Effects of Counseling by Peer Human Advisors vs Computers to Increase Walking in Underserved Populations: The COMPASS Randomized Clinical Trial | Adolescent Medicine | JAMA Internal Medicine | JAMA Network
[Skip to Navigation]
Figure 1.  The Carmen Virtual Advisor
The Carmen Virtual Advisor

Image used with the permission of Timothy Bickmore, PhD.

Figure 2.  Consolidated Standards of Reporting Trials Diagram
Consolidated Standards of Reporting Trials Diagram
Figure 3.  Mean 12-Month Change in Total Walking Minutes per Week
Mean 12-Month Change in Total Walking Minutes per Week

Error bars indicate 95% CIs.

Table 1.  Baseline Descriptive Statistics for the Sample and by Randomization Arm
Baseline Descriptive Statistics for the Sample and by Randomization Arm
Table 2.  Primary and Secondary Outcomes
Primary and Secondary Outcomes
1.
Guthold  R, Stevens  GA, Riley  LM, Bull  FC.  Worldwide trends in insufficient physical activity from 2001 to 2016: a pooled analysis of 358 population-based surveys with 1·9 million participants.   Lancet Glob Health. 2018;6(10):e1077-e1086. doi:10.1016/S2214-109X(18)30357-7 PubMedGoogle ScholarCrossref
2.
Piercy  KL, Troiano  RP, Ballard  RM,  et al.  The Physical Activity Guidelines for Americans.   JAMA. 2018;320(19):2020-2028. doi:10.1001/jama.2018.14854 PubMedGoogle ScholarCrossref
3.
US Dept of Health & Human Services.  Physical Activity Guidelines for Americans, 2nd edition. US Dept of Health & Human Services;2018.
4.
Physical Activity Guidelines Advisory Committee.  2018 Physical Activity Guidelines Advisory Committee Scientific Report. US Dept of Health & Human Services;2018.
5.
US Dept of Health & Human Services. HHS action plan to reduce racial and ethnic disparities: a nation free of disparities in health and health care. US Dept of Health & Human Services; April 2011.
6.
Latulippe  K, Hamel  C, Giroux  D.  Social health inequalities and eHealth: a literature review with qualitative synthesis of theoretical and empirical studies.   J Med internet Res. 2017;19(4):e136. doi:10.2196/jmir.6731 PubMedGoogle Scholar
7.
Castro  CM, Pruitt  LA, Buman  MP, King  AC.  Physical activity program delivery by professionals versus volunteers: the TEAM randomized trial.   Health Psychol. 2011;30(3):285-294. doi:10.1037/a0021980 PubMedGoogle ScholarCrossref
8.
Hooker  SP, Seavey  W, Weidmer  CE,  et al.  The California active aging community grant program: translating science into practice to promote physical activity in older adults.   Ann Behav Med. 2005;29(3):155-165. doi:10.1207/s15324796abm2903_1 PubMedGoogle ScholarCrossref
9.
MacNab  M, Lee  SH, McCloughan  L, Hanley  J, McKinstry  B, Pinnock  H.  Oximetry-supported self-management for chronic obstructive pulmonary disease: mixed method feasibility pilot project.   BMC Health Serv Res. 2015;15:485. doi:10.1186/s12913-015-1135-2 PubMedGoogle ScholarCrossref
10.
Lauer  MS, Gordon  D, Wei  G, Pearson  G.  Efficient design of clinical trials and epidemiological research: is it possible?   Nat Rev Cardiol. 2017;14(8):493-501. doi:10.1038/nrcardio.2017.60 PubMedGoogle ScholarCrossref
11.
SAS Institute Inc.  SAS/STAT User’s Guide: Version 15.1. SAS Institute, Inc; 2018.
12.
King  AC, Friedman  R, Marcus  B,  et al.  Ongoing physical activity advice by humans versus computers: the Community Health Advice by Telephone (CHAT) trial.   Health Psychol. 2007;26(6):718-727. doi:10.1037/0278-6133.26.6.718 PubMedGoogle ScholarCrossref
13.
King  AC, Campero  I, Sheats  JL,  et al.  Testing the comparative effects of physical activity advice by humans vs computers in underserved populations: the COMPASS trial design, methods, and baseline characteristics.   Contemp Clin Trials. 2017;61(October):115-125. doi:10.1016/j.cct.2017.07.020 PubMedGoogle ScholarCrossref
14.
Physical Activity Guidelines Advisory Committee.  Report of the Physical Activity Guidelines Advisory Committee, 2008. Washington, DC: US Dept of Health & Human Services;2008.
15.
Stewart  AL, Mills  KM, King  AC, Haskell  WL, Gillis  D, Ritter  PL.  CHAMPS physical activity questionnaire for older adults: outcomes for interventions.   Med Sci Sports Exerc. 2001;33(7):1126-1141. doi:10.1097/00005768-200107000-00010 PubMedGoogle ScholarCrossref
16.
Thomas  S, Reading  J, Shephard  RJ.  Revision of the Physical Activity Readiness Questionnaire (PAR-Q).   Can J Sport Sci. 1992;17(4):338-345.PubMedGoogle Scholar
17.
Hernández-Nieto  R, Gutiérrez  MC, Moreno-Fernández  F.  Hispanic Map of the United States—2017. Harvard University;2017. doi:10.15427/OR035-11/2017EN
18.
Alcalá  HE, Chen  J, Langellier  BA, Roby  DH, Ortega  AN.  Impact of the Affordable Care Act on health care access and utilization among Latinos.   J Am Board Fam Med. 2017;30(1):52-62. doi:10.3122/jabfm.2017.01.160208 PubMedGoogle ScholarCrossref
19.
Ayala  GX, Vaz  L, Earp  JA, Elder  JP, Cherrington  A.  Outcome effectiveness of the lay health advisor model among Latinos in the United States: an examination by role.   Health Educ Res. 2010;25(5):815-840. doi:10.1093/her/cyq035 PubMedGoogle ScholarCrossref
20.
Pahor  M, Blair  SN, Espeland  M,  et al; LIFE Study Investigators.  Effects of a physical activity intervention on measures of physical performance: results of the Lifestyle Interventions and Independence for Elders Pilot (LIFE-P) study.   J Gerontol A Biol Sci Med Sci. 2006;61(11):1157-1165. doi:10.1093/gerona/61.11.1157 PubMedGoogle ScholarCrossref
21.
King  AC, Bickmore  TW, Campero  MI, Pruitt  LA, Yin  JL.  Employing virtual advisors in preventive care for underserved communities: results from the COMPASS study.   J Health Commun. 2013;18(12):1449-1464. doi:10.1080/10810730.2013.798374 PubMedGoogle ScholarCrossref
22.
Bickmore  T, Gruber  A, Picard  R.  Establishing the computer-patient working alliance in automated health behavior change interventions.   Patient Educ Couns. 2005;59(1):21-30. doi:10.1016/j.pec.2004.09.008 PubMedGoogle ScholarCrossref
23.
Bickmore  TW, Pfeifer  LM, Byron  D,  et al.  Usability of conversational agents by patients with inadequate health literacy: evidence from two clinical trials.   J Health Commun. 2010;15(suppl 2):197-210. doi:10.1080/10810730.2010.499991 PubMedGoogle ScholarCrossref
24.
Kincaid  JP, Braby  R, Mears  J.  Electronic authoring and delivery of technical information.   J Instructional Development. 1988;11(2):8-13. doi:10.1007/BF02904998 Google ScholarCrossref
25.
Fitzsimmons  PR, Michael  BD, Hulley  JL, Scott  GO.  A readability assessment of online Parkinson’s disease information.   J R Coll Physicians Edinb. 2010;40(4):292-296. doi:10.4997/JRCPE.2010.401 PubMedGoogle ScholarCrossref
26.
Wilcox  S, Dowda  M, Griffin  SF,  et al.  Results of the first year of active for life: translation of 2 evidence-based physical activity programs for older adults into community settings.   Am J Public Health. 2006;96(7):1201-1209. doi:10.2105/AJPH.2005.074690 PubMedGoogle ScholarCrossref
27.
Hekler  EB, Buman  MP, Ahn  D, Dunton  G, Atienza  AA, King  AC.  Are daily fluctuations in perceived environment associated with walking?   Psychol Health. 2012;27(9):1009-1020. doi:10.1080/08870446.2011.645213 PubMedGoogle ScholarCrossref
28.
Pahor  M, Guralnik  JM, Ambrosius  WT,  et al; LIFE study investigators.  Effect of structured physical activity on prevention of major mobility disability in older adults: the LIFE study randomized clinical trial.   JAMA. 2014;311(23):2387-2396. doi:10.1001/jama.2014.5616 PubMedGoogle ScholarCrossref
29.
King  AC, Sallis  JF, Frank  LD,  et al.  Aging in neighborhoods differing in walkability and income: associations with physical activity and obesity in older adults.   Soc Sci Med. 2011;73(10):1525-1533. doi:10.1016/j.socscimed.2011.08.032 PubMedGoogle ScholarCrossref
30.
Troiano  RP, McClain  JJ, Brychta  RJ, Chen  KY.  Evolution of accelerometer methods for physical activity research.   Br J Sports Med. 2014;48(13):1019-1023. doi:10.1136/bjsports-2014-093546 PubMedGoogle ScholarCrossref
31.
Colbert  LH, Matthews  CE, Havighurst  TC, Kim  K, Schoeller  DA.  Comparative validity of physical activity measures in older adults.   Med Sci Sports Exerc. 2011;43(5):867-876. doi:10.1249/MSS.0b013e3181fc7162 PubMedGoogle ScholarCrossref
32.
