it.