Health-At-Every-Size and eating behaviors: 1-year follow-up results of a size acceptance intervention.

Background: Poor long-term success observed with current weight-control strategies stresses the relevance to explore new weight management approaches.
Objective: To assess the effects of a Health-At-Every-Size (HAES) intervention on eating behaviors, appetite sensations, metabolic and anthropometric variables, and physical activity levels in women at 6-month and 1-year postintervention.
Design: Randomized controlled trial; measurements at baseline, at the end of the intervention period (4 months), and at 6-month and 1-year postintervention (10 months and 16 months, respectively). Intervention and testing took place from September 2003 to August 2006.
Participants/setting: Premenopausal overweight/obese women (n=144; mean age of 42.3+/-5.6 years), recruited from free-living, general community.
Intervention: Women were randomly assigned to: HAES group (n=48), social support group (n=48), or control group (n=48).
Main Outcome Measures: Eating behaviors (restraint, disinhibition, and susceptibility to hunger), appetite ratings (desire to eat, hunger, fullness, and prospective food consumption), anthropometric and metabolic variables (body mass index, waist and hip circumferences, blood lipid levels, and blood pressure), and engagement in moderate to intense physical activity (energy cost > or =1.2 k cal x kg(-1) x 15 minutes(-1) [> or =4.8 metabolic equivalents]).
Statistical Analyses Performed: Changes for each dependent variable were assessed by linear mixed models according to a group (HAES vs social support vs control) by time (baseline vs 4 months vs 10 months vs 16 months) split-plot design.
Results: Situational susceptibility to disinhibition and susceptibility to hunger significantly decreased over time in both HAES group (-0.9+/-0.2 and -1.3+/-0.5, respectively) and the social support group (-0.4+/- 0.2 and -1.4+/-0.5, respectively). Although eating behavior scores observed at 16 months did not differ between HAES and social support groups (situational susceptibility to disinhibition: 2.5+/-0.2 in HAES group vs 2.7 +/- 0.2 in social support group; susceptibility to hunger: 4.2+/-0.5 in both groups), they were lower in these groups than scores noted in the control group (3.3+/-0.2 for situational susceptibility to disinhibition and 5.9+/-0.5 for susceptibility to hunger).
Conclusions: These results suggest that, when compared to a control group, an HAES approach could have long-term beneficial effects on eating behaviors related to disinhibition and hunger. In addition, our study did not show distinctive effects of the HAES approach in comparison to a social support intervention.