Evenson  KR, Wen  F, Herring  AH.  Associations of accelerometry-assessed and self-reported physical activity and sedentary behavior with all-cause and cardiovascular mortality among US adults.   Am J Epidemiol. 2016;184(9):621-632. doi:10.1093/aje/kww070 PubMedGoogle ScholarCrossref
33.
Evenson  KR, Wen  F, Metzger  JS, Herring  AH.  Physical activity and sedentary behavior patterns using accelerometry from a national sample of United States adults.   Int J Behav Nutr Phys Act. 2015;12:20. doi:10.1186/s12966-015-0183-7 PubMedGoogle ScholarCrossref
34.
Gardiner  PA, Clark  BK, Healy  GN, Eakin  EG, Winkler  EAH, Owen  N.  Measuring older adults’ sedentary time: reliability, validity, and responsiveness.   Med Sci Sports Exerc. 2011;43(11):2127-2133. doi:10.1249/MSS.0b013e31821b94f7 PubMedGoogle ScholarCrossref
35.
Myers  AM, Malott  OW, Gray  E,  et al.  Measuring accumulated health-related benefits of exercise participation for older adults: the Vitality Plus Scale.   J Gerontol A Biol Sci Med Sci. 1999;54(9):M456-M466. doi:10.1093/gerona/54.9.M456 PubMedGoogle ScholarCrossref
36.
King  AC, Haskell  WL, Taylor  CB, Kraemer  HC, DeBusk  RF.  Group- vs home-based exercise training in healthy older men and women: a community-based clinical trial.   JAMA. 1991;266(11):1535-1542. doi:10.1001/jama.1991.03470110081037 PubMedGoogle ScholarCrossref
37.
Winkleby  MA, Jatulis  DE, Frank  E, Fortmann  SP.  Socioeconomic status and health: how education, income, and occupation contribute to risk factors for cardiovascular disease.   Am J Public Health. 1992;82(6):816-820. doi:10.2105/AJPH.82.6.816 PubMedGoogle ScholarCrossref
38.
Horvath  A, Greenberg  L.  Development and validation of the Working Alliance Inventory.   J Counseling Psychol. 1989;36(2):223-233. doi:10.1037/0022-0167.36.2.223 Google ScholarCrossref
39.
King  AC, Hekler  EB, Grieco  LA,  et al.  Harnessing different motivational frames via mobile phones to promote daily physical activity and reduce sedentary behavior in aging adults.   PLoS One. 2013;8(4):e62613. doi:10.1371/journal.pone.0062613 PubMedGoogle Scholar
40.
Kaji  AH, Lewis  RJ.  Noninferiority trials: is a new treatment almost as effective as another?   JAMA. 2015;313(23):2371-2372. doi:10.1001/jama.2015.6645 PubMedGoogle ScholarCrossref
41.
Sundquist  K, Qvist  J, Sundquist  J, Johansson  SE.  Frequent and occasional physical activity in the elderly: a 12-year follow-up study of mortality.   Am J Prev Med. 2004;27(1):22-27. doi:10.1016/j.amepre.2004.03.011 PubMedGoogle ScholarCrossref
42.
Kushi  LH, Fee  RM, Folsom  AR, Mink  PJ, Anderson  KE, Sellers  TA.  Physical activity and mortality in postmenopausal women.   JAMA. 1997;277(16):1287-1292. doi:10.1001/jama.1997.03540400037028 PubMedGoogle ScholarCrossref
43.
Albright  CL, Pruitt  L, Castro  C, Gonzalez  A, Woo  S, King  AC.  Modifying physical activity in a multiethnic sample of low-income women: one-year results from the IMPACT (Increasing Motivation for Physical ACTivity) project.   Ann Behav Med. 2005;30(3):191-200. doi:10.1207/s15324796abm3003_3 PubMedGoogle ScholarCrossref
44.
Finney  JW, Mitchell  RE, Cronkite  RC, Moos  RH.  Methodological issues in estimating main and interactive effects: examples from coping/social support and stress field.   J Health Soc Behav. 1984;25(1):85-98. doi:10.2307/2136706 PubMedGoogle ScholarCrossref
45.
West  BT, Welch  KB, Galecki  AT.  Linear Mixed Models: A Practical Guide Using Statistical Software. CRC Press; 2006. doi:10.1201/9781420010435
46.
Tan  WY, Tabatabai  MA.  A modified winsorized regression procedure for linear models.   J Statistical Computation and Simulation. 2007;30(4):299-313. doi:10.1080/00949658808811103 Google ScholarCrossref
47.
Little  RJA, Rubin  DB.  Statistical Analysis with Missing Data, Second Edition. John Wiley & Sons; 2002. doi:10.1002/9781119013563
48.
Hekler  EB, Buman  MP, Haskell  WL,  et al.  Reliability and validity of CHAMPS self-reported sedentary-to-vigorous intensity physical activity in older adults.   J Phys Act Health. 2012;9(2):225-236. doi:10.1123/jpah.9.2.225 PubMedGoogle ScholarCrossref
49.
Kallogjeri  D, Francis  DO, Piccirillo  JF.  Use of confidence intervals to improve the quality of research results reporting and interpretation.   JAMA Otolaryngol Head Neck Surg. 2019;145(6):491-493. doi:10.1001/jamaoto.2019.0793 PubMedGoogle ScholarCrossref
50.
King  AC, Castro  C, Wilcox  S, Eyler  AA, Sallis  JF, Brownson  RC.  Personal and environmental factors associated with physical inactivity among different racial-ethnic groups of US middle-aged and older-aged women.   Health Psychol. 2000;19(4):354-364. doi:10.1037/0278-6133.19.4.354 PubMedGoogle ScholarCrossref
51.
Wilcox  S, King  AC, Brassington  G, Ahn  D.  Physical activity preferences of middle-aged and older adults: a community analysis.   Journal Aging Physical Activity. 1999;7:386-399. doi:10.1123/japa.7.4.386 Google ScholarCrossref
52.
King  AC, Castro  CM, Buman  MP, Hekler  EB, Urizar  GG  Jr, Ahn  DK.  Behavioral impacts of sequentially versus simultaneously delivered dietary plus physical activity interventions: the CALM trial.   Ann Behav Med. 2013;46(2):157-168. doi:10.1007/s12160-013-9501-y PubMedGoogle ScholarCrossref
53.
Gothe  NP, Wójcicki  TR, Olson  EA,  et al.  Physical activity levels and patterns in older adults: the influence of a DVD-based exercise program.   J Behav Med. 2015;38(1):91-97. doi:10.1007/s10865-014-9581-6 PubMedGoogle ScholarCrossref
54.
Wijsman  CA, Westendorp  RG, Verhagen  EA,  et al.  Effects of a web-based intervention on physical activity and metabolism in older adults: randomized controlled trial.   J Med internet Res. 2013;15(11):e233. doi:10.2196/jmir.2843 PubMedGoogle Scholar
55.
Kurti  AN, Dallery  J.  Internet-based contingency management increases walking in sedentary adults.   J Appl Behav Anal. 2013;46(3):568-581. doi:10.1002/jaba.58 PubMedGoogle ScholarCrossref
56.
Alley  SJ, Kolt  GS, Duncan  MJ,  et al.  The effectiveness of a web 2.0 physical activity intervention in older adults - a randomised controlled trial.   Int J Behav Nutr Phys Act. 2018;15(1):4. doi:10.1186/s12966-017-0641-5 PubMedGoogle ScholarCrossref
57.
King  AC, Hekler  EB, Grieco  LA,  et al.  Effects of three motivationally targeted mobile device applications on initial physical activity and sedentary behavior change in midlife and older adults: a randomized trial.   PLoS One. 2016;11(6):e0156370. doi:10.1371/journal.pone.0156370 PubMedGoogle Scholar
58.
Colón-Semenza  C, Latham  NK, Quintiliani  LM, Ellis  TD.  Peer coaching through mhealth targeting physical activity in people with parkinson disease: Feasibility Study.   JMIR Mhealth Uhealth. 2018;6(2):e42. doi:10.2196/mhealth.8074 PubMedGoogle Scholar
59.
Victorson  D, Banas  J, Smith  J,  et al.  eSalud: designing and implementing culturally competent ehealth research with Latino patient populations.   Am J Public Health. 2014;104(12):2259-2265. doi:10.2105/AJPH.2014.302187 PubMedGoogle ScholarCrossref
60.
López  L, Tan-McGrory  A, Horner  G, Betancourt  JR.  Eliminating disparities among Latinos with type 2 diabetes: effective ehealth strategies.   J Diabetes Complications. 2016;30(3):554-560. doi:10.1016/j.jdiacomp.2015.12.003 PubMedGoogle ScholarCrossref
61.
Rodriquez  EJ, Pérez-Stable  EJ.  The time is now for ehealth research with Latinos.   Am J Public Health. 2017;107(11):1705-1707. doi:10.2105/AJPH.2017.304055 PubMedGoogle ScholarCrossref
62.
Stenholm  S, Vahtera  J, Kawachi  I,  et al.  Patterns of weight gain in middle-aged and older US adults, 1992-2010.   Epidemiology. 2015;26(2):165-168. doi:10.1097/EDE.0000000000000228 PubMedGoogle ScholarCrossref
63.
Bikdeli  B, Welsh  JW, Akram  Y,  et al.  Noninferiority designed cardiovascular trials in highest-impact journals.   Circulation. 2019;140(5):379-389. doi:10.1161/CIRCULATIONAHA.119.040214 PubMedGoogle ScholarCrossref
64.
Wilcox  S, Dowda  M, Leviton  LC,  et al.  Active for life: final results from the translation of two physical activity programs.   Am J Prev Med. 2008;35(4):340-351. doi:10.1016/j.amepre.2008.07.001 PubMedGoogle ScholarCrossref
65.
Wilcox  S, Dowda  M, Wegley  S, Ory  MG.  Maintenance of change in the Active-for-Life initiative.   Am J Prev Med. 2009;37(6):501-504. doi:10.1016/j.amepre.2009.07.016 PubMedGoogle 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
    Original Investigation
    September 28, 2020

