PROJECT SUMMARY/ABSTRACT
Decades of research have demonstrated that behavioral obesity treatments can produce clinically significant
weight losses that improve health and disease risk/severity. However, these treatments have not been
disseminated widely due to high costs and lack of qualified providers. We therefore aimed to develop a fully
automated online obesity treatment that would produce clinically significant weight losses (i.e., = 5 % of initial
body weight) when delivered online in a variety of settings. Our 3-month program, Rx Weight Loss (RxWL),
produced clinically significant mean±SD weight losses of 5.8±4.4 % of initial body in patients referred by
physician (N=154), 5.8 ±5.2 % in worksites (N=75), and 4.2±5.3 % in community settings (N=230). These
weight losses were largely maintained at 6-month follow-up. Given these positive results, healthcare
stakeholders have expressed a desire to pursue dissemination of RxWL. In order to maximize the public health
impact of disseminating RxWL and similar programs, and advance the science of online behavioral intervention
in general, it is imperative to evaluate innovative behavioral intervention components with the potential to
optimize weight loss outcomes. Despite promising mean weight losses, only about half of participants treated
online achieve weight losses of > 5%. Furthermore, the health benefits of weight loss are dose dependent;
optimizing RxWL to produce larger mean weight losses would therefore enhance its health benefits. Optimizing
RxWL would also provide a template for optimizing other online treatments. Because digital health technology
evolves rapidly, we will use the Multiphase Optimization Strategy (MOST) framework to most quickly and
efficiently determine which, if any, of 5 innovative intervention components, alone or in combination, increases
the proportion of patients achieving a =5% weight loss, and mean weight loss, of the RxWL program at 12-
months. The 5 intervention components to be tested are: (a) Web-based virtual reality intervention for training
in basic behavioral weight loss skills; tailored interactive intervention targeting (b) structured physical activity
and (c) dysregulated eating; (d) a platform for social interaction including opportunities for friendly competition,
and (e) interactive video feedback with content tailored to the unique needs of each participant and a focus on
dietary skills. After initial testing with N=32 to refine and finalize the intervention components with feedback
from advisory boards, N=384 individuals with BMI = 25 will be randomized to receive RxWL and 0-5 of the
experimental intervention components in a full factorial experiment. This design will allow us to determine
which intervention components maximize weight loss and whether there are favorable combinations of
components. In addition, by evaluating the effects of each component on proximal outcomes (i.e., mediators)
we will learn not only which components are (or are not) effective but also why or how they exert their effects.
This project advances the science of behavioral obesity treatment, and will directly impact the care of patients
receiving RxWL.
