Physical activity levels often decline with age and over two thirds of adults over the age of 60 sit for more than 8.5 hours of their waking day [1]. Physical activity can improve blood pressure, body composition, and overall health, which can prevent frailty and can contribute to a longer independent life [2,3]. Physical activity guidelines recommend adults over the age of 50 to perform moderate to vigorous physical activity for at least 150 minutes/week [4-6]. The guidelines define inactivity as “insufficient amounts of moderate to vigorous physical activity” and Katzmarzyk has noted that this accounts for only 3% of waking hours [4-7]. However, a sedentary lifestyle and physical inactivity are independent risk factors for chronic disease and a shortened lifespan [8-10]. Sedentary behavior is “any waking behaviour characterized by an energy expenditure of 1.5 or less metabolic equivalents while in a sitting or reclining posture” and includes sitting, television watching, computer use, and travel [11].

Home-based physical activity programs have significant potential for encouraging physical activity in older adults [12]. Yet, although sedentary persons can certainly adopt and continue regular long-term physical activity [13], daily adherence to home-based programs is low [14]. Successful interventions are often expensive and time intensive [15]. Programs that build self-awareness may be most amenable to lasting participation [15]. Goal setting has significant promise when it comes to promoting physical and dietary changes [16]. There is evidence that other behavior change techniques, such as self-monitoring, risk communication, and the use of social support can be beneficial additions to change-based interventions [17]. There is also some evidence that tailored feedback and information helps adults adopt physical activity and health-based activities [18-20].

The main challenge to implementing many of the evidence-based physical activity interventions and behavior change techniques is that they are resource intensive. In most cases, often multiple individuals are required to create tailored programs for individual patients, provide education, and then follow-up to promote adherence [21]. Where resources are scarce and demand is high, it would be ideal to have a simple, inexpensive, patient-managed intervention that incorporates behavior change techniques and can help patients improve and maintain activity levels.

Wearable activity trackers are a rapidly growing health-focused industry. There are many terms for the trackers, including electronic activity monitors, fitness trackers, wearable activity monitors, and wearables, but generally speaking, wearable activity trackers are devices that use sensors to help users automatically track step counts while aiming for a particular step count or activity goal. Pedometers are simple tools to promote self-awareness and self-monitoring of activity levels. Pedometer-based walking programs improve physical activity, body mass index, and blood pressure, even in adults who are already moderately active [22-26]. Improvements are typically greatest in programs that incorporate goals such as 10,000 steps daily [19].

The newest generation of pedometer is known as the wearable activity tracker. Similar to the behavioral strategies used in walking programs, the new trackers promote goal setting, self-efficacy, and tailored feedback through companion mobile apps and websites. The trackers do this by providing visual representations of activity data (step counts, altimetry, calories, sleep) over days and weeks. Some wearable activity trackers can also be used with social networking sites and other lifestyle apps for diet and stress, such as MyFitnessPal, Weight Watchers, and SparkPeople.

Although the new trackers have been designed as wellness devices rather than health or medical devices, they have considerable potential for use in health care. The BodyMedia FIT and Fitbit Zip trackers have been shown to provide valid measures of daily energy expenditures [27-30]. However, activity trackers may be less accurate for older adults with a shuffling, abnormal, or slow gait caused by conditions such as stroke or Parkinson’s disease [31-34]. There is growing evidence that these trackers do improve step count, and while there is acknowledged potential, there is currently little evidence yet that they improve health outcomes such as blood pressure or blood sugar [35-38].

The following conditions are necessary for persuasive technology to promote a behavior: motivation, physical ability, and an effective trigger [36]. Persuasive fitness technologies are attractive because they “automate” behavior change [39]. They offer convenient data collection, analysis, and storage over long periods with immediate automated feedback. Game-based features also offer points, leader boards, badges, trophies, vibrations, and progress bars to promote self-competition and building support and competitive networks, with the end result of encouraging and increasing physical activity [40,41].

Research has been emerging on the design and evaluation of persuasive technologies to promote physical activity. A matched case-control study with individuals who were members of the 10,000 steps Australia program found that individuals invited to use a website or mobile phone app were four times more likely to log their steps and 20 times more likely to achieve 10,000 daily steps [42]. Mobile phone (and app) developers are also starting to build pedometer functions into devices and using the phones’ main screen wallpapers to promote specific behaviors. For example, UbiFit Garden uses images of flowers (activities) and butterflies (goals) on the background screen of a user’s mobile phone to promote physical activity [43].