    Effects of Counseling by Peer Human Advisors vs Computers to Increase Walking in Underserved Populations: The COMPASS Randomized Clinical Trial

    Author Affiliations
    • 1Department of Epidemiology & Population Health, Stanford University School of Medicine, Stanford, California
    • 2Stanford Prevention Research Center, Department of Medicine, Stanford University School of Medicine, Stanford, California
    • 3Now with Global Community Health and Behavioral Science Department, Tulane University School of Public Health and Tropical Medicine, New Orleans, Louisiana
    • 4Now with Omada Health, Inc, San Francisco, California
    • 5Now with Division of Primary Care and Population Health, Department of Medicine, Stanford University School of Medicine, Stanford, California
    • 6Now with Primary Care, Veterans Affairs Palo Alto Health Care System, Livermore, California
    • 7Now with Fair Oaks Health Center, San Mateo County Health System, Redwood City, California
    • 8Now with Department of Health and Kinesiology, Purdue University, West Lafayette, Indiana
    • 9Khoury College of Computer Sciences, Northeastern University, Boston, Massachusetts
    JAMA Intern Med. 2020;180(11):1481-1490. doi:10.1001/jamainternmed.2020.4143
    Key Points

    Question  Can customized counseling by computer achieve weekly walking increases among low-income older adults to an extent similar to those achieved by trained human advisors?

    Findings  In this cluster-randomized noninferiority trial of 245 adults (241 Latino, 98.4%) 50 years or older, the computer advisor intervention achieved significant 12-month increases in weekly walking levels that were no worse than those achieved by trained peer advisors. In addition, the increases were accompanied by similar improvements in relevant clinical risk factors.

    Meaning  The findings of this trial showing weekly walking increases attained in both the computer and human intervention groups of underserved older adults provide support for broadening the range of light-touch physical activity programs that can be offered to inactive older adults.

    Abstract

    Importance  Effective and practical treatments are needed to increase physical activity among those at heightened risk from inactivity. Walking represents a popular physical activity that can produce a range of desirable health effects, particularly as people age.

    Objective  To test the hypothesis that counseling by a computer-based virtual advisor is no worse than (ie, noninferior to) counseling by trained human advisors for increasing 12-month walking levels among inactive adults.

    Design, Setting, and Participants  A cluster-randomized, noninferiority parallel trial enrolled 245 adults between July 21, 2014, and July 29, 2016, with follow-up through September 15, 2017. Data analysis was performed from March 15 to December 20, 2018. The evidence-derived noninferiority margin was 30 minutes of walking per week. Participants included inactive adults aged 50 years and older, primarily of Latin American descent and capable of walking without significant limitations, from 10 community centers in Santa Clara and San Mateo counties, California.

    Interventions  All participants received similar evidence-based, 12-month physical activity counseling at their local community center, with the 10 centers randomized to a computerized virtual advisor program (virtual) or a previously validated peer advisor program (human).

    Main Outcomes and Measures  The primary outcome was change in walking minutes per week over 12 months using validated interview assessment corroborated with accelerometry. Both per-protocol and intention-to-treat analysis was performed.

    Results  Among the 245 participants randomized, 193 were women (78.8%) and 241 participants (98.4%) were Latino. Mean (SD) age was 62.3 (8.4) years (range, 50-87 years), 107 individuals (43.7%) had high school or less educational level, mean BMI was 32.8 (6.8), and mean years residence in the US was 47.4 (17.0) years. A total of 231 participants (94.3%) completed the study. Mean 12-month change in walking was 153.9 min/wk (95% CI, 126.3 min/wk to infinity) for the virtual cohort (n = 123) and 131.9 min/wk (95% CI, 101.4 min/wk to infinity) for the human cohort (n = 122) (difference, 22.0, with lower limit of 1-sided 95% CI, −20.6 to infinity; P = .02); this finding supports noninferiority. Improvements emerged in both arms for relevant clinical risk factors, sedentary behavior, and well-being measures.

    Conclusions and Relevance  The findings of this study indicate that a virtual advisor using evidence-based strategies produces significant 12-month walking increases for older, lower-income Latino adults that are no worse than the significant improvements achieved by human advisors. Changes produced by both programs are commensurate with those reported in previous investigations of these behavioral interventions and provide support for broadening the range of light-touch physical activity programs that can be offered to a diverse population.

