This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
It is well documented that meeting the guideline levels (150 minutes per week) of moderate-to-vigorous physical activity (PA) is protective against chronic disease. Conversely, emerging evidence indicates the deleterious effects of prolonged sitting. Therefore, there is a need to change both behaviors. Self-monitoring of behavior is one of the most robust behavior-change techniques available. The growing number of technologies in the consumer electronics sector provides a unique opportunity for individuals to self-monitor their behavior.
The aim of this study is to review the characteristics and measurement properties of currently available self-monitoring devices for sedentary time and/or PA.
To identify technologies, four scientific databases were systematically searched using key terms related to behavior, measurement, and population. Articles published through October 2015 were identified. To identify technologies from the consumer electronic sector, systematic searches of three Internet search engines were also performed through to October 1, 2015.
The initial database searches identified 46 devices and the Internet search engines identified 100 devices yielding a total of 146 technologies. Of these, 64 were further removed because they were currently unavailable for purchase or there was no evidence that they were designed for, had been used in, or could readily be modified for self-monitoring purposes. The remaining 82 technologies were included in this review (73 devices self-monitored PA, 9 devices self-monitored sedentary time). Of the 82 devices included, this review identified no published articles in which these devices were used for the purpose of self-monitoring PA and/or sedentary behavior; however, a number of technologies were found via Internet searches that matched the criteria for self-monitoring and provided immediate feedback on PA (ActiGraph Link, Microsoft Band, and Garmin Vivofit) and sedentary time (activPAL VT, the Lumo Back, and Darma).
There are a large number of devices that self-monitor PA; however, there is a greater need for the development of tools to self-monitor sedentary time. The novelty of these devices means they have yet to be used in behavior change interventions, although the growing field of wearable technology may facilitate this to change.
Modern environments and technological advancements have radically altered the way we live our lives [
Over the last decade, there has been a plethora of tools developed to support PA and sedentary behavior change, of which the greatest growth has been seen in self-monitoring tools. Self-monitoring is defined as “a person closely and deliberately monitors their own behavior” [
Traditionally, self-monitoring of PA and sedentary time occurred via paper-based journal methods [
This increased availability of electronic self-monitoring devices provides an opportunity for researchers to utilize these novel technologies as an aid for behavior change in PA and sedentary behavior on a large scale. Furthermore, wearable technologies are increasingly integrating health care systems. Recent reports from the National Information Board in a review of the National Health Service in the United Kingdom indicate the need for “citizens” to start playing a more active role in their health care by accessing, entering, and uploading data into their own online medical record. Under these new plans, citizens will be able to access and download their detailed medical records as well as contribute to it with information from their personal wearable technology or biosensors [
Therefore, it seems timely to review the characteristics and measurement properties (eg, wear location, integrated sensors, outcomes measured) of currently available self-monitoring devices, both those marketed to consumers and those used in research settings, that have been (or could be) utilized in, or developed for, real-time self-monitoring of sedentary behavior and/or PA.
The search strategy was built around three groups of keywords: behavior (ie, PA and sedentary behavior), measurement, and population. A detailed description of the keywords used and method of combination can be found in
Scopus, Medline, Web of Science, and the Institute of Electrical and Electronic Engineers (IEEE) databases were searched using these keywords from the inception of the databases to October 1, 2015. In addition, manual searches of personal files were conducted and reference lists of primary studies were screened.
Because of the rapid release of technology in the consumer electronic area, a grey literature search of relevant websites was conducted for technologies that allow for the self-monitoring of PA and sedentary time but may not have made it into the published research to date. Keywords based on the same groups as the database searches were used to search the Internet engines Google, Bing, and Yahoo. Searches were extracted for later review using a specialized browser plug-in. The first 200 search results from each search engine were extracted for further review; this was a pragmatic approach because it was deemed that results after the first 200 were either not relevant or repetitive. This ensured that the results were unaffected by the changing algorithms of Web search engines. Searches were completed on October 1, 2015.
