UBC Faculty Research and Publications

GrOup based physical Activity for oLder adults (GOAL) randomized controlled trial: study protocol Beauchamp, Mark R; Harden, Samantha M; Wolf, Svenja A; Rhodes, Ryan E; Liu, Yan; Dunlop, William L; Schmader, Toni; Sheel, Andrew W; Zumbo, Bruno D; Estabrooks, Paul A Jun 27, 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
52383-12889_2015_Article_1909.pdf [ 671.69kB ]
Metadata
JSON: 52383-1.0223733.json
JSON-LD: 52383-1.0223733-ld.json
RDF/XML (Pretty): 52383-1.0223733-rdf.xml
RDF/JSON: 52383-1.0223733-rdf.json
Turtle: 52383-1.0223733-turtle.txt
N-Triples: 52383-1.0223733-rdf-ntriples.txt
Original Record: 52383-1.0223733-source.json
Full Text
52383-1.0223733-fulltext.txt
Citation
52383-1.0223733.ris

Full Text

STUDY PROTOCOLtd. Wdobesity represent some of the most prevalent chronic physical activity recommendations [5].and Unitedon is 65 years036 [10].re are severall activity thatpolicy factors.Beauchamp et al. BMC Public Health  (2015) 15:592 DOI 10.1186/s12889-015-1909-9meta-analytic reviews suggest that people are more likely1School of Kinesiology, University of British Columbia, Vancouver, BC, CanadaFull list of author information is available at the end of the articleIn the current trial we focus on the social context in whichphysical activity takes place. In particular, the results of* Correspondence: mark.beauchamp@ubc.careduced risk for cardiovascular disease, functional capacity,and quality of life [3]. Further, those who maintain mobilityare more likely to remain in their community of origin forlonger [4], which is often associated with higher personalquality of life. In spite of the myriad health benefitsassociated with physical activity, older adults represent theother developed countries (e.g., AustraliaStates [8, 9]), 14 % of the Canadian populatior older with an estimated rise to 25 % by 2From a population health perspective, thebroad categories of determinants of physicainclude personal, social, environmental, andconditions associated with older adults’ physical inactivity[1, 2]. Engaging in 150 min of moderate activity per weekis associated with marked improvements for older adults’The current prevalence of physical inactivity has beenimplicated in the high rate of provincial and territorialgovernment health spending in Canada [7]. Similar toCardiovascular disease, arthritis, decreased mobility, andBackground: Physical activity has health benefits across the lifespan, yet only 13 % of Canadian older adults aresufficiently active. Results from a number of observational studies indicate that adults display positive preferencesfor exercising with others of a similar age and same gender, and that intra-group age- and gender-similarity areassociated with elevated exercise adherence. However, research has yet to experimentally examine the extent towhich intra-group age- and gender-related similarity affect exercise adherence behaviors.Methods/design: The GrOup-based physical Activity for oLder adults (GOAL) trial is a three-arm randomized controltrial that will examine the efficacy of two different group-based exercise programs for older adults (informed by thetenets of self-categorization theory) in relation to a standard group-based exercise program. Within this manuscriptwe outline the design and proposed evaluation of the GOAL trial. The first arm is comprised of exercise groups madeup of participants of a similar-age and of the same gender; the second arm consists of groups with similar-aged mixedgender participants; the control arm is comprised of mixed-aged mixed gender participants. We aim to compare theadherence rates of participants across conditions, as well as potential moderation effects and mediating mechanisms.Discussion: Results from this trial will inform intervention designs to improve the exercise adherence behaviors ofolder adult. At a systems-level, should support be derived for the efficacy of the interventions tested in this trial, changinggroup composition (i.e., age, gender) represents a feasible program adaptation for physical activity centers.Trial registration: ClinicalTrials.gov # NCT02023632. Registered December 13, 2013.Keywords: Physical activity, Older adults, Adherence, Self-categorization theory, Group-dynamicsBackground least active cohort [5, 6], with only 13.1 % of Canadians overthe age of 65 (Men = 13.7 %, Women = 12.6 % ) meetingGrOup based physical Acadults (GOAL) randomizestudy protocolMark R. Beauchamp1*, Samantha M. Harden1,2,3, Svenja AToni Schmader7, Andrew W. Sheel1, Bruno D. Zumbo8 anAbstract© 2015 Beauchamp et al. This is an Open AccLicense (http://creativecommons.org/licenses/medium, provided the original work is propercreativecommons.org/publicdomain/zero/1.0/Open Accessivity for oLdercontrolled trial:olf1, Ryan E. Rhodes4, Yan Liu1,5, William L. Dunlop1,6,Paul A. Estabrooks2,3,9ess article