    Trial Registration  ClinicalTrials.gov Identifier: NCT02111213

    Introduction

    Physical inactivity is associated with approximately $117 billion per year in US health care costs.1,2 While national physical activity (PA) guidelines emphasize convenient aerobic activities, such as walking,3 fewer than half of adults meet these guidelines.4 Older, low-income, and Latino adults are at particular risk for inactivity owing to factors including reduced access to convenient, customized programs addressing PA barriers4; consequently, these groups have high rates of obesity and other metabolic conditions.5

    Technology-enabled eHealth programs represent potentially cost-efficient and practical means for customized PA guidance to diverse groups.4 Most people targeted by eHealth, however, are well educated, younger than 50 years, and of non-Hispanic White ancestry,4 potentially intensifying health disparities.6

    This investigation tested whether a virtual advisor could increase 12-month walking to an extent similar to a comparably structured human advisor program among Latino adults. The human advisor program was delivered by trained peer advisors—a resource-efficient approach that is well accepted by Latino and other diverse groups7,8 but may be less convenient and scalable than computer-based programs. Randomized clinical trials and translational PA research demonstrating peer/volunteer program effectiveness include over 700 adults from 14 communities, with effect sizes comparable to randomized clinical trials using professional health advisors (eMethods Supplement 1).7,8 Both interventions were considered “light-touch” given that they were delivered using primarily non–health professional staff and resources.9

    Methods
    Design

    COMPASS (Computerized Physical Activity Support for Seniors) was a single-blind, cluster-randomized noninferiority parallel trial10 conducted by Stanford University and Northeastern University. The trial protocol is available in Supplement 2. This study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline. The trial was approved by the Stanford University Institutional Review Board. Participants provided written informed consent in their preferred language (English or Spanish). They received a $10 gift card/assessment. Such modest remunerations have been found to have minimal influence on PA change.4

    Participants were recruited from community centers in Santa Clara and San Mateo counties, California. The centers were randomized in pairs (1:1 allocation) based on locale to either virtual or human advisors based on a computerized randomization sequence (SAS PROC Plan statistical software, version 15.1, SAS Institute Inc).11 Allocation concealment was accomplished by using staff not involved in study procedures. Both arms received a similar 12-month behavioral PA instruction/support program at their designated center based on Active Choices—an individually tailored program with demonstrated effectiveness and translatability7,8,12 that has merited formal recognition by the Administration on Aging, National Council on Aging, and the Centers for Disease Control and Prevention (eMethods in Supplement 1). Study methods were published previously.13

    Participants

    Participants were enrolled from July 21, 2014, to July 29, 2016, at community centers using geographically defined targeted mailings, cultural media-based promotion, and community outreach.13 Follow-up ended September 15, 2017.

    Participants completed assessment instruments. The following eligibility criteria were used13: (1) age 50 years or older; (2) insufficiently active14 (ie, engaged in <100 min/wk of moderate-intensity PA over the past month, based on the Community Healthy Activities Model Program for Seniors [CHAMPS] instrument)15; (3) able to safely engage in moderate-intensity PA, such as walking, based on the PA Readiness Questionnaire16; (4) live less than 5 miles from a study-designated community center; (5) be able to read and understand English or Spanish sufficiently to provide informed consent and participate in study procedures, including computer use; and (6) plan to live in the area for a year.

    Interventions and Advisors

    The interventions focused on walking and similar moderate-intensity PA that have demonstrated health benefits.2 Community centers were used for delivery given their appeal and use by Latino populations17-20 (eMethods in Supplement 1). Sessions occurred 1:1 in a private space within each center. In an introductory session, national guidelines and safety tips were discussed and each participant’s PA history and goals were reviewed. As has become usual practice in PA promotion, participants were encouraged to use a pedometer provided as an additional intervention tool and received a calendar to note advisor appointments, log steps, and record walking minutes.13

    Participants then received up to 28 brief (10-15 minutes) advising sessions across 12 months. Sessions followed a standard, evidence-based protocol7,12,13 (eMethods in Supplement 1) and allowed for customization based on participant preferences and availability. Each session consisted of an introductory dialogue, brief check for health changes, review of pedometer steps and minutes walked since last session, problem-solving around personal PA barriers, goal-setting, and next-session scheduling. Cultural tailoring occurred in a variety of ways (eMethods in Supplement 1).

    The virtual advisor (Figure 1), described elsewhere,13,21 was an interactive, animated computer agent simulating face-to-face counseling using simple speech (synthetic English or Spanish) and nonverbal behaviors (eg, facial cues and hand gestures).22,23 The virtual agent, named Carmen, was successful previously in increasing 4-month walking levels among Latino aging adults relative to controls.21 Individuals interacted with Carmen by touching simple conversation boxes on the computer screen offered in English and Spanish and targeted at a first- to third-grade reading level.23-25 Participants received a private log-in and were encouraged to wear headphones for privacy. They also were encouraged to download their pedometer data on the virtual computer at each session via a USB port. One difference between this system and some other automated health information systems is that participants could interact directly with this system at their convenience, receiving real-time customized advice through the multi-modality visual plus auditory interface21,23 (eMethods in Supplement 1).

    Criteria for human advisors are described in the eMethods in Supplement 1. Volunteers (n = 31) were community residents aged 30 years and older (mean [SD] age, 60.0 [14.1] years) who engaged in regular PA. Eligible individuals underwent 12 hours of Active Choices program training commensurate with previous studies.7,26 Advisors received modest remuneration for the general time commitment accompanying these activities and any travel costs to centers (eMethods in Supplement 1). Advisor oversight and quality assurance were accomplished through monthly supervision meetings, periodic review of advisor postsession logs, and random staff-participant check-ins.13

    Outcome Assessments

    Participants completed primary assessments at their designated center at baseline and 12 months, with a briefer, interim assessment at 6 months. Assessment staff were blinded to center and participant treatment assignment and previous data.

    Change in walking minutes per week was assessed using the 4 walking items from the validated CHAMPS survey (interview format) for older adults (Table 1), which is available in English and Spanish.15,27,28 Such standardized self-report instruments represent the most direct, reliable, and cost-efficient means for assessing specific PA types typically targeted in community interventions, given that device-based assessment tools (accelerometers) capture more general movement beyond such purposeful PA behavior29,30 and currently lack sufficient normative data, especially among older adults, to allow direct linkages with the PA guidelines and evidence base.4 Accelerometry also can underestimate movement, may be less sensitive to change in older and very inactive populations,31 and was found in one investigation to be burdensome for this older Latino population.21 CHAMPS is the only PA questionnaire for older adults shown to be correlated with both doubly-labeled, water-measured PA energy expenditure31 and accelerometers.31 Accelerometry data collection occurred at baseline and 12 months to provide a secondary source of corroboration of daily PA amounts and followed a standard protocol29,30,32,33 (eMethods in Supplement 1).

    Secondary outcomes included CHAMPS total PA and moderate to vigorous PA (MVPA) variables15; sedentary behavior, using a validated 1-week recall survey34; measured body mass index (BMI) (calculated as weight in kilograms divided by height in meters squared); resting blood pressure and heart rate28; and well-being, measured with the 10-item Vitality Plus Scale (scores range from 10 [low well-being] to 50 [high well-being]).35

    Sociodemographic variables were collected using standard questionnaires.36,37 Program safety and adverse events were tracked using standardized protocols.7,36 Twelve-month program satisfaction measures included the Working Alliance Inventory38 and user satisfaction survey.21,22,39 The mean number of advising sessions and total participant-advisor interaction time were captured automatically by the virtual advisor and through session logs kept by human advisors.

    Statistical Analysis

    Data analysis was performed from March 15 to December 20, 2018. A sample size of 226 (113 participants/arm) was calculated to provide 80% power to demonstrate noninferiority of virtual vs human counseling assuming 10% loss to follow-up.40 The effect size was defined as the standardized difference between the 2 interventions in 12-month change in walking minutes per week (Cohen d, with the distribution of the mean difference showing sufficient approximation of a normal distribution). The noninferiority margin was based on a clinically meaningful difference, in one direction, between arms of 30 minutes of walking per week,4,14,41,42 and a within-arm SD of 90, accounting for clustering within centers.43

    Descriptive statistics summarized baseline characteristics. Baseline between-arm differences were tested using independent-sample t tests for continuous variables and Pearson χ2 or exact tests for categorical variables. The primary noninferiority hypothesis was tested using a mixed-effects linear regression model (SAS, version 15.1). Change in 12-month walking minutes per week was the primary outcome, with centered arm assignment as the independent variable, and centered baseline walking value, the interaction between centered arm assignment and centered baseline value, community site, and sex as covariates.44 To account for the cluster randomization design, community center was included as a random intercept term.45 Extreme values of the 12-month outcomes were limited by winsorization to 3 SDs above the mean.46 To account for missing data, multiple imputations were performed by replacing missing 12-month values with a set of plausible values using the option of imputation by fully conditional specification methods (10 imputations were done). Imputation results were then combined.11,47 All reported outcomes were analyzed based on intention-to-treat principles. Using a 1-sided 95% CI constructed for the between-group difference in the primary outcome, noninferiority was deemed to be demonstrated if the lower limit of the CI lay to the right of the noninferiority margin of −0.30.40 For completeness, both intention-to-treat and per-protocol analyses are reported herein.40 Intervention participation variables were compared using t tests. Within-arm pre-post t tests and between-arm 12-month testing of specific secondary outcomes (eg, program satisfaction) were undertaken for descriptive purposes. The significance threshold was α = .05, using 2-tailed tests. Several accelerometry-based corroboratory analyses were conducted,48 taking into account data skewness in this sedentary sample (eMethods in Supplement 1). These analyses included evaluation of median 12-month changes in accelerometry-derived step counts using Wilcoxon signed rank tests, and associations between 12-month reported MVPA and accelerometry-derived MVPA and step counts using Spearman-rank order correlations.

