*Abby King; Erik Hekler, MS, PhD; Lauren Grieco, PhD; Sandra Winter, PhD; Jylana Sheats, PhD, MPH; Matt Buman, PhD; Banny Banerjee; Jesse Cirimele; Thomas N. Robinson, MD, MPH; Beth Mezias; Frank Chen
Stanford University School of Medicine
Oral Presentation – Research Track
Sunday, Sept 30, 2012: 10:18 AM – 10:38 AM – LK120

*Presenting Speaker

Background
Mobile devices offer a potentially powerful means for delivering ongoing advice and support for improving key health behaviors impacting health and wellbeing. While there has been an explosion of mobile phone applications (apps) aimed at physical activity and other health behaviors, few are based on scientific theory and evidence. A major objective of this research is to evaluate the efficacy of smartphone apps harnessing different motivational frames to promote initial physical activity adoption. An additional aim has been to develop apps that are “user-friendly” irrespective of user technology skills or experience.

Methods
Twenty-nine inactive adults ages 45 years and older without smartphone experience were introduced to the smartphones, which they carried for a week to establish baseline physical activity levels via the smartphone’s built-in accelerometer. Participants were then randomized to one of three custom apps focused on increasing physical activity and decreasing sedentary behaviors, or to a commercially available app used as a control (Calorific). The Cognitive app focused on goal-setting, problem-solving, and self-monitoring. The Social app focused on social comparisons, norms, and support. The Affect app fostered attachment to an avatar (a bird) along with positive reinforcement and game-like feedback. The primary outcome—health-enhancing moderate or more vigorous physical activity (MVPA)–was measured continuously using the smartphone’s built-in accelerometer, which was verified against standard Actigraph accelerometry (R2=0.83).

Results
The mean age of the sample was 58 years, 65% were women, 87% were White, and 55% worked full-time. Across 2 months, participants randomized to the Social app engaged in 16.3 more minutes per day (114 more minutes per week) of MVPA compared to Control (p<.01), and those assigned to the Affect app engaged in 11.2 more minutes per day (78 more minutes per week) of MVPA than Control (p<.01). The Cognitive app did not produce significantly more MVPA than Control (p=.39).

Conclusions
Both the Social and Affect apps resulted in clinically meaningful weekly increases in health-enhancing physical activity relative to a Control (nutrition) app in a smartphone-naïve sample of midlife and older adults. A second experiment is being conducted evaluating refined versions of the three apps. Additional questions of interest that are being explored are impacts of the three apps on decreasing daily sedentary behavior and increasing lighter forms of physical activity throughout the day; identification of the key features through which the different apps exert their effects; and gaining a broader understanding of which apps may be most suitable for which types of people. We are also in the initial stages of a project aimed at translating and testing the apps in a largely Latino patient population.