False and true pre-treatment predictors of weight loss in obese patients starting a program for lifestyle change

Weight loss treatment effectiveness and cost-effectiveness may be improved by the identification of patients who are more prone to participate and gain benefit from specific interventions. Aim of the present study is to identify easily available additional predictors of weight loss among data usually present in the medical records of obese/overweight patients attending an outpatient clinic for a non-pharmacological lifestyle change program.

268 patients, 74 men and 195 women (age 43.2 ± 11.9 years, BMI 38.9 ± 6.8 kg/m(2)) were enrolled. Among these patients, only 35.6 % men and 22.7 % women completed the 6-month protocol. Among participants, 50.7 % lost at least 5 % initial body weight after 6 months (SUCCESSES), while 49.3 % failed (FAILURES). Baseline nutritional parameters (total Kcal, lipid, carbohydrate, protein and alcohol intake) were not significantly different in successes when compared to failures, while a significant difference between groups was observed for baseline diastolic blood pressure (DBP); free fat mass (FFM); muscle mass (MM); total body water (TBW); HDL cholesterol; ALT; AST; γGT. After dividing into quartiles the not-normally distributed variables, successes had AST values above median (3rd and 4th quartiles; χ (2) = 0.003). At multivariate analysis (linear regression), the OR was 3.34 (1.42-7.85; p = 0.006).

In our patients, baseline liver enzyme levels (AST in particular), but not baseline quantitative and qualitative dietary intake, were significantly different in successes versus failures and could therefore represent a predictor of success. In conclusion, AST could represent a usually available biomarker that could be used as a predictor of outcome (weight loss) in obese patients starting a lifestyle change program.