Wearable trackers have been increasingly identified as a tool for helping consumers prevent disease and increase physical activity [44,45]. Wearable activity monitors contain a wide range of behavior change techniques that have been often used in clinical behavioral interventions. In 2014, Lyon et al [46] identified 13 fitness trackers that provided tools for self-monitoring, feedback, and environmental change as well as other common behavior change techniques such as goal setting, self-monitoring, and feedback content that closely matched recommendations from social cognitive theory.

Emerging research into persuasive technology and wearable trackers offers significant promise for improving health and fitness. Research into better understanding how their design affects activity and behavior, how visualization helps both motivate and provide awareness, and how feedback can be better understood and used is a growing field [47-50]. Michie et al [51] discussed the wide range of approaches that can result in behavior change [51]. In 2003, Fogg [52] discussed how persuasive technologies employ a wide variety of strategies that influence behavior and activities, most notably drawing out self-monitoring and conditioning [52]. Self-monitoring is one of the most prevalent persuasive technologies; however, the most successful technologies often employ multiple strategies [53].

To date we know little about how older adults perceive new and emerging mobile health (mHealth) tools that include wearable activity trackers. While some studies conclude that health interventions that use technology are less effective than in-person interventions, there is limited evidence to support this conclusion [54-56]. Therefore, there is great potential in building a better understanding of mobile tools that can help older adults become more active on their own time, in their own space. Our objective was to examine the usability and usefulness of wearable activity trackers for older adults living with chronic illness as a first step to better understand how wearable fitness trackers can help older adults become healthier.

Our research design was inspired by a similar study on mobile medication management apps [57]. We used a mixed-methods approach to examine the perceived acceptability of popular wearable technologies designed to promote fitness and healthy living. The research was conducted at the University of Waterloo School of Pharmacy. We qualitatively assessed user acceptance using a thematic analysis, which is reported according to the consolidated criteria for reporting qualitative research [58]. We also developed a technology acceptance questionnaire based on the Technology Acceptance Model (TAM) [59]. We did not develop any of the devices we tested and received ethical approval from the University of Waterloo Office of Research Ethics (Certificate Number 19440).

We made the following assumptions: (1) most older adults are not using wearable activity trackers; (2) many have chronic illnesses that would benefit from increased physical activity; (3) wearable technologies have the potential to improve activity levels; and (4) there are age-related barriers unique to older users. We also assumed that most older adults are not early adopters and we were therefore asking participants to assume the early adopter role for the duration of this study.

To assess acceptability of commercially available wearable activity trackers, we began by identifying all devices available to Canadian consumers as of November 1, 2013. We identified and reviewed 4 devices by Fitbit, 2 devices by Jawbone, and 1 device each from Withings, Misfit, and Nike. The research team selected 4 devices for testing purpose, with each device representing a different feature available with activity trackers (Table 1). All 4 devices used an accelerometer to assess steps (pedometer). We chose the Fitbit Zip because it was inexpensive, could be clipped to clothing, and allowed users to track activity on a simple interactive screen. We selected the Jawbone Up 24 as it could be worn on the wrist and could collect data on sleep quality. We selected the Misfit Shine because it could be worn on the wrist or clothing, was waterproof, and could double as a watch. Finally, we selected the Withings Pulse because it could capture heart rate data.

Focus groups were conducted at the end of each study group. We chose to do a high-level thematic analysis with the purpose of examining and recording the patterns in our data [66]. We chose thematic analysis because it helped us organize the data into themes to offer a rich description of participant experiences. Thematic analysis goes further than just counting phrases or words in a text and works to identify the implicit and explicit ideas of the data [67]. Borrowing from the grounded theory approach, we did not identify any preconceived theories before starting data collection on the usability or usefulness of the wearable activity trackers for older adults with chronic disease [68]. This gave us the opportunity to incorporate fresh viewpoints from the diverse settings of participant lives, rather than from our own perspectives as heavy users of technology. The final dataset used in the thematic analysis included participants’ written notes, researcher observations, and the recorded and transcribed focus group discussions.

