Pre-Enrollment Steps and Run-Ins in Weight Loss Trials: A Meta-Regression

Abstract

Introduction: The generalizability of study findings may be influenced by pre-enrollment trial procedures, including the use of behavioral run-in periods. The study goals were to determine whether behavioral run-in periods and other pre-enrollment processes affect outcomes in randomized trials of behavioral weight loss interventions that have contributed to clinical guidelines.

Methods: The sample was behavioral weight loss intervention trials included in the 2018 U.S. Preventive Services Task Force systematic review. Information on pre-enrollment processes (total steps, in-person steps, behavioral run-in) was abstracted, and meta-regressions were conducted in 2022 to test whether pre-enrollment processes were associated with weight loss at 6 or 12 months and trial retention at 12 months.

Results: Across 80 trials, the median number of total pre-enrollment steps was 2 (range=1-5), and that of in-person pre-enrollment steps was 1 (range=0-4). Almost one-third of the trials (k=24; 30%) used a behavioral run-in. The most common run-in tasks were self-monitoring physical activity (k=9) or both physical activity and diet (k=6). Greater weight loss was observed in trials with behavioral run-ins at 6 months (-2.33 kg; 95% CI= -3.72, -0.93) and, to an attenuated extent, at 12 months (-0.86 kg; 95% CI= -1.72, 0.01) compared to those without run-ins. The total number of pre-enrollment steps was also associated with greater 6-month weight loss (-0.85 kg; 95% CI= -1.59, -0.11). Higher retention was associated with total number of pre-enrollment steps and in-person steps and marginally with the presence of run-ins.

Discussion: The use of more pre-enrollment processes is associated with greater weight loss in behavioral weight loss trials and may impact the generalizability of outcomes.