    Results

    Trial flow is shown in Figure 2. As noted, soon after being recruited for the study and before participant enrollment, one center (randomized to the human advisor arm) experienced unforeseen changes in its administration that disrupted center operations and precluded study participation. To ensure comparable participant enrollment in each study arm, additional participants were enrolled at the remaining 4 centers assigned to that arm. A total of 94.3% of the participants (231/245) (virtual advisor, 95.1%; human advisor, 93.4%) completed 12-month data collection for walking minutes per week. Noncompleters did not differ significantly from completers on baseline demographic or CHAMPS variables, with the exception of baseline minutes of walking briskly, which showed minor differences (virtual advisor: mean minutes per week for noncompleters, 0.0 vs completers, 2.9; P = .003; human advisor: mean minutes per week for noncompleters, 0.0 vs completers, 2.6; P = .009).

    Baseline participant characteristics were similar between the 2 arms (Table 1), with the exception of Latin American country of origin. A total of 241 participants (98.4%) were Latino, 193 participants were women (78.8%), and 52 participants were men (21.2%). Mean (SD) age was 62.3 (8.4) years (range, 50-87 years). One hundred seven participants (43.7%) had an education level of high school or lower. Mean (SD) years of residence in the US was 47.4 (17.0) years. Mean (SD) BMI was 32.8 (6.8), with 156 participants (63.7%) in the obese range. One hundred nineteen participants (48.6%) were taking antihypertensive medications. Reported walking time was low (mean [SD] total, 70 [98] min/wk; approximately 10-minute/d). Baseline accelerometry-derived MVPA per week also was low (mean [SD] virtual advisor, 40.7 [25.8]; human advisor, 43.6 [31.2]). One hundred six participants (43.3%) chose to receive their intervention in Spanish (virtual advisor, 48 [39.0%]; human advisor, 58 [47.5%]) (χ2 = 1.81, P = .18).

    The mean (SD) number of total advising sessions completed for virtual vs human advisors was 18.8 (11.8) (64.8%) vs 18.4 (8.9) (63.4%) (t test, 0.30; P = .76). Mean length of postintroductory advising sessions was significantly shorter in virtual (8.4 [4.8] min/session) vs human (21.0 [7.1] min/session) (t test, 16.0, P < .001). This finding translates into a mean total intervention volume for virtual vs human of 3.2 hours (56.6) vs 6.9 hours (63.0) (t test, 27.8; P < .001). Although both programs encouraged pedometer use and reporting at each session, pedometer reporting was sporadic in both programs across 12 months (eg, month 12 reporting for virtual advisor, 35.8% vs human advisor, 32.8%; χ2 = 0.24; P = .62) (eTable 1 in Supplement 1).

    Outcomes

    Twelve-month change in walking minutes per week, summarized in Table 2 and Figure 3, supported noninferiority (ie, lower limit of the 95% CI [−20.6] lay to the right of the noninferiority margin [−30]). Mean walking increases indicated a somewhat larger increase in the virtual advisor cohort (153.9 min/wk; 95% CI, 126.3 min/wk to infinity vs 131.9 min/wk; 95% CI, 101.4 min/wk to infinity; difference, 22.0, with lower limit of 1-sided 95% CI, −20.6 to infinity; P = .02). Similar results were obtained using per-protocol analyses (mean [SD] increases in walking time for the virtual advisor cohort [n = 117]: 158.6 [217.1] min/wk vs human advisor cohort [n = 114]: 134.8 [192.1] min/wk; P = .02). At baseline, no participant was at the nationally recommended target range of 150 min/wk or more of MVPA3; at 12 months, 29.3% of the virtual advisor and 31.1% of the human advisor cohorts achieved that target (χ2 = 0.10, P = .75).

    Corroborative evidence of PA levels was indicated by 12-month changes in accelerometry-derived mean steps per day, correlations between self-reported and accelerometry-derived PA, and accelerometry comparisons at 12 months between participants reporting PA levels above or below the sample median (eResults in Supplement 1). Median step count increases per day (first to third interquartile range [IQR]) equaled 539.8/d (IQR, −327 to 1600) for the virtual advisor cohort and 304.2/d (IQR, −987 to 1245) for the human advisor cohort (Wilcoxon 2-sample rank test, P = .08). Twelve-month Spearman correlations were significant between reported MVPA and accelerometry-derived MVPA (r = 0.24, P < .001), and mean step counts per day (r = 0.25, P < .001).

    Mean between-arm changes and 95% CIs49 for measured clinical risk factors are reported in Table 2. Intention-to-treat analysis for the virtual advisor cohort identified statistically significant within-arm reductions of diastolic blood pressure (2.3%), heart rate (2.5%), and BMI (2.6%). For the human advisor cohort, significant reductions were identified for systolic blood pressure (2.8%), diastolic blood pressure (3.0%), and BMI (2.7%). For both arms, significant decreases at 12 months were noted for total reported sedentary time (virtual advisor: P = .03 vs human advisor: P = .02) and time watching television/videos virtual advisor: P < .001 vs human advisor: P = .01) (Table 2).

    Both arms reported improvements in overall well-being35 (Table 2), with the magnitude of the within-arm improvement somewhat larger in the human advisor (mean [SE] change, 3.5 [0.6]; t = 4.1; P < .001) compared with the virtual advisor (mean [SE] change, 1.1 [0.6]; t = 1.8; P = .06) cohorts. The human advisor cohort reported significant pre-post improvements in all 10 domains (eg, good appetite and few aches or pain), while the virtual advisor cohort reported significant improvements in 3 domains (sleep well, feel rested, and full of pep and energy) (eTable 2 in Supplement 1).

    Program Satisfaction and Safety

    Twelve-month program satisfaction measures indicated no significant between-arm differences in participants’ comfort with their advisor and overall helpfulness of advisor information, including reaching PA goals (eTable 3 in Supplement 1). Overall ratings of advisor support from the Working Alliance Inventory (bonding subscale) were significantly higher in the human vs virtual advisor cohort (difference, −0.89; 95% CI, −1.28 to −0.50; t = −4.53; P < .001), although, compared with the virtual advisor cohort, human advisor participants indicated that their advisor sessions were too long (difference, −0.52; 95% CI, −0.89 to −0.17; t = 2.9; P = .005). The longer human advisor session length was expected, given that human advisors were able to engage in broader conversational dialogues compared with the virtual advisors.

    Technical problems with the virtual advisor were infrequent and resolved remotely by Northeastern University in partnership with Stanford University. Common problems accompanying software upgrades included simple system shutdowns, graphical issues, and data reading errors. These issues were resolved once reported and did not interfere unduly with the intervention.

    No serious adverse events or deaths were reported for either arm. Seven participants (virtual advisor, 3; human advisor, 4) experiencing milder events (eg, leg pain/discomfort) remained in the study, and all were able to resume PA following the reported event.

    Discussion

    The results of this study broaden evidence-based alternatives for PA advising for older inactive populations. While structured group-based PA programs remain a popular offering throughout the US, more than half of aging adults prefer undertaking all or much of their PA in another format,36,50,51 specifically, “on one’s own with some instruction.”50[p354] A recurring challenge has been finding acceptable alternatives to professionally provided PA instruction/support deliverable in a sustainable fashion. The computer and peer programs tested represent 2 such alternatives. Both programs produced meaningful walking increases—a health-enhancing, convenient, and widely appealing PA.3 Both interventions also have been delivered successfully to older populations through remote channels (eg, telephone and home computer),7,23,52 of relevance to coronavirus disease 2019 and similar outbreaks.

