Can smartphone-based motion sensors and applications measure and influence physical activity? (PEDro synthesis)

By Baskaran Chandrasekaran1, Ashokan Arumugam2  

1Centre for Sports Science, Medicine and Research, School of Allied Health Sciences, Manipal Academy of Higher Education, Karnataka, India
2Department of Community Medicine and Rehabilitation – Physiotherapy Section, Umeå University, SE-901 87, Umeå, Sweden


PEDro synthesis on the following paper: Bort-Roig J, Gilson ND, Puig-Ribera A, et al. Measuring and influencing physical activity with smartphone technology: a systematic review. Sports Med2014;44:671-86.

Regular physical activity (PA) reduces the risk of various chronic diseases such as cardiovascular disorders, hypertension, osteoporosis, depression.Now-a-days a smartphone (SmPh)serves as a promising tool to measure and influence an individual’s PA levels owing to the availability of robust applications (apps) harbouring the features of sophisticated motion sensors.The feasibility and utility of SmPhs to measure and influence PA in healthy and diseased individuals is well documented.3 4  However, Smph-based interventions to promote PA must follow best evidence-informed strategies and guidelines to improve health effects, patient perceptions and compliance with such interventions.


This systematic review was aimed at synthesising evidence on the use of SmPh technology as a tool to measureand influence PA levels of healthy and clinical populations.

Search and inclusion criteria

The literature search was conducted from 2007 to September 2013 in the following databases: Web of Science, MEDLINE, Scopus, PsychINFO, PubMed, ScienceDirect and EBSCO (CINAHL and SPORTDiscus). Keywords used were (physical activity OR exercise OR fitness) AND (smartphone* OR mobile phone* OR cell phone*) AND (intervention OR measurement). The review included full textarticles published in English which reported objective measurement of PA through in-built or external sensors and/or promoted PA via SmPh apps. The studies involving SmPh-based text messages and questionnaires measuring PA, abstracts of conference proceedings, theses, and books were excluded.


Studies evaluating accuracy and/or intervention effects of SmPh-based motion sensors in PA measurement were included. In the intervention studies, the SmPh-based apps integrated with motion sensors (in-built, external, or a combination) were used to set goals, send tailored reminders, and provide feedback in real-time or at later time points on, but not limited to, step counts, duration, intensity, progression, and variation of PA levels during the intervention period. Some intervention studies coupled the SmPh app to a website in order to check the progress of users and provide feedback, share PA levels to peers involved in the intervention, or as a source of reference for experts in the field. Neither risk of bias assessment nor meta-analysis to pool the results of the included studies was done.

Main outcome measures

The review assimilated and summarized a multitude of outcome measures such as accuracy of PA measurement, PA and health-related outcomes, participants’ perceptions and compliance with the usage of the SmPh-based sensors or apps.


Totally 26 studies were included out of 196 full-text articles reviewed. All studies were conducted in economically developed countries with the majority of the studies from the USA (n = 7), Finland (n =3), and Germany (n = 3). The included studies were categorized under three themes: 1.) studies measuring accuracy of smartphone apps coupled with internal or external sensors in measuring PA (n = 9; 107 participants [2 cohorts undergoing cardiac rehabilitation and 10 elderly cohorts]), 2.) studies concurrently measuring accuracy and influence of SmPh apps/sensors on PA during an intervention (n = 1; 9 healthy adults), and 3.) studies that intervened but did not validate the accuracy of smartphone apps/sensors in recording PA (n = 16; 604 participants [250 asymptomatic men and/or women; 8 adolescent girls], 278 obese/overweight, 17 with chronic obstructive pulmonary disease (COPD), 12 with type 2 diabetes, 24 with metabolic syndrome, 17 undergoing cardiac rehabilitation, and 8 office workers). In those studies measuring accuracy and/or intervention effects, PA was assessed with various machine learning algorithms using accelerometer, gyroscope or a similar signal detector as an in-built SmPh sensor (n = 12) and/or an external device (n = 14) with PA data uploaded via Bluetooth, wireless communication, or entered manually into the apps. Most intervention studies were based on one or more theoretical frameworks (cognitive behavioural theory, trans-theoretical model, social cognitive theory, etc.) while some studies did not report the theoretical basis underpinning the interventions.