Two sets of inclusion criteria were developed for research articles and websites. For inclusion in the review, studies were required to (1) include adults aged 18 years or older, (2) be published in English, and (3) describe a device that objectively self-monitors PA, physical inactivity, and/or sedentary time/sitting and can, or has the potential to, provide feedback to the user. Traditionally, there would also be a criteria based around study type; however, in order to obtain the widest variety of devices, this was not included.
For inclusion in the review, only websites from manufacturers were included (ie, blogs or consumer reviews pertaining to technologies of interest were excluded) and devices that had the ability to self-monitor and were available for purchase at the time of the review were included.
Potentially relevant articles were selected by screening titles, screening abstracts, and if abstracts were not available or did not provide sufficient data, the entire article was sought and screened to determine whether it met the inclusion criteria. Relevant websites were selected by screening webpage titles and screening devices on relevant webpages to determine whether it met the inclusion criteria. Data were extracted on standardized forms developed for this review.
Information on the devices was extracted from articles and cross-referenced with the device manufacturer’s information. Validity data on each device was not extracted; instead, articles with relevant validity data, where available [
A 10% subsample of potentially relevant articles retrieved for full-paper screening were extracted by a second author (AL) to determine interrater agreement. Interrater agreement was high (Cohen’s kappa=.81). If any discrepancies arose, these were resolved by discussion between authors.
Each device was designated a self-monitoring code: (1) yes, self-monitors PA (YPA); (2) yes, self-monitors PA and physical inactivity, such as self-monitoring and feedback on lack of movement (YPI); and (3) yes, self-monitors sedentary time (YST).
The different attributes of the self-monitoring devices were based on Control Theory [
Auditory: feedback on behavior provided verbally from device (eg, via Sensoria voice-over feedback regarding ground contact from smartphone/smart MP3)
Vibratory: haptic feedback on predetermined behavioral thresholds provided using vibrations (eg, LumoBack)
Omnipresent: feedback that is visible all the time, usually in the form of a progression bar that changes with advancement toward predetermined goals (eg, Fitbit Flower)
Push notification: the delivery of information regarding behavioral goals from a software app to a computing device without a specific request from the user
Immediate: whether the data/feedback are immediate in its return to the user (eg, LumoBack)
Delayed: whether the data/feedback are delayed in its return to the user (eg, ActiGraph)
Numeric: data are returned in the form of numbers/figures or statistics
Graph: data are returned in the form of graphical representation
Written/textual feedback: data are returned in the form of textual feedback
‘Ometer (omnipresent meter): data are returned in the form of a growing or shrinking picture/image based on completion toward a predetermined goal (eg, UbiFit Garden)
Application: what operating system the mobile app can be accessed on for viewing the data/feedback
Software: the operating systems it can be accessed on if a piece of computing software is present for use at viewing the data
Website: can the data/feedback be viewed on a website?
Goal-setting capability: can predetermined goals be set by the user?
Customization: can the device or mobile app be customized by the end user (eg, via a software development kit)?
Each device was given a score between 1 and 6 for each attribute of behavior change. This score was used to describe two factors: (1) whether or not that device contained that behavior change attribute and (2) to what extent it did or did not contain the attribute. The self-monitoring scoring system that was used for each attribute was (1) yes; (2) yes, difficulties (eg, proximity to computer); (3) yes, lack of evidence to suggest this; (4) no, but present in future iterations; (5) no, but possible (with application programming interface or software development kit); and (6) not described/featured.
This scoring system was meant to be a descriptive tally of the behavior change attributes and not a judgment on the effectiveness of the various features.
Database searches identified 49,956 articles (
Pedometer studies: these were excluded if no evidence could be found that the pedometer in question provided temporally stamped data.
Prototypes: if the device was not commercially available or if no data currently existed for the prototype and only proof of concept information was available were excluded.