    The virtual advisor program proved to be not inferior to the human advisor program in achieving this outcome and used less advisor time. The mean walking increase achieved with the virtual advisor intervention was virtually identical to that found in the earlier pilot experiment that included a no-treatment control group.21 The results of the present trial expand the small PA eHealth evidence base for older adults53-58 and Latino populations59-61 across longer time frames. Both programs also produced reported decreases in prevalent sedentary behaviors, including television/video viewing, that have been linked independently with detrimental health outcomes.3,4

    Although no participant was initially at the federally recommended PA target range,3 approximately 30% in both programs achieved this level at 12 months. While reaching this target range can provide many health benefits, evidence underscores that inactive individuals and older adults can benefit from even small PA increases.4

    The sample had mean baseline BMI levels in the class I obesity range and a 12-month 2.6% BMI reduction across both arms, representing a weight loss of approximately 2.3 kg—an amount supported by national evidence reviews of PA alone.14 Given that US older adults typically experience a yearly weight gain,62 the results indicate that such weight gain may be curtailed through reasonable increases in walking and other moderate PA outside of formal weight control instruction.3

    Both interventions resulted in improvements in well-being, although these improvements were more robust in the human advisor program. These results parallel those from a 12-month PA trial comparing telephone instruction delivered by human advisors vs an automated telehealth system.12 Given the importance of well-being to quality of life and continued PA participation, such differences deserve further investigation.

    Strengths and Limitations

    Among study strengths are a population at risk for various negative outcomes that can be ameliorated by PA, but which has rarely been targeted for PA programming; head-to-head comparison of 2 modes of customized intervention delivery with potentially substantial reach into diverse population segments and minimal reliance on health professionals9; and high retention (>90%) and related methodologic features recommended for noninferiority trials.63 Along with clinically measured weight loss and small blood pressure decreases, accelerometry measurement provided corroborative evidence of PA levels.48

    The trial had limitations. Because of the noninferiority design, which lacked a control arm, it is possible that, although the walking increases achieved were not poorer in the virtual advisor arm, neither arm was successful. This supposition is diminished given that the 12-month increases observed in both arms are commensurate with the moderate or higher effect size estimates achieved across 3 decades of Active Choices randomized clinical trials and translation research.4,7,12,64 Although preliminary pilot-testing of virtual advisor acceptability by older adults of other ethnicities (ie, Filipino, African American, and Asian) suggested that intervention using the virtual advisor program was acceptable, further formal investigation is required. Other areas meriting study include determining cost-effectiveness, investigating peer advising implementation and effectiveness issues, and evaluating behavioral maintenance following program completion. An Active Choices translation trial found that increased PA levels from a 6-month program were maintained through a 6-month follow-up period,65 similar to those found in the original virtual advisor study.21

    Conclusions

    The findings of this trial show that, among inactive older Latino adults, a customized, computer-driven virtual advisor produced meaningful 12-month improvements in weekly walking and similar moderate-intensity activities that were not inferior to the significant improvements achieved by human advisors. The results expand the evidence-based PA program choices that have the potential to reach more diverse and underrepresented population segments.

    Back to top
    Article Information

    Accepted for Publication: July 8, 2020.

    Corresponding Author: Abby C. King, PhD, Department of Epidemiology & Population Health, Stanford University School of Medicine, 259 Campus Dr, HRP-Redwood Building, Room T221, Stanford, CA 94305 (king@stanford.edu).

    Published Online: September 28, 2020. doi:10.1001/jamainternmed.2020.4143

    Author Contributions: Drs King and Ahn had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: King, Castro Sweet, Hauser, Blanco, Bickmore.

    Acquisition, analysis, or interpretation of data: King, Campero, Sheats, Castro Sweet, Hauser, Garcia, Chazaro, Banda, Ahn, Fernandez, Bickmore.

    Drafting of the manuscript: King, Hauser, Banda, Bickmore.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: Ahn, Bickmore.

    Obtained funding: King, Bickmore.

    Administrative, technical, or material support: King, Campero, Sheats, Castro Sweet, Garcia, Chazaro, Blanco, Banda, Fernandez, Bickmore.

    Supervision: King, Campero, Castro Sweet.

    Conflict of Interest Disclosures: Dr King reported receiving grants from the National Institutes of Health (NIH) during the conduct of the study. Ms Campero, Dr Castro Sweet, Ms Garcia, and Mr Chazaro reported receiving grant funding as paid laboratory staff from the NIH during the conduct of the study. No other disclosures were reported.

    Funding/Support: This investigation was supported by Public Health Service grant 5R01HL116448 from the National Heart, Lung, & Blood Institute (Dr King). Dr King also received partial support from US Public Health Services grants 5R01CA211048 and P20CA217199 from the National Cancer Institute, US Public Health Service grant 1U54EB020405 supporting the National Center for Mobility Data Integration and Insight (PI: S. Delp), partial funding from the Catalyst program at Stanford University (PI: S. Delp), a grant from the Discovery Innovation Fund in Basic Biomedical Sciences from Stanford University, the Nutrilite Health Institute Wellness Fund provided by Amway to the Stanford Prevention Research Center, and US Public Health Service grant 1U54MD010724 (PI: M. Cullen). Drs Sheats and Hauser received support from NIH training grant 5T32HL007034-39 (PI: C. Gardner).

    Role of the Funder/Sponsor: The funding organizations 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.

    Additional Contributions: We acknowledge the overall assistance and support of the study peer advisors, participants, staff, and volunteers of participating community centers in Santa Clara and San Mateo Counties, California, and the following individuals in project activities or methods input. Methods input: Michael Baiocchi, PhD, Laurence Baker, PhD, William Haskell, PhD, John Ioannidis, MD, PhD, David Maron, MD, Randall Stafford, MD, PhD, and Sandra Winter, PhD (Stanford University School of Medicine). Data collection activities: Betsy Caravalho, BS, Gustavo Chavez, MD, Ana Cortes, BA, Michele Escobar, BA, Darlyne Esparza, BS, Lidya Esparza-Rivera, BA, Martha Gabaray, BS, Ruby Gonzalez, BA, Cain Murguía, BA, Selene Virgen, BA, Daniel Vuong, BS (Stanford University School of Medicine), and Ha Trinh, PhD, and Langxuan (James) Yin, PhD (Northeastern University). The above-named individuals received no compensation outside of salary.

    Data Sharing Statement: See Supplement 3.

