Assessment of weight maintenance at five years after behavioral weight loss intervention

Maintenance of satisfactory weight even after substantial weight loss is challenging for many individuals. We evaluated subjects in placebo arm of a randomized, double-blind, controlled trial over 3- 60 mos after initial weight loss in non-pharmacologic, weight loss programs (NPWLP). At 26 sites, eligible subjects who had lost >10 kg in NPWLP and maintained loss of >10 kg for 6 mos were randomized to placebo or sibutramine 15 mg/d for follow-up of 5 years. Weight changes for both arms were reported (Obesity 16:S64, 2008). To assess efficacy of behavioral program, only subjects in placebo arm are considered. Subjects lost weight in behavioral low-energy diet program using meal replacements. 180 subjects (women,129; men,51) who achieved maximum weight loss (low-weight) in the NPWLP from 3 to 36 mos prior to enrolling in the trial were eligible. Weight changes from low-weight were computed. In the trial, 1-hr behavioral weight maintenance classes were provided weekly for 12 mos; “Healthy Solutions” classes were provided for 8 consecutive weeks beginning at 18, 24, 30, 36, 42, 48 and 54 mos. Weight changes were assessed for subjects who completed 5 yrs of follow-up and for imputed values for subjects who had observation periods <5 yrs. Values were imputed as follows: 1. the percentage weight change at each quarter for completing subjects was calculated; 2. weight changes were assigned to quartiles; and 3. based on percentage changes at last 2 visits for non-completing subjects, missing values were imputed based on their quartile. Weight regains for yrs 1-5, respectively, were: 1st Quartile, 3.9%, 2.8%, 3.8%, 4.7% and 3.0%; and 4th Quartile, 13.4%, 25.3%, 34.9%, 45.7% and 53.6%. Mean weight loss in NPWLP (n=180) was 29.8 + 1.3 (SE) kg and low-weight was 85.0 + 1.3 kg. Over 5 yrs, completers (n=59) regained 25.4 + 2.9% but maintained weight losses of 8.8 + 2.0 kg below pre-NPWLP weight or 32.9 + 6.9% of peak weight loss. Using imputed weights over 5 yrs subjects (n=180) regained 27.4 + 1.6% but maintained weight losses of 6.8 + 1.2 kg or 21.8 + 4.4%. After 5 yrs, 26.6% of subjects (n=180) were above pre-NPWLP weight while 32.2% were maintaining weight losses of >10% of pre-NPWLP weights. Imputations of weight changes in weight maintenance studies may provide a better approximation of actual changes than other analyses. Imputed weight regain at 1, 2, 3, 4 and 5 yrs were: 11.4%, 18.2%, 22.6%, 26.5% and 27.4 %. Assessing quartiles of weight gain may help identify individuals who require more intensive intervention to avoid excessive weight regains.