The measurement accuracy of SmPh sensors/apps in measuring PA was 52–100%. The measurement accuracy might have varied between studies owing to inherent variations in the precision and calibration of the motion sensors, the position of SmPhs or external sensors worn on the body (the waist-hip area, chest or arm) or carried in a bag or the hand, and algorithms to recognize postural variations and PA.

The intervention studies mostly employed pre-post designs in a single group or two/three groups (experimental vs. comparator/control). Of the 17 intervention studies, seven demonstrated a significant change in bodyweight, cardiovascular risk factors and quality-of-life of the participants. Only five studies objectively reported SmPh-based intervention effects with four of them accounting for a mean increase (800–1104 steps/day, 2 weeks-6 months of intervention) or maintenance of PA levels (>10,000 steps/day over 3 months). Conversely, one pilot studyreported improved step count in one group employing SmPh app to monitor PA (+609 steps [mean]) and reduced step count (-1,017 steps) in the comparator group using SmPh app to monitor and influence PA. These differences might be attributed to low sample sizes and differences in baseline characteristics between groups with COPD.PA profiles, feedback (real-time), social support, seeking expert opinion, and goal setting were the cardinal factorsthat promoted engagement in PA. Disturbing or audible prompts, text messages, and competitive strategies were found to reduce engagement in PA.

The review also warrants further investigations combining different motion-sensing algorithms under free-living conditions and employing in-built SmPh sensors rather than SmPh-coupled external devices for increasing measurement accuracy and influencing PA. Further studies in this area in low- and middle-income countries are warranted.

Limitations/ considerations 

This review seems to be a “scoping review”, as such, rather than a “systematic review” because risk of bias of the included studies was not assessed. Furthermore, there is a need to synthesise the quality of the evidence using standardised recommendations such as the GRADE (Grading of Recommendations Assessment, Development and Evaluation). A lack of meta-analysis on the effects of SmPh-based interventions further ignores the pooled treatment effect and precludes firm conclusions. Overall, AMSTAR 2 score for the review is critically low which implies that the review findings must be interpreted with caution owing to one or more critical methodological flaws.

The study emphasized the need for RCTs on SmPh-based interventions to measure and influence PA in general and clinical populations. However, at present, there is growing evidence on the use of  SmPhs as a tool for  measuring and regulating PA in healthy and diseased populations within the last four years after the review was published.3 4 6-10 A summary of other literature reviews on techniques of PA measurement including SmPh-based motion senors is presented in Table 1.

Table 1. A summary of recent literature reviews on smartphone-based technology on health-realted outcomes

Authors (Year) Study purpose No of studies Key findings relevant to

SmPh-based sensors/apps

Cornet & Holden (2018)3* To review evidence on SmPh-based passive sensing (capture of data without extra effort from participants) for health and wellbeing 35 studies SmPh-based sensors/apps generally appear beneficial for valid and/or reliable passive sensing of PA and other health-related measures, behavioural modification through feedback and improved accountability in users.
McCallum et al. (2018)8$ To review study designs and methods used to collect data, assess effectiveness, engagement and acceptability of SmPh-based sensors and apps measuring PA 111 studies 55% were RCTs; less than 1/3rdof the studies simultaneously investigated effectiveness, engagement and acceptability of SmPh-based sensors and apps; device-generated logs were used for assessing engagement and questionnaires and/or qualitative methods for assessing acceptability.
Dowd et al. (2018)4* To review existing reviews (narrative/systematic) to summarize the effectiveness of different tools/methods employed to measure PA. 63 review articles Overall, the methodological effectiveness of motion sensors is increasing and these sensors appear valid and reliable for PA measurements compared to subjective measures. However, Bort-Roig et al.’s review was the only systematic review on SmPh-based motion sensors/apps included.
Zhao et al. (2016)10# To review the effectiveness of SmPh-based apps in influencing health-related behaviour for a multitude of health conditions 23 studies 73% of studies found significant effects on targeted behavioural change measured by one or more outcome measures (increasing PA, alcohol addiction, coping sadness, reducing smoking, etc.) and 82% studies showed improved compliance (60-100%) with the apps used in the intervention groups. The most common behavioural modification technique used was self-monitoring.

