[Skip to Content]
[Skip to Content Landing]
Figure 1.
Flow Diagram of Screening Stages
Flow Diagram of Screening Stages

No unpublished studies were found. RCT indicates randomized clinical trial.

Figure 2.
Subgroup Meta-analysis of Delivery and Consultation Type
Subgroup Meta-analysis of Delivery and Consultation Type

NA indicates not applicable; SMD, standardized mean difference.

aFor studies using accelerometers, mean values reflect minutes of moderate-to-vigorous physical activity. For studies using pedometers, mean values reflect total daily steps.

Table 1.  
Meta-analysis of Accelerometer and Pedometer Interventions for Increased Physical Activity (Primary and Secondary Outcomes)
Meta-analysis of Accelerometer and Pedometer Interventions for Increased Physical Activity (Primary and Secondary Outcomes)
Table 2.  
Univariable and Multivariable Metaregressions for Physical Activity Outcomes
Univariable and Multivariable Metaregressions for Physical Activity Outcomes
1.
New York State. Office of Mental Health. Cardiometabolic risk: what is it and what can I do about it? https://www.omh.ny.gov/omhweb/psyckes_medicaid/brochures/cardio.html. Accessed October 2018.
2.
Kivimäki  M, Pentti  J, Ferrie  JE,  et al; IPD-Work Consortium.  Work stress and risk of death in men and women with and without cardiometabolic disease: a multicohort study.  Lancet Diabetes Endocrinol. 2018;6(9):705-713. doi:10.1016/S2213-8587(18)30140-2PubMedGoogle ScholarCrossref
3.
National Institute for Health and Care Excellence. Type 2 diabetes in adults: management (NICE guideline). https://www.nice.org.uk/guidance/ng28/resources/type-2-diabetes-in-adults-management-pdf-1837338615493. Published December 2, 2015. Accessed April 2019.
4.
Diabetes Times. Diabetes NHS costs could hit £17 billion. https://diabetestimes.co.uk/diabetes-nhs-costs-could-hit-17-billion/. Published October 25, 2016. Accessed January 2019.
5.
McNamara  E, Hudson  Z, Taylor  SJ.  Measuring activity levels of young people: the validity of pedometers.  Br Med Bull. 2010;95:121-137. doi:10.1093/bmb/ldq016PubMedGoogle ScholarCrossref
6.
Kvaavik  E, Batty  GD, Ursin  G, Huxley  R, Gale  CR.  Influence of individual and combined health behaviors on total and cause-specific mortality in men and women: the United Kingdom Health and Lifestyle Survey.  Arch Intern Med. 2010;170(8):711-718. doi:10.1001/archinternmed.2010.76PubMedGoogle ScholarCrossref
7.
Public Health England. Physical activity: applying All Our Health. https://www.gov.uk/government/publications/physical-activity-applying-all-our-health/physical-activity-applying-all-our-health. Updated January 9, 2018. Accessed September 10, 2018.
8.
National Institute for Health and Clinical Excellence. Physical activity: walking and cycling (Public Health Guidance PH41). https://www.nice.org.uk/guidance/ph41. Published November 2012. Accessed September 10, 2018.
9.
Centers for Disease Control and Prevention (CDC).  School health guidelines to promote healthy eating and physical activity.  MMWR Recomm Rep. 2011;60(RR-5):1-76.PubMedGoogle Scholar
10.
National Institute for Health and Clinical Excellence. Physical activity in the workplace (Public Health Guidance PH13). https://www.nice.org.uk/guidance/ph13. Published May 2008. Accessed August 30, 2017.
11.
de Souto Barreto  P.  Time to challenge public health guidelines on physical activity.  Sports Med. 2015;45(6):769-773. doi:10.1007/s40279-015-0326-7PubMedGoogle ScholarCrossref
12.
Afshin  A, Babalola  D, Mclean  M,  et al.  Information technology and lifestyle: a systematic evaluation of internet and mobile interventions for improving diet, physical activity, obesity, tobacco, and alcohol use.  J Am Heart Assoc. 2016;5(9):e003058. doi:10.1161/JAHA.115.003058PubMedGoogle ScholarCrossref
13.
Ayabe  M, Ishii  K, Takayama  K, Aoki  J, Tanaka  H.  Comparison of interdevice measurement difference of pedometers in younger and older adults.  Br J Sports Med. 2010;44(2):95-99. doi:10.1136/bjsm.2007.045179PubMedGoogle ScholarCrossref
14.
Bravata  DM, Smith-Spangler  C, Sundaram  V,  et al.  Using pedometers to increase physical activity and improve health: a systematic review.  JAMA. 2007;298(19):2296-2304. doi:10.1001/jama.298.19.2296PubMedGoogle ScholarCrossref
15.
Qiu  S, Cai  X, Chen  X, Yang  B, Sun  Z.  Step counter use in type 2 diabetes: a meta-analysis of randomized controlled trials.  BMC Med. 2014;12:36. doi:10.1186/1741-7015-12-36PubMedGoogle ScholarCrossref
16.
Public Health England. Adult obesity and type 2 diabetes. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/338934/Adult_obesity_and_type_2_diabetes_.pdf. Published July 2014. Accessed February 5, 2019.
17.
Higgins JPT, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0. London, UK: Cochrane Collaboration; 2011.
18.
Liberati  A, Altman  DG, Tetzlaff  J,  et al.  The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration.  BMJ. 2009;339:b2700. doi:10.1136/bmj.b2700PubMedGoogle ScholarCrossref
19.
Brickwood  K-J, Watson  G, O’Brien  J, Williams  AD.  Consumer-based wearable activity trackers increase physical activity participation: systematic review and meta-analysis.  JMIR Mhealth Uhealth. 2019;7(4):e11819-e11819. doi:10.2196/11819PubMedGoogle ScholarCrossref
20.
Department of Health and Human Services. 2018 Physical Activity Guidelines Advisory Committee Scientific Report. https://health.gov/paguidelines/second-edition/report/pdf/PAG_Advisory_Committee_Report.pdf. Accessed August 8, 2019.
21.
Montgomery  P, Underhill  K, Gardner  F, Operario  D, Mayo-Wilson  E.  The Oxford Implementation Index: a new tool for incorporating implementation data into systematic reviews and meta-analyses.  J Clin Epidemiol. 2013;66(8):874-882. doi:10.1016/j.jclinepi.2013.03.006PubMedGoogle ScholarCrossref
22.
Higgins  JPT, Altman  DG, Gøtzsche  PC,  et al; Cochrane Bias Methods Group; Cochrane Statistical Methods Group.  The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials.  BMJ. 2011;343:d5928. doi:10.1136/bmj.d5928PubMedGoogle ScholarCrossref
23.
R [computer program]. Version 3.3.3. Vienna, Austria: R Project for Statistical Computing.
24.
Hedges  LV.  Distribution theory for Glass’s estimator of effect size and related estimators.  J Educ Behav Stat. 1981;6(2):107-128. doi:10.3102/10769986006002107Google ScholarCrossref
25.
Cohen J.  Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Oxfordshire, United Kingdom: Taylor & Francis; 1977.
26.
Higgins  JP, Thompson  SG.  Quantifying heterogeneity in a meta-analysis.  Stat Med. 2002;21(11):1539-1558. doi:10.1002/sim.1186PubMedGoogle ScholarCrossref
27.
IntHout  J, Ioannidis  JP, Borm  GF.  The Hartung-Knapp-Sidik-Jonkman method for random effects meta-analysis is straightforward and considerably outperforms the standard DerSimonian-Laird method.  BMC Med Res Methodol. 2014;14(1):25. doi:10.1186/1471-2288-14-25PubMedGoogle ScholarCrossref
28.
Cayir  Y, Aslan  SM, Akturk  Z.  The effect of pedometer use on physical activity and body weight in obese women.  Eur J Sport Sci. 2015;15(4):351-356. doi:10.1080/17461391.2014.940558PubMedGoogle ScholarCrossref
29.
Greaney  ML, Askew  S, Wallington  SF, Foley  PB, Quintiliani  LM, Bennett  GG.  The effect of a weight gain prevention intervention on moderate-vigorous physical activity among black women: the Shape Program.  Int J Behav Nutr Phys Act. 2017;14(1):139. doi:10.1186/s12966-017-0596-6PubMedGoogle ScholarCrossref
30.
Holliday  A, Burgin  A, Fernandez  EV, Fenton  SAM, Thielecke  F, Blannin  AK.  Points-based physical activity: a novel approach to facilitate changes in body composition in inactive women with overweight and obesity.  BMC Public Health. 2018;18(1):261. doi:10.1186/s12889-018-5125-2PubMedGoogle ScholarCrossref
31.
Pekmezi  D, Ainsworth  C, Joseph  RP,  et al.  Pilot trial of a home-based physical activity program for African American women.  Med Sci Sports Exerc. 2017;49(12):2528-2536. doi:10.1249/MSS.0000000000001370PubMedGoogle ScholarCrossref
32.
Christle  JW, Schlumberger  A, Haller  B, Gloeckl  R, Halle  M, Pressler  A.  Individualized vs. group exercise in improving quality of life and physical activity in patients with cardiac disease and low exercise capacity: results from the DOPPELHERZ trial.  Disabil Rehabil. 2017;39(25):2566-2571. doi:10.1080/09638288.2016.1242174PubMedGoogle ScholarCrossref
33.
Frederix  I, Hansen  D, Coninx  K,  et al.  Medium-term effectiveness of a comprehensive internet-based and patient-specific telerehabilitation program with text messaging support for cardiac patients: randomized controlled trial.  J Med Internet Res. 2015;17(7):e185. doi:10.2196/jmir.4799PubMedGoogle ScholarCrossref
34.
Engel  L, Lindner  H.  Impact of using a pedometer on time spent walking in older adults with type 2 diabetes.  Diabetes Educ. 2006;32(1):98-107. doi:10.1177/0145721705284373PubMedGoogle ScholarCrossref
35.
Van Dyck  D, De Greef  K, Deforche  B,  et al.  The relationship between changes in steps/day and health outcomes after a pedometer-based physical activity intervention with telephone support in type 2 diabetes patients.  Health Educ Res. 2013;28(3):539-545. doi:10.1093/her/cyt038PubMedGoogle ScholarCrossref
36.
Yates  T, Edwardson  CL, Henson  J,  et al.  Walking Away From Type 2 Diabetes: a cluster randomized controlled trial.  Diabet Med. 2017;34(5):698-707. doi:10.1111/dme.13254PubMedGoogle ScholarCrossref
