Socio-demographic determinants of physical activity and app usage from smartphone data

Long-term usage of a commercial mHealth app: A "multiple-lives" perspective

Background

Emerging evidence suggests that individuals use mHealth apps in multiple disjointed ways in the real-world-individuals, for example, may engage, take breaks, and re-engage with these apps. To our knowledge, very few studies have adopted this 'multiple-live' perspective to analyze long-term usage of a physical activity (PA) app. This study aimed to examine the duration of use, as well as the frequency, length, and timing of streaks (uninterrupted periods of use) and breaks (uninterrupted periods of non-use) within a popular commercial PA app called Carrot Rewards over 12 months. We also examined sociodemographic correlates of usage.

Method

This retrospective observational study analyzed data from 41,207 Carrot Rewards users participating in the "Steps" walking program from June/July 2016 to June/July 2017. We measured four usage indicators: duration of use, frequency and length of streaks and breaks, time to first break, and time to resume second streak. We also extracted information regarding participants' age, gender, province, and proxy indicators of socioeconomic status derived from census data. We used descriptive statistics to summarize usage patterns, Kaplan-Meier curves to illustrate the time to first break and time to resume second streak. We used linear regressions and Cox Proportional Hazard regression models to examine sociodemographic correlates of usage.

Results

Over 60% of the participants used Carrot Rewards for ≥6 months and 29% used it for 12 months (mean = 32.59 ± 18.435 weeks). The frequency of streaks and breaks ranged from 1 to 9 (mean = 1.61 ± 1.04 times). The mean streak and break length were 20.22 ± 18.26 and 16.14 ± 15.74 weeks, respectively. The median time to first break was 18 weeks across gender groups and provinces; the median time for participants to resume the second streak was between 12 and 32 weeks. Being female, older, and living in a community with greater post-secondary education levels were associated with increased usage.

Conclusion

This study provides empirical evidence that long-term mHealth app usage is possible. In this context, it was common for users to take breaks and re-engage with Carrot Rewards. When designing and evaluating PA apps, therefore, interventionists should consider the 'multiple-lives' perspective described here, as well as the impact of gender and age.

Profile of adults users of smartphone applications for monitoring the level of physical activity and associated factors: A cross-sectional study

Background

There are currently more than 200 million smartphones in Brazil. The potential of mobile technologies for favorable changes in health behavior such as physical activity has been previously described in the literature. Results of surveys in developed countries indicate that applications (APPs) are developed for people who are better educated, younger, and with higher incomes compared to non-users. However, the profile of users in developing countries like Brazil is not well-known. Understanding the profile of APP users might ease the development turned to physically inactive people and those at higher cardiovascular risk. Furthermore, the physiological and functional factors associated with the use of such APP are unknown.

Objectives

To characterize the profile of APP users to monitor the physical activity level (PAL) and assess the demographic, socioeconomic, clinical, physiological, and functional characteristics associated with the use of smartphone APPs to monitor physical activity in Brazilian adults.

Methods

We assessed 176 asymptomatic men and 178 women (43 ± 12 years; 27 ± 5 kg/m2). We initially asked participants about their current use of a smartphone APP containing PAL monitoring functionality, such as exercise session logs and/or step counts. In a cross-sectional design, we investigated schooling, socioeconomic status (Critério Brasil), and classic self-reported cardiovascular risk factors. We evaluated several physiological and functional variables such as maximum O2 consumption on a treadmill (VO2 max), blood pressure, body composition (bioelectrical impedance), handgrip strength, and isokinetic muscle strength of the dominant lower limb. Participants used a triaxial accelerometer for 7 days to quantify daily physical activity. We also assessed health-related quality of life (WHOQOL BREF), perceived stress (PSS14 Scale), and the built environment (NEWS Scale). We compared continuous variables using the Student's t-test and categorical variables using the χ2 test, between APP users and non-users. After univariate analysis, we included the main variables associated with the use of APP in a multiple logistic regression model.

Results

One hundred and two participants (28.3%), unrelated to gender, reported using a smartphone APP for physical activity at the time of assessment. Except for perceived stress and the built environment that were not associated with the use of APP, users of APP were younger and had higher education, lower cardiovascular risk, better socioeconomic status, a better quality of life, better cardiorespiratory function, better body composition, greater physical fitness and more moderate to vigorous physical activity in daily life. The results of the multiple logistic regression showed that age, hypertension, VO2 max, socioeconomic status (Critério Brasil), and quality of life (WHOQOL BREF total score) were the variables most significantly associated with the use of the APP.

Conclusions

Our results indicate that smartphone APPs to monitor physical activity are developed for younger adults with better socioeconomic status, lower cardiovascular risk, higher quality of life, and greater cardiorespiratory fitness. Greater efforts are needed to develop a science-based APP for people who most need this technology, enabling greater potential to prevent undesirable health outcomes in asymptomatic adults.

Clustering Accelerometer Activity Patterns from the UK Biobank Cohort

Many researchers are beginning to adopt the use of wrist-worn accelerometers to objectively measure personal activity levels. Data from these devices are often used to summarise such activity in terms of averages, variances, exceedances, and patterns within a profile. In this study, we report the development of a clustering utilising the whole activity profile. This was achieved using the robust clustering technique of k-medoids applied to an extensive data set of over 90,000 activity profiles, collected as part of the UK Biobank study. We identified nine distinct activity profiles in these data, which captured both the pattern of activity throughout a week and the intensity of the activity: "Active 9 to 5", "Active", "Morning Movers", "Get up and Active", "Live for the Weekend", "Moderates", "Leisurely 9 to 5", "Sedate" and "Inactive". These patterns are differentiated by sociodemographic, socioeconomic, and health and circadian rhythm data collected by UK Biobank. The utility of these findings are that they sit alongside existing summary measures of physical activity to provide a way to typify distinct activity patterns that may help to explain other health and morbidity outcomes, e.g., BMI or COVID-19. This research will be returned to the UK Biobank for other researchers to use.

Characterisation of Temporal Patterns in Step Count Behaviour from Smartphone App Data: An Unsupervised Machine Learning Approach

The increasing ubiquity of smartphone data, with greater spatial and temporal coverage than achieved by traditional study designs, have the potential to provide insight into habitual physical activity patterns. This study implements and evaluates the utility of both K-means clustering and agglomerative hierarchical clustering methods in identifying weekly and yearlong physical activity behaviour trends. Characterising the demographics and choice of activity type within the identified clusters of behaviour. Across all seven clusters of seasonal activity behaviour identified, daylight saving was shown to play a key role in influencing behaviour, with increased activity in summer months. Investigation into weekly behaviours identified six clusters with varied roles, of weekday versus weekend, on the likelihood of meeting physical activity guidelines. Preferred type of physical activity likewise varied between clusters, with gender and age strongly associated with cluster membership. Key relationships are identified between weekly clusters and seasonal activity behaviour clusters, demonstrating how short-term behaviours contribute to longer-term activity patterns. Utilising unsupervised machine learning, this study demonstrates how the volume and richness of secondary app data can allow us to move away from aggregate measures of physical activity to better understand temporal variations in habitual physical activity behaviour.