Assessing the quality of mobile phone apps for weight management: user-centered study with employees from a Lebanese university

Background: Evaluating the quality of mobile health apps for weight loss and weight management isimportant to understand whether these can be used for obesity prevention and treatment. Recent reviews call for more research on multidimensional aspects of app quality, especially involving end users, as there are already many expert reviews on this domain. However, no quantitative study has investigated how lay persons see popular apps for weight management and perceive different dimensions of app quality.

Objective: This study aimed to explore how laypersons evaluate the quality of 6 free weight management apps (My Diet Coach, SparkPeople, Lark, MyFitnessPal, MyPlate, and MyDiet Diary), which achieved the highest quality ratings in a related andrecent expert review.

Methods: A user-centered study was conducted with 36 employees of a Lebanese university. Participants enrolled in the study on a rolling basis between October 2016 and March 2017. Participants were randomly assigned an app to use for 2 weeks. App quality was evaluated at the end of the trial period using the Mobile App Rating Scale userversion (uMARS). uMARS assesses the dimensions of engagement, functionality, aesthetics, information, and subjectivequality on 5-point scales. Internal consistency and interrater agreement were examined. The associations between uMARS scores and users’ demographic characteristics were also explored using nonparametric tests. Analyses were completed in November 2017. 

Results: Overall, the 6 apps were of moderately good quality (median uMARS score 3.6, interquartile range[IQR] 0.3). The highest total uMARS scores were achieved by Lark (mean 4.0 [SD 0.5]) and MyPlate (mean 3.8 [SD 0.4]), which alsoachieved the highest subjective quality scores (Lark:mean 3.3 [SD 1.4]; MyPlate: mean 3.3[SD 0.8]). Functionality was thedomain with the highest rating (median 3.9, IQR 0.3), followed by aesthetics (median 3.7, IQR 0.5), information (median 3.7, IQR 0.1), and engagement (median 3.3, IQR 0.2). Subjective quality was judged low (median2.5, IQR 0.9). Overall, subjective quality wasstrongly and positively related (P<.001)with total uMARS score (ρ=.75), engagement (ρ=.68), information, and aesthetics (ρ=.60)but not functionality (ρ=.40; P=.02). Higher engagement scoreswere reported among healthy (P=.003) andobese individuals (P=.03), who also showedhigher total uMARS (P=.04) and subjective quality (P=.05)scores. 

Conclusions: Although the apps were considered highly functional, they were relatively weak in engagement and subjective quality scores, indicating a low propensity of usingthe apps in the future. As engagement was the subdomain most strongly associated with subjective quality, app developers and researchers should focuson creating engaging apps, holding constant the functionality, aesthetics, and information quality. The tested apps (in particular Lark and MyPlate) were perceived as more engaging and of higher quality among healthy, obese individuals, making them a promising mode of delivery for self-directed interventions promoting weight control among the sampled population or in similar and comparable settings.