37.
Fayehun  AF, Olowookere  OO, Ogunbode  AM, Adetunji  AA, Esan  A.  Walking prescription of 10 000 steps per day in patients with type 2 diabetes mellitus: a randomised trial in Nigerian general practice.  Br J Gen Pract. 2018;68(667):e139-e145. doi:10.3399/bjgp18X694613PubMedGoogle ScholarCrossref
38.
De Greef  K, Deforche  B, Tudor-Locke  C, De Bourdeaudhuij  I.  A cognitive-behavioural pedometer-based group intervention on physical activity and sedentary behaviour in individuals with type 2 diabetes.  Health Educ Res. 2010;25(5):724-736. doi:10.1093/her/cyq017PubMedGoogle ScholarCrossref
39.
Unick  JL, Gaussoin  SA, Hill  JO,  et al.  Four-year physical activity levels among intervention participants with type 2 diabetes.  Med Sci Sports Exerc. 2016;48(12):2437-2445. doi:10.1249/MSS.0000000000001054PubMedGoogle ScholarCrossref
40.
Unick  JL, Gaussoin  SA, Hill  JO,  et al; Look AHEAD Research Group. Objectively assessed physical activity and weight loss maintenance among individuals enrolled in a lifestyle intervention.  Obesity (Silver Spring). 2017;25(11):1903-1909. doi:10.1002/oby.21971PubMedGoogle ScholarCrossref
41.
Andersen  E, Ekelund  U, Anderssen  SA.  Effects of reducing sedentary time on glucose metabolism in immigrant Pakistani men.  Med Sci Sports Exerc. 2015;47(4):775-781. doi:10.1249/MSS.0000000000000460PubMedGoogle ScholarCrossref
42.
Araiza  P, Hewes  H, Gashetewa  C, Vella  CA, Burge  MR.  Efficacy of a pedometer-based physical activity program on parameters of diabetes control in type 2 diabetes mellitus.  Metabolism. 2006;55(10):1382-1387. doi:10.1016/j.metabol.2006.06.009PubMedGoogle ScholarCrossref
43.
Baillot  A, Vallée  CA, Mampuya  WM,  et al.  Effects of a pre-surgery supervised exercise training 1 year after bariatric surgery: a randomized controlled study.  Obes Surg. 2018;28(4):955-962. doi:10.1007/s11695-017-2943-8PubMedGoogle ScholarCrossref
44.
Borland  M, Rosenkvist  A, Cider  A.  A group-based exercise program did not improve physical activity in patients with chronic heart failure and comorbidity: a randomized controlled trial.  J Rehabil Med. 2014;46(5):461-467. doi:10.2340/16501977-1794PubMedGoogle ScholarCrossref
45.
Cowie  A, Thow  MK, Granat  MH, Mitchell  SL.  A comparison of home and hospital-based exercise training in heart failure: immediate and long-term effects upon physical activity level.  Eur J Cardiovasc Prev Rehabil. 2011;18(2):158-166. doi:10.1177/1741826710389389PubMedGoogle ScholarCrossref
46.
Cupples  M, Dean  A, Tully  MA,  et al.  Using pedometer step-count goals to promote physical activity in cardiac rehabilitation: a feasibility study of a controlled trial.  Int J Phys Med Rehabil. 2013;1(7):157. doi:10.4172/2329-9096.1000157Google Scholar
47.
Dasgupta  K, Rosenberg  E, Joseph  L,  et al; SMARTER Trial Group.  Physician step prescription and monitoring to improve ARTERial health (SMARTER): A randomized controlled trial in patients with type 2 diabetes and hypertension.  Diabetes Obes Metab. 2017;19(5):695-704. doi:10.1111/dom.12874PubMedGoogle ScholarCrossref
48.
De Greef  K, Deforche  B, Tudor-Locke  C, De Bourdeaudhuij  I.  Increasing physical activity in Belgian type 2 diabetes patients: a three-arm randomized controlled trial.  Int J Behav Med. 2011;18(3):188-198. doi:10.1007/s12529-010-9124-7PubMedGoogle ScholarCrossref
49.
De Greef  KP, Deforche  BI, Ruige  JB,  et al.  The effects of a pedometer-based behavioral modification program with telephone support on physical activity and sedentary behavior in type 2 diabetes patients.  Patient Educ Couns. 2011;84(2):275-279. doi:10.1016/j.pec.2010.07.010PubMedGoogle ScholarCrossref
50.
Devi  R, Powell  J, Singh  S.  A web-based program improves physical activity outcomes in a primary care angina population: randomized controlled trial.  J Med Internet Res. 2014;16(9):e186. doi:10.2196/jmir.3340PubMedGoogle ScholarCrossref
51.
Frederix  I, Van Driessche  N, Hansen  D,  et al.  Increasing the medium-term clinical benefits of hospital-based cardiac rehabilitation by physical activity telemonitoring in coronary artery disease patients.  Eur J Prev Cardiol. 2015;22(2):150-158. doi:10.1177/2047487313514018PubMedGoogle ScholarCrossref
52.
Guiraud  T, Granger  R, Gremeaux  V,  et al.  Telephone support oriented by accelerometric measurements enhances adherence to physical activity recommendations in noncompliant patients after a cardiac rehabilitation program.  Arch Phys Med Rehabil. 2012;93(12):2141-2147. doi:10.1016/j.apmr.2012.06.027PubMedGoogle ScholarCrossref
53.
Houle  J, Doyon  O, Vadeboncoeur  N, Turbide  G, Diaz  A, Poirier  P.  Innovative program to increase physical activity following an acute coronary syndrome: randomized controlled trial.  Patient Educ Couns. 2011;85(3):e237-e244. doi:10.1016/j.pec.2011.03.018PubMedGoogle ScholarCrossref
54.
Houle  J, Doyon  O, Vadeboncoeur  N, Turbide  G, Diaz  A, Poirier  P.  Effectiveness of a pedometer-based program using a socio-cognitive intervention on physical activity and quality of life in a setting of cardiac rehabilitation.  Can J Cardiol. 2012;28(1):27-32. doi:10.1016/j.cjca.2011.09.020PubMedGoogle ScholarCrossref
55.
Izawa  KP, Watanabe  S, Hiraki  K,  et al.  Determination of the effectiveness of accelerometer use in the promotion of physical activity in cardiac patients: a randomized controlled trial.  Arch Phys Med Rehabil. 2012;93(11):1896-1902. doi:10.1016/j.apmr.2012.06.015PubMedGoogle ScholarCrossref
56.
Katzmarzyk  PT, Champagne  CM, Tudor-Locke  C,  et al.  A short-term physical activity randomized trial in the lower Mississippi Delta.  PLoS One. 2011;6(10):e26667. doi:10.1371/journal.pone.0026667PubMedGoogle ScholarCrossref
57.
Kirk  A, Barnett  J, Leese  G, Mutrie  N.  A randomized trial investigating the 12-month changes in physical activity and health outcomes following a physical activity consultation delivered by a person or in written form in Type 2 diabetes: Time2Act.  Diabet Med. 2009;26(3):293-301. doi:10.1111/j.1464-5491.2009.02675.xPubMedGoogle ScholarCrossref
58.
Miyamoto  T, Fukuda  K, Oshima  Y, Moritani  T.  Non-locomotive physical activity intervention using a tri-axial accelerometer reduces sedentary time in type 2 diabetes.  Phys Sportsmed. 2017;45(3):245-251. doi:10.1080/00913847.2017.1350084PubMedGoogle ScholarCrossref
59.
Paschali  AA, Goodrick  GK, Kalantzi-Azizi  A, Papadatou  D, Balasubramanyam  A.  Accelerometer feedback to promote physical activity in adults with type 2 diabetes: a pilot study.  Percept Mot Skills. 2005;100(1):61-68. doi:10.2466/pms.100.1.61-68PubMedGoogle ScholarCrossref
60.
Ribeiro  F, Oliveira  NL, Silva  G,  et al.  Exercise-based cardiac rehabilitation increases daily physical activity of patients following myocardial infarction: subanalysis of two randomised controlled trials.  Physiotherapy. 2017;103(1):59-65. doi:10.1016/j.physio.2015.12.002PubMedGoogle ScholarCrossref
61.
Tudor-Locke  C, Bell  RC, Myers  AM,  et al.  Controlled outcome evaluation of the First Step Program: a daily physical activity intervention for individuals with type II diabetes.  Int J Obes Relat Metab Disord. 2004;28(1):113-119. doi:10.1038/sj.ijo.0802485PubMedGoogle ScholarCrossref
62.
Young  L, Hertzog  M, Barnason  S.  Effects of a home-based activation intervention on self-management adherence and readmission in rural heart failure patients: the PATCH randomized controlled trial.  BMC Cardiovasc Disord. 2016;16(1):176-176. doi:10.1186/s12872-016-0339-7PubMedGoogle ScholarCrossref
63.
Bjørgaas  MR, Vik  JT, Stølen  T, Lydersen  S, Grill  V.  Regular use of pedometer does not enhance beneficial outcomes in a physical activity intervention study in type 2 diabetes mellitus.  Metabolism. 2008;57(5):605-611. doi:10.1016/j.metabol.2007.12.002PubMedGoogle ScholarCrossref
64.
Baskerville  R, Ricci-Cabello  I, Roberts  N, Farmer  A.  Impact of accelerometer and pedometer use on physical activity and glycaemic control in people with type 2 diabetes: a systematic review and meta-analysis.  Diabet Med. 2017;34(5):612-620. doi:10.1111/dme.13331PubMedGoogle ScholarCrossref
65.
Riley  RD, Thompson  JR, Abrams  KR.  An alternative model for bivariate random-effects meta-analysis when the within-study correlations are unknown.  Biostatistics. 2008;9(1):172-186. doi:10.1093/biostatistics/kxm023PubMedGoogle ScholarCrossref
66.
Kelly  MP, Barker  M.  Why is changing health-related behaviour so difficult?  Public Health. 2016;136:109-116. doi:10.1016/j.puhe.2016.03.030PubMedGoogle ScholarCrossref
67.
Tudor-Locke  C, Craig  CL, Brown  WJ,  et al.  How many steps/day are enough? for adults.  Int J Behav Nutr Phys Act. 2011;8:79-79. doi:10.1186/1479-5868-8-79PubMedGoogle ScholarCrossref
68.
Physical Activity Guidelines Advisory Committee.  Physical Activity Guidelines Advisory Committee Report, 2008. Washington, DC: US Department of Health and Human Services; 2008.
69.
O’Donovan  G, Blazevich  AJ, Boreham  C,  et al.  The ABC of physical activity for health: a consensus statement from the British Association of Sport and Exercise Sciences.  J Sports Sci. 2010;28(6):573-591. doi:10.1080/02640411003671212PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Views 3,401
    Original Investigation
    Public Health
    October 9, 2019