Health outcome: articles were excluded if they examined the relationship between behavior (eg, sedentary behavior and/or PA) and a particular health outcome (eg, blood pressure, lipid profile) and the measurement tool of choice was not the main focus of the article.
Miscellaneous: articles were excluded if the purpose of the study was to examine a new algorithm or data processing procedure for device analysis.
The remaining 125 studies (on 46 devices) and 90 websites yielded 146 devices (see
The remaining 82 [
Study/website selection.
Technologies found that can be used to self-monitor and provide feedback on PA ordered by number of self-monitoring attributes that were found to be present in the technologies.
The proportion of devices that could be used to self-monitor and provided feedback on physical activity with specific self-monitoring attributes.
Technologies found that could be used to self-monitor and provided feedback on sedentary time ordered by number of feedback elements in the technologies.
The proportion of devices that could be used to self-monitor and provided feedback on sedentary time with specific self-monitoring attributes.
The present systematic review sought to identify current measurement technologies available that could be used for real-time self-monitoring of sedentary time and/or PA. The review identified 125 articles on 46 device and 90 websites, for a combined total of 146 technologies that monitor sedentary time and/or PA. Of these, 82 devices were considered capable of self-monitoring sedentary time and/or PA. These devices can be used by researchers, clinicians, and the general public.
Technologies that self-monitor PA mainly come from the consumer health and fitness market. In general, these devices consist of an accelerometer for activity measurement (steps, calories burned, distance traveled) with varying secondary sensors, including gyroscope, inclinometer, lux sensors, skin sweat sensors, and other sensors that provide additional pieces of information. However, these devices will provide feedback only on PA and increases in PA do not automatically lead to decreases in sedentary time [
There are devices from both the commercial and research sectors that self-monitor sedentary behavior. These devices tend to measure sedentary time in two different ways. Firstly, posture sensors measure sedentary time either through an accelerometer in conjunction with gravitational components and proprietary algorithms (eg, activPAL) or through the alignment of the area of the body surrounding the pelvic area (ie, pelvic alignment is different depending on standing, sitting, and lying). The other way technologies tend to measure sedentary time is via pressure sensors. These pressure sensors are either located in a sock, shoe, or chair. When placed in a sock or shoe, the pressure can determine standing when there is pressure on the sensor and when there is less pressure the wearer is sitting or lying. Located on a chair, there is a simple binary outcome: when the pressure sensor is active the user is sitting and when it is inactive there is no sitting behavior at that site.
Both these types of devices usually provide feedback either via vibratory feedback (eg, Jawbone UP) or via an omnipresent display on the device (eg, Garmin Vivofit). These devices tend to, but not exclusively, connect to a mobile app for feedback on the PA and sedentary time. For PA, this usually takes the form of energy expenditure or proprietary company points (eg, Nike Fuel). For sedentary time, this usually takes the form of time spent sitting (eg, LumoBack) These mobile apps allow the wearer to receive real-time continuous feedback along with goal-setting capabilities and customization of type and timing of feedback; this is an aspect not traditionally offered by research devices.
With the plethora of devices now available (see
Self-monitoring technologies need to provide real-time feedback on aspects of PA and sitting that are personalized and relevant to the individual (ie, the attributes of real-time feedback must resonate with the individual and not be simply information that has been presupposed for them). Additionally, the immediate feedback should be of a low cognitive load so that it can resonate immediately with the end user [
Not every user will have the same needs and the presentation of actionable information will need to be tailored to fit individual needs. In addition, simply providing more medical data to patients not only fails to guarantee improved outcomes, but also could potentially lead to negative consequences [
In order for the promise of wearable technology to be fully realized, consumers, providers, and health care systems must be able to trust the reliability, privacy, and security of their data as well as the devices that collect and share it. Although regulatory oversight is often considered to be an impediment to the rapid propagation of innovative technologies, the existence of potential scams that could harm the end user dictates the need for some level of oversight. Globally, there is a great deal of uncertainty around wearable technology regulation; there are numerous countries that have no regulatory framework, whereas the other countries that do have a framework are still in their infancy and being actively refined [
Wearable technology users are also concerned about the privacy and ownership of their health data. In the era of big data, it is critical that the terms of ownership of personal data, most especially medical data, be unmistakably stated—not buried in the commonly unread and then accepted terms of use agreements—with users required to explicitly consent whenever their data are sold or transmitted to others [
One of the benefits of mHealth is easier accessibility to pertinent health care data, but this increased availability to both consumers and providers creates the potential for substantial security risks. Because of the small size of the device, it becomes easier to inadvertently lose or easier to steal, which may mean that the information stored on the device becomes accessible to others.