    References
    1.
    Guthold  R, Stevens  GA, Riley  LM, Bull  FC.  Worldwide trends in insufficient physical activity from 2001 to 2016: a pooled analysis of 358 population-based surveys with 1·9 million participants.   Lancet Glob Health. 2018;6(10):e1077-e1086. doi:10.1016/S2214-109X(18)30357-7 PubMedGoogle ScholarCrossref
    2.
    Piercy  KL, Troiano  RP, Ballard  RM,  et al.  The Physical Activity Guidelines for Americans.   JAMA. 2018;320(19):2020-2028. doi:10.1001/jama.2018.14854 PubMedGoogle ScholarCrossref
    3.
    US Dept of Health & Human Services.  Physical Activity Guidelines for Americans, 2nd edition. US Dept of Health & Human Services;2018.
    4.
    Physical Activity Guidelines Advisory Committee.  2018 Physical Activity Guidelines Advisory Committee Scientific Report. US Dept of Health & Human Services;2018.
    5.
    US Dept of Health & Human Services. HHS action plan to reduce racial and ethnic disparities: a nation free of disparities in health and health care. US Dept of Health & Human Services; April 2011.
    6.
    Latulippe  K, Hamel  C, Giroux  D.  Social health inequalities and eHealth: a literature review with qualitative synthesis of theoretical and empirical studies.   J Med internet Res. 2017;19(4):e136. doi:10.2196/jmir.6731 PubMedGoogle Scholar
    7.
    Castro  CM, Pruitt  LA, Buman  MP, King  AC.  Physical activity program delivery by professionals versus volunteers: the TEAM randomized trial.   Health Psychol. 2011;30(3):285-294. doi:10.1037/a0021980 PubMedGoogle ScholarCrossref
    8.
    Hooker  SP, Seavey  W, Weidmer  CE,  et al.  The California active aging community grant program: translating science into practice to promote physical activity in older adults.   Ann Behav Med. 2005;29(3):155-165. doi:10.1207/s15324796abm2903_1 PubMedGoogle ScholarCrossref
    9.
    MacNab  M, Lee  SH, McCloughan  L, Hanley  J, McKinstry  B, Pinnock  H.  Oximetry-supported self-management for chronic obstructive pulmonary disease: mixed method feasibility pilot project.   BMC Health Serv Res. 2015;15:485. doi:10.1186/s12913-015-1135-2 PubMedGoogle ScholarCrossref
    10.
    Lauer  MS, Gordon  D, Wei  G, Pearson  G.  Efficient design of clinical trials and epidemiological research: is it possible?   Nat Rev Cardiol. 2017;14(8):493-501. doi:10.1038/nrcardio.2017.60 PubMedGoogle ScholarCrossref
    11.
    SAS Institute Inc.  SAS/STAT User’s Guide: Version 15.1. SAS Institute, Inc; 2018.
    12.
    King  AC, Friedman  R, Marcus  B,  et al.  Ongoing physical activity advice by humans versus computers: the Community Health Advice by Telephone (CHAT) trial.   Health Psychol. 2007;26(6):718-727. doi:10.1037/0278-6133.26.6.718 PubMedGoogle ScholarCrossref
    13.
    King  AC, Campero  I, Sheats  JL,  et al.  Testing the comparative effects of physical activity advice by humans vs computers in underserved populations: the COMPASS trial design, methods, and baseline characteristics.   Contemp Clin Trials. 2017;61(October):115-125. doi:10.1016/j.cct.2017.07.020 PubMedGoogle ScholarCrossref
    14.
    Physical Activity Guidelines Advisory Committee.  Report of the Physical Activity Guidelines Advisory Committee, 2008. Washington, DC: US Dept of Health & Human Services;2008.
    15.
    Stewart  AL, Mills  KM, King  AC, Haskell  WL, Gillis  D, Ritter  PL.  CHAMPS physical activity questionnaire for older adults: outcomes for interventions.   Med Sci Sports Exerc. 2001;33(7):1126-1141. doi:10.1097/00005768-200107000-00010 PubMedGoogle ScholarCrossref
    16.
    Thomas  S, Reading  J, Shephard  RJ.  Revision of the Physical Activity Readiness Questionnaire (PAR-Q).   Can J Sport Sci. 1992;17(4):338-345.PubMedGoogle Scholar
    17.
    Hernández-Nieto  R, Gutiérrez  MC, Moreno-Fernández  F.  Hispanic Map of the United States—2017. Harvard University;2017. doi:10.15427/OR035-11/2017EN
    18.
    Alcalá  HE, Chen  J, Langellier  BA, Roby  DH, Ortega  AN.  Impact of the Affordable Care Act on health care access and utilization among Latinos.   J Am Board Fam Med. 2017;30(1):52-62. doi:10.3122/jabfm.2017.01.160208 PubMedGoogle ScholarCrossref
    19.
    Ayala  GX, Vaz  L, Earp  JA, Elder  JP, Cherrington  A.  Outcome effectiveness of the lay health advisor model among Latinos in the United States: an examination by role.   Health Educ Res. 2010;25(5):815-840. doi:10.1093/her/cyq035 PubMedGoogle ScholarCrossref
    20.
    Pahor  M, Blair  SN, Espeland  M,  et al; LIFE Study Investigators.  Effects of a physical activity intervention on measures of physical performance: results of the Lifestyle Interventions and Independence for Elders Pilot (LIFE-P) study.   J Gerontol A Biol Sci Med Sci. 2006;61(11):1157-1165. doi:10.1093/gerona/61.11.1157 PubMedGoogle ScholarCrossref
    21.
    King  AC, Bickmore  TW, Campero  MI, Pruitt  LA, Yin  JL.  Employing virtual advisors in preventive care for underserved communities: results from the COMPASS study.   J Health Commun. 2013;18(12):1449-1464. doi:10.1080/10810730.2013.798374 PubMedGoogle ScholarCrossref
    22.
    Bickmore  T, Gruber  A, Picard  R.  Establishing the computer-patient working alliance in automated health behavior change interventions.   Patient Educ Couns. 2005;59(1):21-30. doi:10.1016/j.pec.2004.09.008 PubMedGoogle ScholarCrossref
    23.
    Bickmore  TW, Pfeifer  LM, Byron  D,  et al.  Usability of conversational agents by patients with inadequate health literacy: evidence from two clinical trials.   J Health Commun. 2010;15(suppl 2):197-210. doi:10.1080/10810730.2010.499991 PubMedGoogle ScholarCrossref
    24.
    Kincaid  JP, Braby  R, Mears  J.  Electronic authoring and delivery of technical information.   J Instructional Development. 1988;11(2):8-13. doi:10.1007/BF02904998 Google ScholarCrossref
    25.
    Fitzsimmons  PR, Michael  BD, Hulley  JL, Scott  GO.  A readability assessment of online Parkinson’s disease information.   J R Coll Physicians Edinb. 2010;40(4):292-296. doi:10.4997/JRCPE.2010.401 PubMedGoogle ScholarCrossref
    26.
    Wilcox  S, Dowda  M, Griffin  SF,  et al.  Results of the first year of active for life: translation of 2 evidence-based physical activity programs for older adults into community settings.   Am J Public Health. 2006;96(7):1201-1209. doi:10.2105/AJPH.2005.074690 PubMedGoogle ScholarCrossref
    27.
    Hekler  EB, Buman  MP, Ahn  D, Dunton  G, Atienza  AA, King  AC.  Are daily fluctuations in perceived environment associated with walking?   Psychol Health. 2012;27(9):1009-1020. doi:10.1080/08870446.2011.645213 PubMedGoogle ScholarCrossref
    28.
    Pahor  M, Guralnik  JM, Ambrosius  WT,  et al; LIFE study investigators.  Effect of structured physical activity on prevention of major mobility disability in older adults: the LIFE study randomized clinical trial.   JAMA. 2014;311(23):2387-2396. doi:10.1001/jama.2014.5616 PubMedGoogle ScholarCrossref
    29.
    King  AC, Sallis  JF, Frank  LD,  et al.  Aging in neighborhoods differing in walkability and income: associations with physical activity and obesity in older adults.   Soc Sci Med. 2011;73(10):1525-1533. doi:10.1016/j.socscimed.2011.08.032 PubMedGoogle ScholarCrossref
    30.
    Troiano  RP, McClain  JJ, Brychta  RJ, Chen  KY.  Evolution of accelerometer methods for physical activity research.   Br J Sports Med. 2014;48(13):1019-1023. doi:10.1136/bjsports-2014-093546 PubMedGoogle ScholarCrossref
    31.
    Colbert  LH, Matthews  CE, Havighurst  TC, Kim  K, Schoeller  DA.  Comparative validity of physical activity measures in older adults.   Med Sci Sports Exerc. 2011;43(5):867-876. doi:10.1249/MSS.0b013e3181fc7162 PubMedGoogle ScholarCrossref
    32.
    Evenson  KR, Wen  F, Herring  AH.  Associations of accelerometry-assessed and self-reported physical activity and sedentary behavior with all-cause and cardiovascular mortality among US adults.   Am J Epidemiol. 2016;184(9):621-632. doi:10.1093/aje/kww070 PubMedGoogle ScholarCrossref
    33.
    Evenson  KR, Wen  F, Metzger  JS, Herring  AH.  Physical activity and sedentary behavior patterns using accelerometry from a national sample of United States adults.   Int J Behav Nutr Phys Act. 2015;12:20. doi:10.1186/s12966-015-0183-7 PubMedGoogle ScholarCrossref