    Accelerometer- and Pedometer-Based Physical Activity Interventions Among Adults With Cardiometabolic Conditions: A Systematic Review and Meta-analysis

    Author Affiliations
    • 1National Institute for Health Research School for Primary Care Research, Manchester Academic Health Science Centre, University of Manchester, United Kingdom
    • 2Department of Primary Care and Public Health, Brighton and Sussex Medical School, University of Brighton, United Kingdom
    JAMA Netw Open. 2019;2(10):e1912895. doi:10.1001/jamanetworkopen.2019.12895
    Key Points español 中文 (chinese)

    Question  Is the use of accelerometers and/or pedometers associated with increased physical activity in people with cardiometabolic conditions?

    Findings  This systematic review and meta-analysis of 32 randomized clinical trials (4856 participants) found that accelerometer- and pedometer-based interventions were associated with small to medium short-term improvements in physical activity and that consultations with health professionals and pedometers were associated with improved physical activity.

    Meaning  Accelerometer- and pedometer-based interventions for people with cardiometabolic conditions have demonstrated encouraging results, but levels of physical activity remain below the targets set by clinical recommendations.

    Abstract

    Importance  Accelerometers and pedometers are accessible technologies that could have a role in encouraging physical activity (PA) in line with current recommendations. However, there is no solid evidence of their association with PA in participants with 1 or more cardiometabolic conditions such as diabetes, prediabetes, obesity, and cardiovascular disease.

    Objectives  To assess the association of accelerometer- and pedometer-based interventions with increased activity and other improved health outcomes in adults with cardiometabolic conditions and to examine characteristics of the studies that could influence the association of both interventions in improving PA.

    Data Sources  Records from MEDLINE, Embase, the Cochrane Central Register of Controlled Trials, the Cumulative Index to Nursing and Allied Health, and PsycINFO were searched from inception until August 2018 with no language restriction.

    Study Selection  Randomized clinical trials or cluster randomized clinical trials evaluating the use of wearable technology devices such as pedometers and accelerometers as motivating and monitoring tools for increasing PA were included. After removing duplicates, the searches retrieved 5762 references. Following abstract and title screening of 1439 references and full-text screening of 107 studies, 36 studies met inclusion criteria.

    Data Extraction and Synthesis  Mean difference in PA was assessed by random-effects meta-analysis. Where the scale was different across studies, the standardized mean difference was used instead. Heterogeneity was quantified using the I2 statistic and explored using mixed-effects metaregression. This study was registered with PROSPERO and followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guideline.

    Main Outcomes and Measures  The primary outcome was objectively measured PA in the short to medium term (postintervention to 8 months’ follow-up).

    Results  Thirty-six randomized clinical trials (20 using accelerometers and 16 using pedometers) involving 5208 participants were eligible for review. Meta-analysis involving 32 of these trials (4856 participants) showed medium improvements in PA: accelerometers and pedometers combined vs comparator showed a small significant increase in PA overall (standardized mean difference, 0.39 [95% CI, 0.28-0.51]; I2 = 60% [95% CI, 41%-73%]) in studies of short to medium follow-up over a mean (SD) of 32 (28.6) weeks. Multivariable metaregression showed improved association with PA for complex interventions that involved face-to-face consultation sessions with facilitators (β = 0.36; 95% CI, 0.17-0.55; P < .001) and pedometer-based interventions (β = 0.30; 95% CI, 0.08-0.52; P = .002).

    Conclusions and Relevance  In this study, complex accelerometer- and pedometer-based interventions led to significant small to medium improvements in PA levels of people with cardiometabolic conditions. However, longer-term trials are needed to assess their performance over time. This study found no evidence that simple self-monitored interventions using either pedometers or accelerometers are associated with improvements in PA.

    Introduction

    A large proportion of the population experiences cardiometabolic conditions such as type 2 diabetes, prediabetes states (eg, obesity), and cardiovascular disease.1,2 In the United Kingdom, 3.3 million people have been diagnosed with type 2 diabetes and most of them experience additional cardiometabolic conditions or risks, including obesity, increased blood pressure, disturbed blood lipid levels, and a tendency to develop thrombosis and cardiovascular disease.3 The increasingly high prevalence of cardiometabolic conditions combined with demographic changes means that the costs of cardiometabolic disease will account for more than 20% of the entire UK National Health Service (NHS) budget in the next 20 years, and most of these costs are avoidable.4

    Despite their detrimental health and economic impacts, cardiometabolic conditions are mainly lifestyle related and can be improved by targeting unhealthy lifestyle behaviors. In particular, low physical activity (PA) is a fundamental modifiable risk behavior for people with cardiometabolic conditions and a major opportunity for intervention.5 Addressing very low levels of PA has the potential to prevent premature death more than any other risk factor,6 including smoking, alcohol use, or stress-related illness. Recognizing its importance, several public health guidelines recommend reaching and maintaining health-enhancing levels of PA7 and promote PA interventions in the community and the workplace.8-10 However, promoting PA in people with cardiometabolic conditions remains a challenge.11

    Objective monitoring devices may help people with cardiometabolic conditions improve their PA levels and health behaviors. These devices include simple monitoring devices such as pedometers (step count devices whose results can be recorded daily in a log book) or accelerometers (more technologically advanced devices containing time-based movement sensors, monitors of time and intensity of activity and inactivity, and monitors of heart rate and calories burned). These devices have recently become very popular for motivating, monitoring, and increasing PA in people with a range of chronic conditions, including those with cardiometabolic conditions.12 Both types of devices are simple, relatively inexpensive, user friendly, and potentially motivational.13