distributed under the terms of the Creative Commons Attributionby/4.0), which permits unrestricted use, distribution, and reproduction in anyly credited. The Creative Commons Public Domain Dedication waiver (http://) applies to the data made available in this article, unless otherwise stated.Beauchamp et al. BMC Public Health  (2015) 15:592 Page 2 of 11to sustain their involvement in physical activity programsif they are provided with the opportunity to exercisewith others in social, or group-based, settings ratherthan on their own [11, 12]. In line with this body ofevidence, group-based physical activity programs havebeen identified as a particularly effective means ofpromoting sustained physical activity involvement amongolder adults [13, 14], and also provide an important meansof maintaining quality of life and reducing the potentialdebilitating effects of social isolation that older adultsoften encounter [15].Despite the potential for groups to sustain long-termphysical activity behaviors, there appears to be animportant caveat that comes with exercising withothers: If people perceive themselves to be similar toother members of a given group, in terms of salientunderlying qualities, this corresponds positively withtheir attraction to, and level of involvement within, thatgroup [16–20]. If, however, people perceive themselves tobe distinctly different from others within their socialgroup, this is likely to undermine their attraction to, andinvolvement in that group [16–20]. Recent researchsuggests that across the adult age spectrum people reporta positive preference for exercising within groups that arecomprised of others their own age [16, 21], and when theyparticipate in such classes they display higher levels ofadherence to the group [18]. In a similar regard, peoplereport comparable positive preferences for same-genderrather than mixed-gender physical activity group settings[19]. This preference exists for both males and femalesalthough as demonstrated in our recent research thestrength of this effect appears slightly stronger for women(d = .76) than for men (d = .30) [19]. As a complement tothese findings, a recent case study of a highly efficacious‘similar-age-same-gender’ physical activity program forolder adults [22] demonstrated noteworthy rates ofadherence (with over 45 % of its membership adherent forover 10 years and approximately 70 % adherent for morethan 5 years). In light of findings from the above observa-tional studies [16–20] as well as those of the recent casestudy [22], there is now sufficient evidence to support thedevelopment and application of group-based physicalactivity programs for older adults that incorporate theseage- and gender-based considerations, and testing theefficacy of these programs through use of a RandomizedControlled Trial (RCT) design.Conceptual frameworkSelf-categorization theory [23, 24] serves as the conceptualframework for this trial. This theory was developed byTurner and colleagues [23] and built upon its precursorsocial identity theory [25, 26]. Initially, social identitytheory purported that people not only develop a sense ofpersonal (i.e., individualized) identity through relianceupon factors that make them unique, but also possesssocial identities, based on their membership in socialgroups [25, 26]. When a social identity is made salient,individuals tend to favor persons who share membershipin the applicable social group (i.e., in-group members)over those from other social groups (i.e., out-groupmembers). Although social identity theory included recog-nition of the fact that social identities will carry implicationsfor both within- and between-group behavior, the predom-inant focus of this framework centered on between-group(i.e., intergroup) processes [25]. To explicate the cognitiveprocesses by which people categorize themselves andothers, and define themselves in terms of membershipwithin different social groups, Turner and his colleaguesdeveloped self-categorization theory [23]. This theoryfocuses to a much greater extent on within-group(i.e., intragroup) processes than social identity theory.The underlying premise behind self-categorizationtheory is that people place themselves and others intosocial categories on the basis of a set of underlyingattributes that are particularly salient, and this processof social categorization shapes a range of attitudes,emotions, and behaviors [23]. Specifically, according toself-categorization theory people are generally attracted toothers with whom they share membership in a givencategory (i.e., “birds of a feather flock together”) andrepelled by those with whom they do not share categorymembership [23]. There is a growing body of evidencesupporting the notion that the extent to which peopleself-identify as being similar to, or different from, otherswithin physical activity group contexts, on the basisof