Abstract
Background. Pharmacotherapy, associated with a comprehensive weight loss intervention, has emerged as a promising therapeutic approach in adolescents. Identification of subjects who best respond to a pharmacological intervention remains difficult. Objective. To compare the value of early weight loss after 12 weeks of treatment with placebo or orlistat (120 mg three times a day) in predicting treatment outcome after 52 weeks. Methods. Secondary analysis of a randomized control trial in 182 placebo-treated and 357 orlistat-treated obese adolescents (Body mass index [BMI] ≥2 kg/m2 above the 95th percentile). Results. Percent weight change at 12 weeks was positively correlated with percent change in weight (r2 ≥ 0.41), BMI (r2 ≥ 0.33) and waist circumference (r2 ≥ 0.20) at 52 weeks in both the placebo and orlistat groups (P < 0.001). A weight loss ≥5% of baseline weight at 12 weeks was associated with a mean weight loss of 8.1% (95% CI: 6.4 to 9.7) at the study end that was independent of treatment. Subjects in the orlistat group were 2.44 times (95% CI: 1.34 to 4.46) more likely to experience a weight loss ≥5% after 12 weeks than subjects in the placebo group (P = 0.0028). Conclusions. Early weight loss predicts a favourable outcome in both placebo-treated and orlistat-treated subjects but is more than 2 times more likely to occur in the orlistat group. Addition of orlistat should be considered as part of a weight loss intervention but reevaluated after 3 months of treatment.
Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. M.R.'s time (Statistician) was reimbursed by the sponsor of the study (Roche). The agreement was in place prior to data analysis. The analysis of the data was performed independently from the sponsor using the raw data from the study and Roche was not involved in the writing of the manuscript or in the decision to publish. The results and opinions expressed in the present submission exclusively reflect those of the authors (J.P.C. and M.R.).