Effectiveness of "run-ins" at predicting adherence in a behavioral weight loss efficacy trial.

There is limited research on whether run-in procedures predict participant adherence during behavioral efficacy trials. This study examined whether information from behavioral run-ins (food diary completion, questionnaire completion, and staff interview) predict intervention adherence, trial retention, and trial outcomes in a behavioral weight loss trial. Using run-in data, trial staff predicted which participants would have high, moderate, or low trial adherence. Participants with predicted high or moderate adherence were randomized. Results showed that predicted high adherers had better intervention adherence (session attendance and completion of self-monitoring records) and superior trial outcomes (i.e. weight loss). Run-in data did not predict trial retention. Results suggest that run-ins may be effective at identifying participants adherent to intervention protocols, thereby enhancing internal validity of behavioral efficacy trials.