Are we sure that Mobile Health is really mobile? An examination of mobile device use during two remotely-delivered weight loss interventions

Abstract

Background: The "m" in mHealth is often thought of as the ability to receive health information and monitor behaviors on the go. Little is known about how people actually use mobile vs. traditional access methods and if access method affects engagement and health outcomes.

Methods: This study examines the 3-month outcomes of two mobile weight loss interventions (Pounds Off Digitally (POD) and mobile POD (mPOD)) where participants were required to own a mobile device for study entry and received weight loss information via podcast. Only participants in both studies who were randomized to receive the same theory-based podcast (TBP) were used in this analysis. In POD, 41 participants were randomized to the TBP condition (37 to a control not included in this analyses). In mPOD, 49 participants were randomized to the TBP (n=49) and 47 to the TBP+mobile group (a self-monitoring app and Twitter app for social support). The goal of this study is to examine how participants accessed study components and to examine how type of device impacts engagement and weight loss.

Results: Examining data from both studies in aggregate, despite a mobile delivery method, 58% of participants reported using a non-mobile device to access the majority of the podcasts (desktop computers), 76% accessed the podcasts mostly at their home or work, and 62% were mainly non-mobile (e.g., sitting at work) when listening. Examining objective download data for mPOD, 49% of downloads (2889/5944) originated from non-mobile delivery methods vs. mobile platforms (3055/5944). At 3 months, 55% of Twitter posts originated from the website (n=665 posts) vs. a mobile app (n=540; 45%). There was no difference in the number of podcasts participants reported listening to by device. There were more Twitter posts by mobile app users (51±11) than Twitter website users (23±6, p<0.05). There was a trend (p=0.055) in greater weight loss among mobile users for podcasts (-3.5±0.5%) as compared to non-mobile users (-2.5±0.5%). Weight loss was significantly greater in Twitter mobile app users (-5.6±0.9%) than website users (-2.2±0.5%, p<0.01).

Conclusion: Type of device used for podcast listening did not affect engagement but there was a trend toward greater weight loss among mobile users. Method of Twitter posting was associated with engagement and weight loss with mobile app users posting more to Twitter and losing more weight.

Keywords: Technology; Telemedicine; Weight loss.