As consumer demand for wearable sensor increases, health care providers will face the possibility of being inundated by a flood of patient data. This will create a number of difficult challenges, including the potential requirement for 24/7 oversight, the need to summarize multiparameter, continuously collected data into a usable and clinically meaningful format [
The strengths of this review are the systematic approach taken and the comprehensive range of technologies found. However, there are some limitations. Due to the nature of articles included, it was not possible to present data on the validity and reliability of the devices in their ability to measure sedentary time. Similarly, due to the fact that self-monitoring using objective measurement tools is in its infancy, there are gaps in the literature as to whether these devices truly work as self-monitors; consequently, we cannot comment on how useful or valid they are in these settings. However, validity data are important. Users of self-monitoring technologies must be able to trust in the feedback that is being returned to them otherwise they may become disenfranchised with the tool and the behavior change tool. Therefore, incorporating important valid data with the feedback tools means additional value can be added to the consumers and potentially more potent behavior change.
In conclusion, the authors believe that this review is the first of its kind to systematically describe the wide breadth of devices that self-monitor and provide feedback on PA and sedentary behavior. There has been an explosion in the number of devices that measure PA and there is a greater need for the development of tools that specifically measure sitting time. Cooperative work between engineers, computer scientists, and academics in relevant fields is needed to develop these technologies that provide real-time, personalized, context-aware feedback to aid in the reduction in sitting time and its detrimental effect on cardiometabolic health independent of PA. This could potentially lead to the use of these devices in a health care setting as part of the increasing value-based care systems that are starting to arise in the United States or as a diagnostic tool, which is beginning to be implemented in the National Health Service in the United Kingdom.
This scoping review provides a record of a plethora of devices with information on their capabilities both in terms of their ability to measure behavior and to provide feedback to the user, providing a foundation for clinical, research, and public health use. Future studies are needed to further investigate the validity of these devices and their feasibility in increasing PA and/or decreasing sedentary time and the public health impact this may produce.
Example of devices discussed in this review. Clockwise from top left: Fitbit Charge HR, Garmin Vivofit, Jawbone Up, Nike Fuelband SE, Lumo Back Posture Sensor, and a mobile app.
Search Strategy.
Supplementary Table 1 - All Devices.
Supplementary Table 2 - Devices that self monitor physical activity.
Supplementary Table 3 - Devices that Self-monitor Sedentary behaviour.
physical activity
moderate-to-vigorous physical activity
TY and CLE acknowledge support from the National Institute for Health Research Collaboration for Leadership in Applied Health Research and Care-East Midlands (NIHR CLAHRC-EM) and the Leicester Clinical Trials Unit, as well as the NIHR Leicester-Loughborough Diet, Lifestyle and Physical Activity Biomedical Research Unit, which is a partnership between University Hospitals of Leicester NHS Trust, Loughborough University, and the University of Leicester. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR, or the Department of Health.
This work was supported in part by a Medical Research Council grant in LLHW Promoting Physical Activity in Older Age 2012 (MRC Reference: MR/K025090/1). This work was additionally supported by an EPSRC Small Equipment Grant titled: Upgrading the Small Equipment base for Early Career Researchers in the Engineering and Physical Sciences (EP/K03118X/1).
None declared.