    The first systematic review published more than a decade ago pooled the results of 8 trials involving outpatient participants with mixed conditions (ie, diabetes, chronic obstructive pulmonary disease, sedentary lifestyle, hypertension). Their results showed that pedometer-based interventions had a promising association with PA levels.14 A more recent systematic review that exclusively focused on people with type 2 diabetes identified 12 trials and showed that the use of monitoring devices was associated with a medium short-term increase in PA.15 However, one major limitation in the current evidence base is that the large overlap among cardiometabolic conditions (eg, diabetes, obesity, and cardiac disease) has not been taken into consideration. For example, the prevalence of adults with type 2 diabetes associated with overweight or obese status is approximately 90% in the United Kingdom.16 Therefore, it is important to summarize the evidence across people who experience 1 or more of these cardiometabolic conditions. The advantage of this broader approach is that a more robust investigation of specific factors that are potentially responsible for the improved PA levels in the intervention groups is feasible (owing to the larger number of eligible trials). For example, there is limited evidence about the role of intervention components, delivery, and patient factors in increasing PA and improving health outcomes. There is also no evidence, to our knowledge, on whether any PA benefits are sustained long term. These are major barriers for the wider use of these monitoring devices in the care of people with cardiometabolic conditions.

    To our knowledge, the study described in this article is the most comprehensive systematic review with meta-analysis to date examining whether interventions using monitoring devices (pedometers and/or accelerometers) are associated with improvements in PA levels and health outcomes, including blood glucose levels, blood pressure, cholesterol levels, body weight, and body mass index (BMI) among people with cardiometabolic conditions. We also used metaregression to examine whether the increased PA levels in the intervention groups over comparators were moderated by the characteristics of interventions (type of monitoring device, setting daily goals, use of consultations with facilitators, evaluation length, use of a theoretical framework, and uptake rate) and patients (sex and index condition).

    Methods

    The review was conducted and reported in accordance with the Cochrane Handbook17 and the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guideline.18 The protocol was registered on PROSPERO (CRD42018104448).

    Search Methods

    Searches were performed in the Cochrane Central Register of Controlled Trials, Cumulative Index to Nursing and Allied Health, Embase, MEDLINE, and PsycINFO from inception until August 2018 with no language restriction. Search updates were performed in August 2019 and no further eligible studies were identified. We used combinations of Medical Subject Headings terms and text words in “diabetes,” “obesity,” “cardiovascular disease,” “pedometers,” “accelerometers” and “step-counters.” The full search strategy in MEDLINE is available in eTable 1 in the Supplement. Additional studies were obtained from screening the reference lists of included trials and previous systematic reviews. We also contacted experts in the field to inquire about unpublished studies. Trial registers (ClinicalTrials.gov, ISRCTN, the World Health Organization International Clinical Trials Registry Platform portal, and OpenTrials.net) were also searched to identify any unpublished or ongoing trials.

    Eligibility Criteria
    Population

    Eligible individuals included adults (aged ≥18 years) with a diagnosis of type 2 diabetes (or at risk for type 2 diabetes), obesity or overweight, and cardiovascular disease. For obesity classification, the World Health Organization definition was used to standardize across studies.19 We excluded studies of people diagnosed with stroke and studies of people immediately after surgery.

    Intervention

    Randomized clinical trials (RCTs) or cluster RCTs evaluating the use of monitoring devices such as pedometers and accelerometers as motivating and monitoring tools for increasing PA were included. We excluded trials that required participants to be hospitalized, trials in which assessors were not blinded to the wearable technology, and trials that used a wearable technology to measure the association of a pharmacological treatment with an individual’s ability to be physically active.

    Comparator

    Any comparator (eg, usual care, control, no intervention, or a minimal intervention with step counters used only for counting steps) was included. All comparators were treated the same in the analysis; however, a sensitivity analysis was performed for usual care groups only.

    Outcome

    The primary outcome was objectively measured PA levels in the short term (eg, postintervention to 6 months’ follow-up). Secondary outcomes were long-term levels of PA, self-reported PA, body weight (kilograms) or BMI (calculated as weight in kilograms divided by height in meters squared), blood glucose level (hemoglobin A1c [HbA1c] percentage), blood pressure (systolic or diastolic [millimeters of mercury]), and cholesterol levels (total, high-density lipoprotein, and low-density lipoprotein [milligrams per deciliter]). Studies were excluded if they only measured secondary outcomes.

    Data Collection and Extraction

    Titles and abstracts were assessed by 3 of us (A.H., M.P., and C.A.). Data extraction was conducted by 1 of us (A.H.) and checked by a second reviewer (C.A.) for consistency. A modified version of the Cochrane Public Health Group’s data extraction template20 was used after pilot testing on 5 studies to ensure reliability. We used the Oxford Implementation Index to assess implementation of the intervention and contextual factors.21 This was adapted for the purposes of this review.

    Assessment of Risk of Bias

    Risk of bias for each study was assessed by 2 of us (A.H. and C.A.) using the Cochrane Risk of Bias tool.22 The blinding of participants and personnel was not included in the risk of bias assessment, as many studies did not report this domain as it was not possible to blind participants while using the technology device. If further information was required on any aspect of study design or outcome, we sought related publications and trial protocols and contacted study authors. For cluster RCTs, the Cochrane handbook section 16.3.2 was consulted.

    Missing Data

    Study authors were contacted by email where there were missing or unclear data (eg, relating to the primary outcome). Studies for which insufficient primary data were available (eg, missing data cannot be obtained) were excluded from the meta-analysis but not the review.

    Data Synthesis and Analysis

    The statistical analysis proceeded in 2 stages. First, DerSimonian-Laird inverse variance weighting random-effects meta-analyses23 were conducted to determine the association of the interventions with improved primary and secondary outcomes compared with controls. If the control group varied considerably, we performed a sensitivity analysis for just the usual care group. For dichotomous outcomes, relative risks and their 95% confidence intervals were calculated, and for continuous outcomes, standardized mean differences (SMDs) were calculated using the Hedges g.24 The SMDs, or associations, were interpreted according to the Cohen rule of interpretation.25 Physical activity outcomes were separated by intervention measure, ie, objectively (daily step count for pedometers, moderate-to-vigorous PA [MVPA] and total PA for accelerometers) or self-reported. For studies that included both interventions (ie, both steps and MVPA as outcome measures), we applied a conservative approach by pooling the overall PA measure by halving the number of patients in both groups for both outcomes. Pooled associations with 95% confidence intervals are presented, and forest plots with I2 (with test-based 95% confidence intervals)26 are used to display statistical heterogeneity between studies. Where a study contributed more than 1 intervention group to the analysis, we combined them while avoiding double counting of the control group. For cases in which there were insufficient data to include in meta-analyses, we synthesized these narratively in the results.

    Second, mixed-effects univariable and multivariable (multilevel) metaregression analyses were conducted in R statistical software version 3.4.3 (R Project for Statistical Computing) to examine the influence of a number of study-level covariates on the pooled association with the primary outcome (PA levels). The multilevel aspect of the regression model allows for potential clustering by including random effects for both intervention and study. Eight covariates were selected and coded following consensus procedures and informed by the Oxford Implementation Index: sex (100% female vs mixed sex), age (≥50 years), index condition (ie, type 2 diabetes, overweight or obese, cardiovascular disease), type of device, consultations with facilitators, intervention length, goal set for PA, theory-based intervention, and intervention uptake. Results were considered statistically significant at P < .15 using 2-tailed tests. Covariates meeting our significance criterion were entered into a multivariable metaregression model. The P < .15 threshold was conservative to avoid prematurely discounting potentially important explanatory variables, and adjusted tests were used for controlling type I error.27 All analyses were performed in the R version 3.4.3. For each meta-analysis with 10 studies or more, funnel plots, Begg test, and Egger test were used to examine potential publication bias. The trim-and-fill method was used as a sensitivity analysis to observe factors associated with publication bias.

    Results

    After removing duplicates, the search retrieved 5762 references. Following abstract and title screening of 1439 references and full-text screening of 107 studies, 36 studies met our inclusion criteria (20 studies using accelerometers and 16 studies using pedometers) (Figure 1). No unpublished studies were identified.

    Characteristics of Included Studies
    Location, Setting, and Participant Characteristics

    Most studies were conducted in either the United States (9 studies) or the United Kingdom (6 studies). The settings of the studies varied and included hospitals, primary care, medical and community centers, and universities. The 36 studies involved 5208 participants (eTable 2 in the Supplement). Most studies included adults with a mean age between 32 and 71 years, with 13 studies focusing on older adults with a mean age older than 60 years. Four studies28-31 included women only and the remainder involved both sexes. The target populations recruited in the studies were predominantly those diagnosed with type 2 diabetes (16 studies), but others included cardiovascular diseases (13 studies) and obese or mixed obese and overweight participants (7 studies).