The TAM identifies two key beliefs as the primary reason for behaviors that encourage computer and technology acceptance: perceived usefulness and perceived ease of use, with the idea being that if people think that using a technology will make their life easier or think a technology is easy to use, then they will be more likely to adopt the technology [69]. In particular, TAM identifies that using technology, both new and established, is associated with behavioral intention, specifically that people form intentions to use technologies when they have a positive feeling about the technology in question. TAM identified the following areas that can influence adoption: external variables, perceived usefulness (U), perceived ease of use (e), attitude toward using (A), behavioral intention to use (BI), and actual system use. It postulates that BI = A + U and stated that people form intentions about using new technologies based on how they perceive the technology will improve their performance.

Our data were coded and analyzed in NVivo (QSR International) in 3 stages. For the initial stage of analysis, 1 independent researcher (KM) coded data by briefly summarizing each line and then paragraph of data. In the second stage of coding, the codes were combined into themes by 2 researchers (KM and KG). In the third and final stage, each theme was populated by representative quotations (KM and KG).

We recruited a purposive sample of 32 participants aged between 52 and 85 years (mean 64 years; Table 2). Participants were included if they were over the age of 50 and self-reported a diagnosis of one of the following: hypertension, hyperlipidemia, diabetes, osteoarthritis, rheumatoid arthritis, and osteoporosis (including a history of non-traumatic fracture of the wrist, spine or hip). We also included participants who had a significant family history of cardiovascular disease or diabetes (a parent, sibling, or child has developed cardiovascular disease or type E diabetes before the age of 55) and a parental history of hip fracture.

Because our goal was to explore acceptability, which is based on participant’s perceptions of usability and usefulness, a sample size of 32 was adequate. We also continued sampling until data saturation was reached and no new ideas or issues were identified [70].

Of the 32 participants, 30 completed the testing of all devices; 2 users dropped out after testing the pedometer (initial stage), citing acute viral illness. All the wearable activity trackers tested had moderate acceptability, with the standard pedometers having the lowest acceptability to users (Table 4).

In general, all the wearable activity trackers tested had a similar score for each item of the TAM questionnaire and were rated higher than the standard pedometer in all items but 13 and 14, both of which are related to baseline knowledge of the technology. This coincides with the exit interview ratings, where participants constantly rated the standard pedometer as their least preferred option.

The language participants used about wearable activity trackers was also notable. Participants would pick a favorite device, and stick with that device, even if through later discussion they identified more negatives than positives about that device. The common reasons for liking or disliking a device are how they looked, be it for subtle or fashion-based factors, and ease of use. The participants’ perceived comfort level with the devices was another notable aspect of this study that drove how participants ranked the devices. We also found that the wearable trackers performed highest on items assessing ease of use, namely, Item 10 “Overall, the activity trackers were easy to use” and Item 14 “I have the knowledge necessary to use the activity tracker.” The highest rated item was Question 16, “The activity tracker was comfortable to wear.” This was reflected by low scores for Item 7 “I found it easy to learn to operate the activity tracker” for the Withings and pedometer (2.84) compared with the Fitbit (3.27), Jawbone (3.13), and Shine (3.00), which contributed to decisions to purchase or not purchase (Table 3).

After testing the devices, none of our participants planned to purchase a Mio or Sportline pedometer, whereas 22 of the 30 participants who completed the study said they would purchase a wearable activity tracker. The participants were asked which device or devices they would potentially purchase, and after trying the devices, 30% (N=32) felt they would buy a Jawbone Up 24, 30% would purchase a Misfit Shine, 33% felt they would buy a Withings Pulse, and 40% felt they would buy a Fitbit Zip. At the completion of the study, 73% (22/30) felt they would purchase a tracker and 67% (20/30) felt they would purchase a device for a friend or family member. Those who felt they would not purchase a tracker included reasons of cost, lack of interest, and complexity of the devices. Tables 5 and 6 present the positive and negative associated with individual fitness trackers, respectively.

In this study, 4 overarching themes emerged to describe how acceptable wearable activity trackers are for older adults with chronic disease. The first theme is that new and emerging consumer health technologies are likely to be outside the older user’s perceived comfort zone. However, the second theme is that after a brief trial period, users can appreciate that wearable activity trackers improve self-awareness and goal setting. The third theme was that wearable activity trackers are ultimately more useful as motivators than as quantifiers. The final and fourth theme was that older adults are unlikely to adopt wearable activity trackers if the trackers are not sold and managed as health care devices.