Lifestyle interventions, while effective at reducing weight and diabetes risk, are intensive (i.e., requiring 16 or more face-to-face visits) which has prevented widespread implementation. Mobile technology may reduce intervention intensity while preserving outcomes by assisting in the delivery of behavioral strategies, but very little research has explored this. Weight loss mobile applications are proliferating in the open market;however our work and others2 show that the range of evidence-based strategies addressed by these apps is narrow, primarily including self-monitoring, prompts, goal-setting, and sometimes a social network.3 A key strategy missing across apps in both the market and research is problem solving, an essential component of behavioral weight loss interventions.4 We propose to develop and test the feasibility of Smart Coach, a weight loss mobile application that includes common features such as self-monitoring, goal setting, and a social network, but even more importantly, an avatar-facilitated, idiographic problem solving feature that processes information intelligently to help patients identify solutions to their weight loss problems. We hypothesize tha Smart Coach when combined with a lower intensity (half the sessions) weight loss intervention will be more effective than a lower intensity weight loss intervention alone, with biggest differences observable after face-to- face visits end. Using a "crowd-sourcing" model, we will populate a database with problems and solutions via 1) expert-delivered problem solving sessions with a sample of obese participants trying to lose weight and 2) a pre-pilot test of the app. Using principles of "artificial intelligence" we will convert the algorithm of problem solving counseling into the mobile application so that it may perform this strategy on its own, but based on expert and crowd-sourced information. We will then use a series of iterative steps involving qualitative research methods (usability testing, focus groups, and pre-piloting) to refine the tool A randomized pilot feasibility trial will test the feasibility and initial effects of the Smart Coah mobile application when paired with a shortened (8- week) behavioral weight loss intervention relative to a shortened behavioral weight intervention alone. Feasibility outcomes include frequency and duration of usage of the mobile app and each feature, recruitment, and retention. We will also do exploratory analyses comparing conditions on problem solving skills and weight loss at 8- and 16-weeks. Data will support a larger efficacy trial of a Smart Coach-assisted brief behavioral weight loss intervention relative to a brief behavioral weight loss intervention alone. Our overarching goal is to develop mobile technology that reduces the intensity of lifestyle interventions as far as possible while preserving weight loss outcomes, to ultimately broaden the reach to people and settings that currently have little access.

Public Health Relevance

Mobile technology that effectively delivers behavioral strategies to reduce the intensity of weight loss intervention while also preserving outcomes is needed to increase the impact and reach of behavioral weight loss interventions. A key strategy missing across mobile applications developed in both the market and research is problem solving, an active ingredient and central feature of behavioral weight loss interventions. We propose to develop and test the feasibility of Smart Coach, a weight loss mobile application that includes common features such as self-monitoring, goal setting, and a social network, but even more importantly an avatar-facilitated, idiographic problem solving feature that processes information intelligently to help patients identify solutions to their weight loss problems. We hypothesize that Smart Coach when combined with a lower intensity (half the sessions as the standard program) weight loss intervention will be more effective than a lower intensity weight loss intervention alone.

Agency
National Institute of Health (NIH)
Institute
National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK)
Type
Exploratory/Developmental Grants (R21)
Project #
1R21DK098556-01
Application #
8492838
Study Section
Psychosocial Risk and Disease Prevention Study Section (PRDP)
Program Officer
Kuczmarski, Robert J
Project Start
2013-09-15
Project End
2015-05-31
Budget Start
2013-09-15
Budget End
2014-05-31
Support Year
1
Fiscal Year
2013
Total Cost
$263,542
Indirect Cost
$92,067
Name
University of Massachusetts Medical School Worcester
Department
Internal Medicine/Medicine
Type
Schools of Medicine
DUNS #
603847393
City
Worcester
State
MA
Country
United States
Zip Code
01655