    Intervention Characteristics

    The interventions mainly focused on increasing PA, preventing disease, and managing weight (eTable 3 in the Supplement). Nine studies used a theoretical framework consisting of social cognitive approaches (eg, health belief model, theory of planned behavior, or transtheoretical model); however, behavior change outcomes did not appear to have been captured in their results. Fewer than one-third of the studies (12 trials) tested simple pedometer or accelerometer interventions (ie, after an initial consultation session, patients were provided with the accelerometer or pedometer and a log book to self-monitor their outcomes using written instructions with no additional support by facilitators or health care professionals), whereas 24 studies tested more complex interventions that also involved consultation sessions (ie, patients were supported by facilitators who were mainly health professionals via face-to-face consultations and/or telecommunications during the intervention). The median (range) duration for receiving the intervention was 7 months (2 weeks to 4 years), indicating considerable variation in duration. The mean (SD) follow-up was 32 (28.6) weeks.

    Risk of Bias

    The quality of the studies was variable (eFigure 1 in the Supplement). Twenty-two studies (61%) had a low risk of bias for the random sequence generation, and 15 studies (42%) had low risk for allocation concealment. Only 2 studies (6%)32,33 were deemed high risk for this criterion. Similarly, blinding of outcome assessment was moderately reported, with 19 studies (53%) showing low risk; however, 7 studies (19%) reported high risk for this domain. Criteria for incomplete outcome data were mostly satisfied across studies, displaying low risk in 23 studies (64%); however, 7 studies (19%) reported high risk. For selective reporting, only 2 studies (6%)17,34 exhibited high risk of bias.

    Association of the Interventions With PA
    Primary Outcome: PA Improvement

    Twenty-two of the 36 studies were included in the meta-analysis (4856 participants). Summary estimates from the meta-analyses are presented in Table 1. Across all studies involving interventions with monitoring devices vs comparators there was a small to medium significant increase in PA over approximately an 8-month period (SMD, 0.39 [95% CI, 0.28-0.51]; I2 = 60% [95% CI, 41%-73%]) (eFigure 2 in the Supplement). Accelerometer-based interventions demonstrated a small increase in PA compared with comparators (SMD, 0.30; 95% CI, 0.16-0.44; n = 20 studies), and pedometer-based interventions demonstrated a medium increase of PA compared with comparators (SMD, 0.52; 95% CI, 0.32-0.72; n = 15) (eFigure 3 in the Supplement). For pedometer-based interventions, the PA measures translated to 1702.85 steps per day (95% CI, 1066.67-2339.03 steps per day) for intervention vs the usual care group. This value is generally lower than recommendations set out by governments and agencies globally. Heterogeneity was high for pedometer use (I2 = 72% [95% CI, 53%-83%]) and moderate for accelerometer use (I2 = 52% [95% CI, 20%-71%]). The cumulative plot (eFigure 4 in the Supplement) of PA performance based on total session times showed that programs with longer periods of engagement in PA generally performed better. However, 18 studies did not report the length of the sessions and therefore could not be included.

    Moderators of Association With PA (Univariable and Multivariable Metaregression)

    The results of the univariable and multivariable analyses are shown in Table 2. Interventions using consultations with a health professional (β = 0.32; 95% CI, 0.10-0.55; P = .002), pedometer-based interventions (β = 0.24; 95% CI, 0.004-0.48; P = .05), and the inclusion of predominately male participants in studies (β = 0.25; 95% CI, 0.03-0.42; P = .03) were the only factors associated with improved PA levels in the univariable regression analyses. The remaining factors, including index diagnosis of participants, age of participants, length of the intervention, goal setting, underpinning the intervention with a theoretical framework, intervention uptake, and risk of bias scores, were not associated with PA level and were not eligible for inclusion in the multivariable regression analysis. The overall multivariable model was statistically significant (χ23 = 17.46; P < .001) and reduced the I2 statistic from 70% to 46%. Both factors associated with improved PA, consultations with a health professional (β = 0.36; 95% CI, 0.17-0.55; P < .001) and pedometer vs accelerometer (β = 0.30; 95% CI, 0.08-0.52; P = .002), remained significant in the multivariable model. Thus, interventions involving regular consultations by health professionals (compared with self-monitoring only) and pedometer-based interventions (compared with accelerometers) were the main 2 factors associated with improved PA levels.

    We conducted a post hoc subgroup analysis using the variables of enhanced consultation and monitoring device whereby studies17,29-32,35-62 were divided into 4 groups to best visualize the results of metaregression analyses (Figure 2). Pedometer interventions incorporating consultations with health professionals were associated with greater increases in PA (SMD, 0.73; 95% CI, 0.50-0.97; n = 10) and supervised accelerometer interventions were associated small to medium PA increases (SMD, 0.29; 95% CI, 0.14-0.43; n = 12) vs comparators. In contrast, interventions without consultations had a weaker association with PA for pedometers (SMD, 0.18; 95% CI, 0.07-0.29; n = 5) and no association with PA for accelerometers (SMD, 0.28; 95% CI, −0.05 to 0.61; n = 8).

    Secondary Outcomes

    Interventions with monitoring devices were associated with small but statistically significant reductions in blood glucose (HbA1c percentage) vs comparators (MD, −0.25%; 95% CI, −0.45% to −0.06%; n = 13) (Table 1). Pedometer-based interventions were associated with the greatest effectiveness (MD, −0.40; 95% CI, −0.55 to −0.25; n = 10), whereas accelerometer performance was not significant across 2 studies. No association was found for systolic and diastolic blood pressure, total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, BMI, and weight. Other measures of PA, including self-reported PA, metabolic equivalents of tasks, and activity times, were not significant (eFigure 5 in the Supplement).

    Publication Bias

    Publication bias was detected for pedometer and accelerometer use by visual inspection of the funnel plots (eFigure 6 in the Supplement) and as indicated by Begg test (z = 3.301; P = .002) and Egger test (z = 3.2484; bias coefficient, 1.64; 95% CI, 0.74-2.53; P = .001). The trim-and-fill method was used to adjust for potential missing studies.

    Narrative Synthesis

    There were 4 pedometer studies that reported nonamenable data for meta-analysis; 2 studies30,31 reported a significant increase in number of steps and the other 2 studies34,63 reported no increase in the number of steps between intervention groups.

    Discussion

    This systematic review and meta-analysis shows that both accelerometers and pedometers are associated with small to medium improvements in PA among people with cardiometabolic conditions. However, this association was only present over the short to medium term. Metaregressions suggest that the type of device and consultations with health professionals appear to be key factors associated with PA improvements obtained by these interventions. The greatest increases in PA levels for people with cardiometabolic conditions were achieved by complex interventions combining the use of pedometers with regular consultation sessions with a health professional (face-to-face or remotely). For example, in a pedometer-based intervention with additional support targeting patients with type 2 diabetes, Van Dyck and colleagues35 found that increases in PA persisted in the medium term, with those in the intervention group reporting an additional 2967 steps per day on average 1 year later. Accelerometer-based interventions had no association with PA levels compared with comparators when simple self-monitoring was used without additional consultations. Pedometers showed a small significant association with PA when simple self-monitoring of the intervention was used compared with more complex delivered interventions. For example, Yates and colleagues36 found that, at 12 months, those in the pedometer group reported 383 more steps per day than the control group. Some evidence did show that more complex interventions, enhanced with consultations and longer periods of PA engagement (with longer or more sessions) did improve PA levels. However, 19 trials did not report the total PA engagement time. Interventions were also associated with reduced HbA1c levels. For example, in an intervention incorporating pedometers and clinician support, Fayehun et al37 found levels of HbA1c to be significantly lower in the intervention group at an 11 week follow-up. Although the change across all included studies was statically significant compared with usual care, the clinical relevance of the magnitude of change was small (−0.25%). Other secondary health outcomes like blood pressure, cholesterol levels, weight, and BMI showed results that were not statically significant.

    Our findings are consistent with the results of earlier systematic reviews15,64 involving populations at most risk, which suggests medium improvements in PA levels at short-term follow-up assessments in response to pedometer- or accelerometer-based interventions. The most recent review64 reported a medium increase in PA (SMD, 0.57; 95% CI, 0.24-0.91) but focused solely on people with type 2 diabetes. Moreover, our extensive searches and broader eligibility criteria in terms of population resulted in pooling PA outcomes from 32 studies, up to 3 times the number of RCTs included in previous reviews. We have also assessed the association of several study-level factors with PA improvements. Our analyses support the importance of complex interventions involving consultation sessions with health professionals for boosting PA benefits for people with cardiometabolic conditions. Furthermore, a recent review19 of consumer-based wearable activity trackers in general populations indicated improved associations with PA levels but limited availability of long-term follow-up data. While this study is related to our review with a focus on electronic devices for monitoring PA, it focused on studies conducted among healthy general populations and not those at risk for chronic conditions.