 *Systematic reviews; $Inter-disciplinary scoping review; #comprehensive thematic literature review.

Clinical implications 

The study demonstrated average (52%) to excellent (100%) accuracy of SmPh-based motion sensors/apps in measuring PA (walking, running, stair negotiation, sedentary postures, etc.). The SmPh-based interventions were found to be feasible, well perceived compared to paper-based PA diaries, and the users were highly compliant with those interventions that provided feedback and allowed goal-setting. However, some intervention studies required manual entry of pedometer step counts into the SmPh apps. According to the review, evidence on the effectiveness of SmPh interventions on PA promotion is limited owing to a lack of randomized controlled trials, omission of risk of bias associated with the studies, small sample sizes (n < 20 in 7/17 studies), short intervention periods and only marginal increase in PA levels in certain studies. Interested readers are referred to the reviews summarized in Table1 and other relevant literature to identify the best possible evidence in this research area.

Competing Interests

None declared

Corresponding author:

Ashokan Arumugam:


  1. Warburton DER, Nicol CW, Bredin SSD. Health benefits of physical activity: the evidence. CMAJ 2006;174(6):801-09. doi: 10.1503/cmaj.051351
  2. del Rosario MB, Redmond SJ, Lovell NH. Tracking the Evolution of Smartphone Sensing for Monitoring Human Movement. Sensors (Basel)2015;15(8):18901-33. doi: 10.3390/s150818901
  3. Cornet VP, Holden RJ. Systematic review of smartphone-based passive sensing for health and wellbeing. J Biomed Inform2018;77:120-32. doi: 10.1016/j.jbi.2017.12.008
  4. Dowd KP, Szeklicki R, Minetto MA, et al. A systematic literature review of reviews on techniques for physical activity measurement in adults: a DEDIPAC study. Int J Behav Nutr Phys Act2018;15(1):15. doi: 10.1186/s12966-017-0636-2
  5. Nguyen HQ, Gill DP, Wolpin S, et al. Pilot study of a cell phone-based exercise persistence intervention post-rehabilitation for COPD. International Journal of Chronic Obstructive Pulmonary Disease2009;4:301-13.
  6. Coughlin SS, Whitehead M, Sheats JQ, et al. A Review of Smartphone Applications for Promoting Physical Activity. Jacobs J Community Med2016;2(1):021.
  7. Jee H. Review of researches on smartphone applications for physical activity promotion in healthy adults. J Exerc Rehabil2017;13(1):3-11. doi: 10.12965/jer.1732928.464
  8. McCallum C, Rooksby J, Gray CM. Evaluating the Impact of Physical Activity Apps and Wearables: Interdisciplinary Review. JMIR Mhealth Uhealth2018;6(3):e58. doi: 10.2196/mhealth.9054
  9. Schoeppe S, Alley S, Van Lippevelde W, et al. Efficacy of interventions that use apps to improve diet, physical activity and sedentary behaviour: a systematic review. Int J Behav Nutr Phys Act2016;13(1):127. doi: 10.1186/s12966-016-0454-y
  10. Zhao J, Freeman B, Li M. Can Mobile Phone Apps Influence People’s Health Behavior Change? An Evidence Review. J Med Internet Res2016;18(11):e287. doi: 10.2196/jmir.5692

(Visited 167 times, 1 visits today)