    34.
    Gardiner  PA, Clark  BK, Healy  GN, Eakin  EG, Winkler  EAH, Owen  N.  Measuring older adults’ sedentary time: reliability, validity, and responsiveness.   Med Sci Sports Exerc. 2011;43(11):2127-2133. doi:10.1249/MSS.0b013e31821b94f7 PubMedGoogle ScholarCrossref
    35.
    Myers  AM, Malott  OW, Gray  E,  et al.  Measuring accumulated health-related benefits of exercise participation for older adults: the Vitality Plus Scale.   J Gerontol A Biol Sci Med Sci. 1999;54(9):M456-M466. doi:10.1093/gerona/54.9.M456 PubMedGoogle ScholarCrossref
    36.
    King  AC, Haskell  WL, Taylor  CB, Kraemer  HC, DeBusk  RF.  Group- vs home-based exercise training in healthy older men and women: a community-based clinical trial.   JAMA. 1991;266(11):1535-1542. doi:10.1001/jama.1991.03470110081037 PubMedGoogle ScholarCrossref
    37.
    Winkleby  MA, Jatulis  DE, Frank  E, Fortmann  SP.  Socioeconomic status and health: how education, income, and occupation contribute to risk factors for cardiovascular disease.   Am J Public Health. 1992;82(6):816-820. doi:10.2105/AJPH.82.6.816 PubMedGoogle ScholarCrossref
    38.
    Horvath  A, Greenberg  L.  Development and validation of the Working Alliance Inventory.   J Counseling Psychol. 1989;36(2):223-233. doi:10.1037/0022-0167.36.2.223 Google ScholarCrossref
    39.
    King  AC, Hekler  EB, Grieco  LA,  et al.  Harnessing different motivational frames via mobile phones to promote daily physical activity and reduce sedentary behavior in aging adults.   PLoS One. 2013;8(4):e62613. doi:10.1371/journal.pone.0062613 PubMedGoogle Scholar
    40.
    Kaji  AH, Lewis  RJ.  Noninferiority trials: is a new treatment almost as effective as another?   JAMA. 2015;313(23):2371-2372. doi:10.1001/jama.2015.6645 PubMedGoogle ScholarCrossref
    41.
    Sundquist  K, Qvist  J, Sundquist  J, Johansson  SE.  Frequent and occasional physical activity in the elderly: a 12-year follow-up study of mortality.   Am J Prev Med. 2004;27(1):22-27. doi:10.1016/j.amepre.2004.03.011 PubMedGoogle ScholarCrossref
    42.
    Kushi  LH, Fee  RM, Folsom  AR, Mink  PJ, Anderson  KE, Sellers  TA.  Physical activity and mortality in postmenopausal women.   JAMA. 1997;277(16):1287-1292. doi:10.1001/jama.1997.03540400037028 PubMedGoogle ScholarCrossref
    43.
    Albright  CL, Pruitt  L, Castro  C, Gonzalez  A, Woo  S, King  AC.  Modifying physical activity in a multiethnic sample of low-income women: one-year results from the IMPACT (Increasing Motivation for Physical ACTivity) project.   Ann Behav Med. 2005;30(3):191-200. doi:10.1207/s15324796abm3003_3 PubMedGoogle ScholarCrossref
    44.
    Finney  JW, Mitchell  RE, Cronkite  RC, Moos  RH.  Methodological issues in estimating main and interactive effects: examples from coping/social support and stress field.   J Health Soc Behav. 1984;25(1):85-98. doi:10.2307/2136706 PubMedGoogle ScholarCrossref
    45.
    West  BT, Welch  KB, Galecki  AT.  Linear Mixed Models: A Practical Guide Using Statistical Software. CRC Press; 2006. doi:10.1201/9781420010435
    46.
    Tan  WY, Tabatabai  MA.  A modified winsorized regression procedure for linear models.   J Statistical Computation and Simulation. 2007;30(4):299-313. doi:10.1080/00949658808811103 Google ScholarCrossref
    47.
    Little  RJA, Rubin  DB.  Statistical Analysis with Missing Data, Second Edition. John Wiley & Sons; 2002. doi:10.1002/9781119013563
    48.
    Hekler  EB, Buman  MP, Haskell  WL,  et al.  Reliability and validity of CHAMPS self-reported sedentary-to-vigorous intensity physical activity in older adults.   J Phys Act Health. 2012;9(2):225-236. doi:10.1123/jpah.9.2.225 PubMedGoogle ScholarCrossref
    49.
    Kallogjeri  D, Francis  DO, Piccirillo  JF.  Use of confidence intervals to improve the quality of research results reporting and interpretation.   JAMA Otolaryngol Head Neck Surg. 2019;145(6):491-493. doi:10.1001/jamaoto.2019.0793 PubMedGoogle ScholarCrossref
    50.
    King  AC, Castro  C, Wilcox  S, Eyler  AA, Sallis  JF, Brownson  RC.  Personal and environmental factors associated with physical inactivity among different racial-ethnic groups of US middle-aged and older-aged women.   Health Psychol. 2000;19(4):354-364. doi:10.1037/0278-6133.19.4.354 PubMedGoogle ScholarCrossref
    51.
    Wilcox  S, King  AC, Brassington  G, Ahn  D.  Physical activity preferences of middle-aged and older adults: a community analysis.   Journal Aging Physical Activity. 1999;7:386-399. doi:10.1123/japa.7.4.386 Google ScholarCrossref
    52.
    King  AC, Castro  CM, Buman  MP, Hekler  EB, Urizar  GG  Jr, Ahn  DK.  Behavioral impacts of sequentially versus simultaneously delivered dietary plus physical activity interventions: the CALM trial.   Ann Behav Med. 2013;46(2):157-168. doi:10.1007/s12160-013-9501-y PubMedGoogle ScholarCrossref
    53.
    Gothe  NP, Wójcicki  TR, Olson  EA,  et al.  Physical activity levels and patterns in older adults: the influence of a DVD-based exercise program.   J Behav Med. 2015;38(1):91-97. doi:10.1007/s10865-014-9581-6 PubMedGoogle ScholarCrossref
    54.
    Wijsman  CA, Westendorp  RG, Verhagen  EA,  et al.  Effects of a web-based intervention on physical activity and metabolism in older adults: randomized controlled trial.   J Med internet Res. 2013;15(11):e233. doi:10.2196/jmir.2843 PubMedGoogle Scholar
    55.
    Kurti  AN, Dallery  J.  Internet-based contingency management increases walking in sedentary adults.   J Appl Behav Anal. 2013;46(3):568-581. doi:10.1002/jaba.58 PubMedGoogle ScholarCrossref
    56.
    Alley  SJ, Kolt  GS, Duncan  MJ,  et al.  The effectiveness of a web 2.0 physical activity intervention in older adults - a randomised controlled trial.   Int J Behav Nutr Phys Act. 2018;15(1):4. doi:10.1186/s12966-017-0641-5 PubMedGoogle ScholarCrossref
    57.
    King  AC, Hekler  EB, Grieco  LA,  et al.  Effects of three motivationally targeted mobile device applications on initial physical activity and sedentary behavior change in midlife and older adults: a randomized trial.   PLoS One. 2016;11(6):e0156370. doi:10.1371/journal.pone.0156370 PubMedGoogle Scholar
    58.
    Colón-Semenza  C, Latham  NK, Quintiliani  LM, Ellis  TD.  Peer coaching through mhealth targeting physical activity in people with parkinson disease: Feasibility Study.   JMIR Mhealth Uhealth. 2018;6(2):e42. doi:10.2196/mhealth.8074 PubMedGoogle Scholar
    59.
    Victorson  D, Banas  J, Smith  J,  et al.  eSalud: designing and implementing culturally competent ehealth research with Latino patient populations.   Am J Public Health. 2014;104(12):2259-2265. doi:10.2105/AJPH.2014.302187 PubMedGoogle ScholarCrossref
    60.
    López  L, Tan-McGrory  A, Horner  G, Betancourt  JR.  Eliminating disparities among Latinos with type 2 diabetes: effective ehealth strategies.   J Diabetes Complications. 2016;30(3):554-560. doi:10.1016/j.jdiacomp.2015.12.003 PubMedGoogle ScholarCrossref
    61.
    Rodriquez  EJ, Pérez-Stable  EJ.  The time is now for ehealth research with Latinos.   Am J Public Health. 2017;107(11):1705-1707. doi:10.2105/AJPH.2017.304055 PubMedGoogle ScholarCrossref
    62.
    Stenholm  S, Vahtera  J, Kawachi  I,  et al.  Patterns of weight gain in middle-aged and older US adults, 1992-2010.   Epidemiology. 2015;26(2):165-168. doi:10.1097/EDE.0000000000000228 PubMedGoogle ScholarCrossref
    63.
    Bikdeli  B, Welsh  JW, Akram  Y,  et al.  Noninferiority designed cardiovascular trials in highest-impact journals.   Circulation. 2019;140(5):379-389. doi:10.1161/CIRCULATIONAHA.119.040214 PubMedGoogle ScholarCrossref
    64.
    Wilcox  S, Dowda  M, Leviton  LC,  et al.  Active for life: final results from the translation of two physical activity programs.   Am J Prev Med. 2008;35(4):340-351. doi:10.1016/j.amepre.2008.07.001 PubMedGoogle ScholarCrossref
    65.
    Wilcox  S, Dowda  M, Wegley  S, Ory  MG.  Maintenance of change in the Active-for-Life initiative.   Am J Prev Med. 2009;37(6):501-504. doi:10.1016/j.amepre.2009.07.016 PubMedGoogle ScholarCrossref
    ×