    Strengths and Limitations

    The major strength of this systematic review and meta-analysis is that we searched 5 major databases for relevant literature and used well-established statistical methods, including pairwise meta-analysis and multilevel multivariable metaregression, to explore the full association of associated factors. However, there are also limitations. We performed metaregression to explore the heterogeneity observed in the main analyses, but important uncertainties remain regarding risk of bias assessments with many unclear domains, whereas participant characteristics such as age and sex are based on aggregate data. Therefore, the metaregression results should be interpreted with caution. Owing to the heterogeneous nature of the 13 cardiovascular disease trials studied and the limited number of trials (7) that focused on overweight or obese participants, we chose to combine the trials under the definition of a cardiometabolic condition. We included only those study populations at risk; however, we understand that this is a potential limitation as studies that did not explicitly report this type of population may have been missed. A network meta-analysis was not performed because all the evidence comparing interventions with one another is reliant on indirect evidence only; therefore, quantifying the change in PA was easier to assess in the pairwise analysis. We initially planned to perform a bivariate meta-analysis estimating the overall correlation between outcomes65; however, owing to the large within-study variation, we were not able to estimate correlations for PA outcomes and were only able to look at outcomes for HbA1c levels, BMI, weight, and cholesterol levels. Although we performed extensive searches to identify all relevant published and unpublished studies in both the ClinicalTrials.gov and OpenTrials.net databases, our formal tests indicated small study bias, indicating publication bias. As recommended, we dealt with small study bias using the trim-and-fill method. Also, we did not look at behavior change outcomes such as those reported in line with the theoretical domains framework, as only 1 trial mentioned explicitly that such outcomes would be collected.66

    Implications for Future Research and Practice

    We found that the use of accelerometers increased the levels of PA by approximately an SMD of 0.30 increase in MVPA and that pedometers increase steps per day by an SMD of 0.52 or an MD of 1703 steps per day among people with cardiometabolic conditions. These values are generally lower than the recommendations of the most recent 2018 Physical Activity Guidelines Advisory Committee Scientific Report20 by the US Department of Health and Human Services and other recommendations set out by governments and agencies globally.67 For instance, the UK National Obesity Forum classifies 3000 to 6000 steps per day as sedentary, Northern Ireland’s Public Health Agency promotes an additional 3000 steps, and the America on the Move campaign suggests an additional 2000 steps each day to stop weight gain. For accelerometers, public health guidelines endorse 30 minutes (at times up to 60 minutes) per day (or 150-210 minutes per week) of MVPA, typically in bouts of at least 10 minutes.68,69 This could not be assessed because the total MVPA session times in minutes were rarely reported in the studies.

    We found evidence that complex interventions that combine the use of monitoring devices (particularly pedometers) with regular consultations with health professionals might be an effective way of increasing PA and reaching the recommendations set out by governments and agencies for people with cardiometabolic conditions. It is likely that giving feedback and lifestyle advice to patients on a regular basis supported the effectiveness of these interventions. Several studies have suggested positive associations of multimodal pedometer interventions with PA levels in a range of populations, including those with type 2 diabetes and cardiac conditions,38 but this is the first study, to our knowledge, to highlight the role of regular consultations in the association of pedometer interventions with PA in people with cardiometabolic conditions. This finding warrants consideration in future trials and further investigation using more robust methods such as individual participant data meta-analyses. Moreover, we only found 2 studies that reported associations between accelerometer or pedometer use and PA levels after a 1-year follow-up period.39,40 Long-term follow-up assessments are needed to generate evidence regarding the sustainability of PA increases over time. Such long-term assessments have a greater potential to affect outcome performance and potentially explain more about the intervention program than current short-term assessments.

    Conclusions

    This systematic review found that, in participants with cardiometabolic conditions, the use of monitoring devices increased the levels of PA by approximately 1703 steps per day (SMD = 0.52) and increased MVPA by a SMD of 0.22. However, the evidence is only over a short to medium period. Complex pedometer and accelerometer interventions that used complementary consultations with health professionals appear to be the most promising in improving PA among people with cardiometabolic conditions. Understanding the association between accelerometer- and pedometer-based interventions and PA over the longer term could have major implications in the care of people with cardiometabolic conditions.

    Back to top
    Article Information

    Accepted for Publication: August 19, 2019.

    Published: October 9, 2019. doi:10.1001/jamanetworkopen.2019.12895

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2019 Hodkinson A et al. JAMA Network Open.

    Corresponding Author: Alexander Hodkinson, PhD, Centre for Primary Care, Division of Population Health, Health Services Research & Primary Care, Williamson Building, Oxford Road, Manchester M13 9PL, United Kingdom (alexander.hodkinson@manchester.ac.uk).

    Author Contributions: Dr Hodkinson had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Hodkinson, Kontopantelis, van Marwijk, Bower, Panagioti.

    Acquisition, analysis, or interpretation of data: Hodkinson, Kontopantelis, Adeniji, van Marwijk, McMillan, Panagioti.

    Drafting of the manuscript: Hodkinson, van Marwijk, Bower, Panagioti.

    Critical revision of the manuscript for important intellectual content: Hodkinson, Kontopantelis, Adeniji, van Marwijk, McMillan, Panagioti.

    Statistical analysis: Hodkinson, Kontopantelis, Adeniji, Panagioti.

    Obtained funding: Kontopantelis, van Marwijk, Bower, Panagioti.

    Administrative, technical, or material support: Hodkinson, Adeniji.

    Supervision: Hodkinson, van Marwijk, Bower, Panagioti.

    Conflict of Interest Disclosures: None reported.

    Funding/Support: This study was funded by the National Institute for Health Research School for Primary Care Research.

    Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    Disclaimer: The views expressed are those of the authors and not necessarily those of the National Institute for Health Research or the Department of Health and Social Care.

    Additional Contributions: We thank the evidence synthesis working group at National Institute of Health Research who funded this project and ongoing work.

    References
    1.
    New York State. Office of Mental Health. Cardiometabolic risk: what is it and what can I do about it? https://www.omh.ny.gov/omhweb/psyckes_medicaid/brochures/cardio.html. Accessed October 2018.
    2.
    Kivimäki  M, Pentti  J, Ferrie  JE,  et al; IPD-Work Consortium.  Work stress and risk of death in men and women with and without cardiometabolic disease: a multicohort study.  Lancet Diabetes Endocrinol. 2018;6(9):705-713. doi:10.1016/S2213-8587(18)30140-2PubMedGoogle ScholarCrossref
    3.
    National Institute for Health and Care Excellence. Type 2 diabetes in adults: management (NICE guideline). https://www.nice.org.uk/guidance/ng28/resources/type-2-diabetes-in-adults-management-pdf-1837338615493. Published December 2, 2015. Accessed April 2019.
    4.
    Diabetes Times. Diabetes NHS costs could hit £17 billion. https://diabetestimes.co.uk/diabetes-nhs-costs-could-hit-17-billion/. Published October 25, 2016. Accessed January 2019.
    5.
    McNamara  E, Hudson  Z, Taylor  SJ.  Measuring activity levels of young people: the validity of pedometers.  Br Med Bull. 2010;95:121-137. doi:10.1093/bmb/ldq016PubMedGoogle ScholarCrossref
    6.
    Kvaavik  E, Batty  GD, Ursin  G, Huxley  R, Gale  CR.  Influence of individual and combined health behaviors on total and cause-specific mortality in men and women: the United Kingdom Health and Lifestyle Survey.  Arch Intern Med. 2010;170(8):711-718. doi:10.1001/archinternmed.2010.76PubMedGoogle ScholarCrossref
    7.
    Public Health England. Physical activity: applying All Our Health. https://www.gov.uk/government/publications/physical-activity-applying-all-our-health/physical-activity-applying-all-our-health. Updated January 9, 2018. Accessed September 10, 2018.
    8.
    National Institute for Health and Clinical Excellence. Physical activity: walking and cycling (Public Health Guidance PH41). https://www.nice.org.uk/guidance/ph41. Published November 2012. Accessed September 10, 2018.
    9.
    Centers for Disease Control and Prevention (CDC).  School health guidelines to promote healthy eating and physical activity.  MMWR Recomm Rep. 2011;60(RR-5):1-76.PubMedGoogle Scholar
    10.
    National Institute for Health and Clinical Excellence. Physical activity in the workplace (Public Health Guidance PH13). https://www.nice.org.uk/guidance/ph13. Published May 2008. Accessed August 30, 2017.
    11.
    de Souto Barreto  P.  Time to challenge public health guidelines on physical activity.  Sports Med. 2015;45(6):769-773. doi:10.1007/s40279-015-0326-7PubMedGoogle ScholarCrossref
    12.
    Afshin  A, Babalola  D, Mclean  M,  et al.  Information technology and lifestyle: a systematic evaluation of internet and mobile interventions for improving diet, physical activity, obesity, tobacco, and alcohol use.  J Am Heart Assoc. 2016;5(9):e003058. doi:10.1161/JAHA.115.003058PubMedGoogle ScholarCrossref
    13.
    Ayabe  M, Ishii  K, Takayama  K, Aoki  J, Tanaka  H.  Comparison of interdevice measurement difference of pedometers in younger and older adults.  Br J Sports Med. 2010;44(2):95-99. doi:10.1136/bjsm.2007.045179PubMedGoogle ScholarCrossref
    14.
    Bravata  DM, Smith-Spangler  C, Sundaram  V,  et al.  Using pedometers to increase physical activity and improve health: a systematic review.  JAMA. 2007;298(19):2296-2304. doi:10.1001/jama.298.19.2296PubMedGoogle ScholarCrossref
    15.
    Qiu  S, Cai  X, Chen  X, Yang  B, Sun  Z.  Step counter use in type 2 diabetes: a meta-analysis of randomized controlled trials.  BMC Med. 2014;12:36. doi:10.1186/1741-7015-12-36PubMedGoogle ScholarCrossref
    16.
    Public Health England. Adult obesity and type 2 diabetes. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/338934/Adult_obesity_and_type_2_diabetes_.pdf. Published July 2014. Accessed February 5, 2019.
    17.
    Higgins JPT, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0. London, UK: Cochrane Collaboration; 2011.
    18.
    Liberati  A, Altman  DG, Tetzlaff  J,  et al.  The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration.  BMJ. 2009;339:b2700. doi:10.1136/bmj.b2700PubMedGoogle ScholarCrossref
    19.
    Brickwood  K-J, Watson  G, O’Brien  J, Williams  AD.  Consumer-based wearable activity trackers increase physical activity participation: systematic review and meta-analysis.  JMIR Mhealth Uhealth. 2019;7(4):e11819-e11819. doi:10.2196/11819PubMedGoogle ScholarCrossref
    20.
    Department of Health and Human Services. 2018 Physical Activity Guidelines Advisory Committee Scientific Report. https://health.gov/paguidelines/second-edition/report/pdf/PAG_Advisory_Committee_Report.pdf. Accessed August 8, 2019.
    21.
    Montgomery  P, Underhill  K, Gardner  F, Operario  D, Mayo-Wilson  E.  The Oxford Implementation Index: a new tool for incorporating implementation data into systematic reviews and meta-analyses.  J Clin Epidemiol. 2013;66(8):874-882. doi:10.1016/j.jclinepi.2013.03.006PubMedGoogle ScholarCrossref
    22.
    Higgins  JPT, Altman  DG, Gøtzsche  PC,  et al; Cochrane Bias Methods Group; Cochrane Statistical Methods Group.  The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials.  BMJ. 2011;343:d5928. doi:10.1136/bmj.d5928PubMedGoogle ScholarCrossref
    23.
    R [computer program]. Version 3.3.3. Vienna, Austria: R Project for Statistical Computing.
    24.
    Hedges  LV.  Distribution theory for Glass’s estimator of effect size and related estimators.  J Educ Behav Stat. 1981;6(2):107-128. doi:10.3102/10769986006002107Google ScholarCrossref
    25.
    Cohen J.  Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Oxfordshire, United Kingdom: Taylor & Francis; 1977.
    26.
    Higgins  JP, Thompson  SG.  Quantifying heterogeneity in a meta-analysis.  Stat Med. 2002;21(11):1539-1558. doi:10.1002/sim.1186PubMedGoogle ScholarCrossref
    27.
    IntHout  J, Ioannidis  JP, Borm  GF.  The Hartung-Knapp-Sidik-Jonkman method for random effects meta-analysis is straightforward and considerably outperforms the standard DerSimonian-Laird method.  BMC Med Res Methodol. 2014;14(1):25. doi:10.1186/1471-2288-14-25PubMedGoogle ScholarCrossref
    28.
    Cayir  Y, Aslan  SM, Akturk  Z.  The effect of pedometer use on physical activity and body weight in obese women.  Eur J Sport Sci. 2015;15(4):351-356. doi:10.1080/17461391.2014.940558PubMedGoogle ScholarCrossref
    29.
    Greaney  ML, Askew  S, Wallington  SF, Foley  PB, Quintiliani  LM, Bennett  GG.  The effect of a weight gain prevention intervention on moderate-vigorous physical activity among black women: the Shape Program.  Int J Behav Nutr Phys Act. 2017;14(1):139. doi:10.1186/s12966-017-0596-6PubMedGoogle ScholarCrossref
    30.
    Holliday  A, Burgin  A, Fernandez  EV, Fenton  SAM, Thielecke  F, Blannin  AK.  Points-based physical activity: a novel approach to facilitate changes in body composition in inactive women with overweight and obesity.  BMC Public Health. 2018;18(1):261. doi:10.1186/s12889-018-5125-2PubMedGoogle ScholarCrossref
    31.
    Pekmezi  D, Ainsworth  C, Joseph  RP,  et al.  Pilot trial of a home-based physical activity program for African American women.  Med Sci Sports Exerc. 2017;49(12):2528-2536. doi:10.1249/MSS.0000000000001370PubMedGoogle ScholarCrossref
    32.
    Christle  JW, Schlumberger  A, Haller  B, Gloeckl  R, Halle  M, Pressler  A.  Individualized vs. group exercise in improving quality of life and physical activity in patients with cardiac disease and low exercise capacity: results from the DOPPELHERZ trial.  Disabil Rehabil. 2017;39(25):2566-2571. doi:10.1080/09638288.2016.1242174PubMedGoogle ScholarCrossref
    33.
    Frederix  I, Hansen  D, Coninx  K,  et al.  Medium-term effectiveness of a comprehensive internet-based and patient-specific telerehabilitation program with text messaging support for cardiac patients: randomized controlled trial.  J Med Internet Res. 2015;17(7):e185. doi:10.2196/jmir.4799PubMedGoogle ScholarCrossref
    34.
    Engel  L, Lindner  H.  Impact of using a pedometer on time spent walking in older adults with type 2 diabetes.  Diabetes Educ. 2006;32(1):98-107. doi:10.1177/0145721705284373PubMedGoogle ScholarCrossref
    35.
    Van Dyck  D, De Greef  K, Deforche  B,  et al.  The relationship between changes in steps/day and health outcomes after a pedometer-based physical activity intervention with telephone support in type 2 diabetes patients.  Health Educ Res. 2013;28(3):539-545. doi:10.1093/her/cyt038PubMedGoogle ScholarCrossref
    36.
    Yates  T, Edwardson  CL, Henson  J,  et al.  Walking Away From Type 2 Diabetes: a cluster randomized controlled trial.  Diabet Med. 2017;34(5):698-707. doi:10.1111/dme.13254PubMedGoogle ScholarCrossref
    37.
    Fayehun  AF, Olowookere  OO, Ogunbode  AM, Adetunji  AA, Esan  A.  Walking prescription of 10 000 steps per day in patients with type 2 diabetes mellitus: a randomised trial in Nigerian general practice.  Br J Gen Pract. 2018;68(667):e139-e145. doi:10.3399/bjgp18X694613PubMedGoogle ScholarCrossref
    38.
    De Greef  K, Deforche  B, Tudor-Locke  C, De Bourdeaudhuij  I.  A cognitive-behavioural pedometer-based group intervention on physical activity and sedentary behaviour in individuals with type 2 diabetes.  Health Educ Res. 2010;25(5):724-736. doi:10.1093/her/cyq017PubMedGoogle ScholarCrossref
    39.
    Unick  JL, Gaussoin  SA, Hill  JO,  et al.  Four-year physical activity levels among intervention participants with type 2 diabetes.  Med Sci Sports Exerc. 2016;48(12):2437-2445. doi:10.1249/MSS.0000000000001054PubMedGoogle ScholarCrossref
    40.
    Unick  JL, Gaussoin  SA, Hill  JO,  et al; Look AHEAD Research Group. Objectively assessed physical activity and weight loss maintenance among individuals enrolled in a lifestyle intervention.  Obesity (Silver Spring). 2017;25(11):1903-1909. doi:10.1002/oby.21971PubMedGoogle ScholarCrossref
    41.
    Andersen  E, Ekelund  U, Anderssen  SA.  Effects of reducing sedentary time on glucose metabolism in immigrant Pakistani men.  Med Sci Sports Exerc. 2015;47(4):775-781. doi:10.1249/MSS.0000000000000460PubMedGoogle ScholarCrossref
    42.
    Araiza  P, Hewes  H, Gashetewa  C, Vella  CA, Burge  MR.  Efficacy of a pedometer-based physical activity program on parameters of diabetes control in type 2 diabetes mellitus.  Metabolism. 2006;55(10):1382-1387. doi:10.1016/j.metabol.2006.06.009PubMedGoogle ScholarCrossref
    43.
    Baillot  A, Vallée  CA, Mampuya  WM,  et al.  Effects of a pre-surgery supervised exercise training 1 year after bariatric surgery: a randomized controlled study.  Obes Surg. 2018;28(4):955-962. doi:10.1007/s11695-017-2943-8PubMedGoogle ScholarCrossref
    44.
    Borland  M, Rosenkvist  A, Cider  A.  A group-based exercise program did not improve physical activity in patients with chronic heart failure and comorbidity: a randomized controlled trial.  J Rehabil Med. 2014;46(5):461-467. doi:10.2340/16501977-1794PubMedGoogle ScholarCrossref
    45.
    Cowie  A, Thow  MK, Granat  MH, Mitchell  SL.  A comparison of home and hospital-based exercise training in heart failure: immediate and long-term effects upon physical activity level.  Eur J Cardiovasc Prev Rehabil. 2011;18(2):158-166. doi:10.1177/1741826710389389PubMedGoogle ScholarCrossref
    46.
    Cupples  M, Dean  A, Tully  MA,  et al.  Using pedometer step-count goals to promote physical activity in cardiac rehabilitation: a feasibility study of a controlled trial.  Int J Phys Med Rehabil. 2013;1(7):157. doi:10.4172/2329-9096.1000157Google Scholar
    47.
    Dasgupta  K, Rosenberg  E, Joseph  L,  et al; SMARTER Trial Group.  Physician step prescription and monitoring to improve ARTERial health (SMARTER): A randomized controlled trial in patients with type 2 diabetes and hypertension.  Diabetes Obes Metab. 2017;19(5):695-704. doi:10.1111/dom.12874PubMedGoogle ScholarCrossref
    48.
    De Greef  K, Deforche  B, Tudor-Locke  C, De Bourdeaudhuij  I.  Increasing physical activity in Belgian type 2 diabetes patients: a three-arm randomized controlled trial.  Int J Behav Med. 2011;18(3):188-198. doi:10.1007/s12529-010-9124-7PubMedGoogle ScholarCrossref
    49.
    De Greef  KP, Deforche  BI, Ruige  JB,  et al.  The effects of a pedometer-based behavioral modification program with telephone support on physical activity and sedentary behavior in type 2 diabetes patients.  Patient Educ Couns. 2011;84(2):275-279. doi:10.1016/j.pec.2010.07.010PubMedGoogle ScholarCrossref
    50.
    Devi  R, Powell  J, Singh  S.  A web-based program improves physical activity outcomes in a primary care angina population: randomized controlled trial.  J Med Internet Res. 2014;16(9):e186. doi:10.2196/jmir.3340PubMedGoogle ScholarCrossref
    51.
    Frederix  I, Van Driessche  N, Hansen  D,  et al.  Increasing the medium-term clinical benefits of hospital-based cardiac rehabilitation by physical activity telemonitoring in coronary artery disease patients.  Eur J Prev Cardiol. 2015;22(2):150-158. doi:10.1177/2047487313514018PubMedGoogle ScholarCrossref
    52.
    Guiraud  T, Granger  R, Gremeaux  V,  et al.  Telephone support oriented by accelerometric measurements enhances adherence to physical activity recommendations in noncompliant patients after a cardiac rehabilitation program.  Arch Phys Med Rehabil. 2012;93(12):2141-2147. doi:10.1016/j.apmr.2012.06.027PubMedGoogle ScholarCrossref
    53.
    Houle  J, Doyon  O, Vadeboncoeur  N, Turbide  G, Diaz  A, Poirier  P.  Innovative program to increase physical activity following an acute coronary syndrome: randomized controlled trial.  Patient Educ Couns. 2011;85(3):e237-e244. doi:10.1016/j.pec.2011.03.018PubMedGoogle ScholarCrossref
    54.
    Houle  J, Doyon  O, Vadeboncoeur  N, Turbide  G, Diaz  A, Poirier  P.  Effectiveness of a pedometer-based program using a socio-cognitive intervention on physical activity and quality of life in a setting of cardiac rehabilitation.  Can J Cardiol. 2012;28(1):27-32. doi:10.1016/j.cjca.2011.09.020PubMedGoogle ScholarCrossref
    55.
    Izawa  KP, Watanabe  S, Hiraki  K,  et al.  Determination of the effectiveness of accelerometer use in the promotion of physical activity in cardiac patients: a randomized controlled trial.  Arch Phys Med Rehabil. 2012;93(11):1896-1902. doi:10.1016/j.apmr.2012.06.015PubMedGoogle ScholarCrossref
    56.
    Katzmarzyk  PT, Champagne  CM, Tudor-Locke  C,  et al.  A short-term physical activity randomized trial in the lower Mississippi Delta.  PLoS One. 2011;6(10):e26667. doi:10.1371/journal.pone.0026667PubMedGoogle ScholarCrossref
    57.
    Kirk  A, Barnett  J, Leese  G, Mutrie  N.  A randomized trial investigating the 12-month changes in physical activity and health outcomes following a physical activity consultation delivered by a person or in written form in Type 2 diabetes: Time2Act.  Diabet Med. 2009;26(3):293-301. doi:10.1111/j.1464-5491.2009.02675.xPubMedGoogle ScholarCrossref
    58.
    Miyamoto  T, Fukuda  K, Oshima  Y, Moritani  T.  Non-locomotive physical activity intervention using a tri-axial accelerometer reduces sedentary time in type 2 diabetes.  Phys Sportsmed. 2017;45(3):245-251. doi:10.1080/00913847.2017.1350084PubMedGoogle ScholarCrossref
    59.
    Paschali  AA, Goodrick  GK, Kalantzi-Azizi  A, Papadatou  D, Balasubramanyam  A.  Accelerometer feedback to promote physical activity in adults with type 2 diabetes: a pilot study.  Percept Mot Skills. 2005;100(1):61-68. doi:10.2466/pms.100.1.61-68PubMedGoogle ScholarCrossref
    60.
    Ribeiro  F, Oliveira  NL, Silva  G,  et al.  Exercise-based cardiac rehabilitation increases daily physical activity of patients following myocardial infarction: subanalysis of two randomised controlled trials.  Physiotherapy. 2017;103(1):59-65. doi:10.1016/j.physio.2015.12.002PubMedGoogle ScholarCrossref
    61.
    Tudor-Locke  C, Bell  RC, Myers  AM,  et al.  Controlled outcome evaluation of the First Step Program: a daily physical activity intervention for individuals with type II diabetes.  Int J Obes Relat Metab Disord. 2004;28(1):113-119. doi:10.1038/sj.ijo.0802485PubMedGoogle ScholarCrossref
    62.
    Young  L, Hertzog  M, Barnason  S.  Effects of a home-based activation intervention on self-management adherence and readmission in rural heart failure patients: the PATCH randomized controlled trial.  BMC Cardiovasc Disord. 2016;16(1):176-176. doi:10.1186/s12872-016-0339-7PubMedGoogle ScholarCrossref
    63.
    Bjørgaas  MR, Vik  JT, Stølen  T, Lydersen  S, Grill  V.  Regular use of pedometer does not enhance beneficial outcomes in a physical activity intervention study in type 2 diabetes mellitus.  Metabolism. 2008;57(5):605-611. doi:10.1016/j.metabol.2007.12.002PubMedGoogle ScholarCrossref
    64.
    Baskerville  R, Ricci-Cabello  I, Roberts  N, Farmer  A.  Impact of accelerometer and pedometer use on physical activity and glycaemic control in people with type 2 diabetes: a systematic review and meta-analysis.  Diabet Med. 2017;34(5):612-620. doi:10.1111/dme.13331PubMedGoogle ScholarCrossref
    65.
    Riley  RD, Thompson  JR, Abrams  KR.  An alternative model for bivariate random-effects meta-analysis when the within-study correlations are unknown.  Biostatistics. 2008;9(1):172-186. doi:10.1093/biostatistics/kxm023PubMedGoogle ScholarCrossref
    66.
    Kelly  MP, Barker  M.  Why is changing health-related behaviour so difficult?  Public Health. 2016;136:109-116. doi:10.1016/j.puhe.2016.03.030PubMedGoogle ScholarCrossref
    67.
    Tudor-Locke  C, Craig  CL, Brown  WJ,  et al.  How many steps/day are enough? for adults.  Int J Behav Nutr Phys Act. 2011;8:79-79. doi:10.1186/1479-5868-8-79PubMedGoogle ScholarCrossref
    68.
    Physical Activity Guidelines Advisory Committee.  Physical Activity Guidelines Advisory Committee Report, 2008. Washington, DC: US Department of Health and Human Services; 2008.
    69.
    O’Donovan  G, Blazevich  AJ, Boreham  C,  et al.  The ABC of physical activity for health: a consensus statement from the British Association of Sport and Exercise Sciences.  J Sports Sci. 2010;28(6):573-591. doi:10.1080/02640411003671212PubMedGoogle ScholarCrossref
    ×