social categories such as age and gender influencestheir attraction to, and level of involvement within, thatgroup [16–20].MediatorsIn a seminal position paper on developing effective physicalactivity interventions, Baranowski and colleagues suggestedthat in order for interventions to be effective in changingbehavior, a sound understanding of the ‘key’ psychologicaldeterminants (i.e., mediators) of behavior change isrequired [27]. In the context of this trial, two theoreticalmediators – group cohesion and affective attitudes – willbe examined to explain the a priori expected relationsbetween involvement in age- and gender-congruentphysical activity groups and their adherence behaviors.Group cohesion is defined as “a dynamic process that isreflected in the tendency for a group to stick together andremain united in the pursuit of its instrumental objectivesand/or for the satisfaction of member affective needs”(p. 213 [28]), and includes both task and social compo-nents. A core theoretical tenet of self-categorization theorycorresponds to the similarity-attraction hypothesis, wherebypeople are more likely to feel attracted to those with whomBeauchamp et al. BMC Public Health  (2015) 15:592 Page 3 of 11they perceive themselves to be similar [23]. As such, itstands to reason that when older adults participate ingroups comprised of those of the same age and gender, theywould be expected to demonstrate greater attraction to thegroup’s social and task activities, and also perceive thegroup to be more united (i.e., higher levels of cohesion)[20]. Furthermore, in light of the fact that cohesionhas consistently been found to predict improvementsin physical activity adherence behavior [11, 12], we wouldexpect that the covariance of the assigned conditionson adherence (at 3 and 6 months) will be explained(mediated) by older adults’ perceptions of task andsocial cohesion.A second theoretical mediator that will be examinedin this research corresponds to older adults’ affectiveattitudes. Affective judgments play an important role inmany key theories of behavior change and can bedefined as the overall pleasure/displeasure and enjoymentexpected from a given activity [29]. For example, in thetheory of planned behavior, affective judgments areconceptualized through affective attitudes [29], withinsocial cognitive theory they are conceptualized throughaffective outcome expectations [30], and within self-determination theory they are conceptualized within theintrinsic motivation regulation [31]. In this trial weoperationalize the affective attitudes construct as concep-tualized within the theory of planned behavior [32].Affective attitudes correspond to how enjoyable (orunenjoyable) an activity is perceived to be, whichcontrasts with instrumental attitudes that are concernedwith how useful (or useless) that activity is perceived. In arecent meta-analytic review, Rhodes and colleagues [33]demonstrated that affective (rather than instrumental)attitudes significantly predicted physical activity behaviors.That is, people tend to engage in physical activity on thebasis of whether they enjoy that activity, and not on thebasis of whether it is perceived to carry some future healthbenefits. This is also consistent with an extensive body ofresearch in social psychology that has found the level ofaffect (enjoyment) experienced in a given situation is aconsistent predictor of the amount of time people chooseto spend in that situation [34]. In light of our previousfindings that people report a general preference forage-matched [16] and same-gender groups [19] we wouldexpect that the covariance of the assigned conditionson adherence (at 3 and 6 months) will be explained(mediated) by older adults’ affective attitudes (enjoyment)towards those contexts.Aims and hypothesesDrawing from the tenets of self-categorization theoryand previous observational research our primary researchquestion was concerned with whether older adults sustaintheir involvement in physical activity programs (over threeand six months) when they participate in groups that arecomprised of members of a similar age and same gender,relative to those taking part in similar-age but mixed-gender classes. Both of these conditions will be comparedto the adherence of older adults within standard (mixed-age mixed-gender) ‘control’ physical activity groups.Second, we were interested in whether, at the end ofa 3-month physical activity program, older adults inSASG (similar-age same-gender) classes will re-enrollin SASG classes over the following 3-month period(6 months in total) to a greater extent when compared toolder adults in the SAMG (similar-age mixed-gender) andMAMG (mixed-age mixed-gender) conditions. Our thirdresearch question was concerned with whether any groupdifferences among these adherence outcomes can beexplained through a mediation model (with cohesionand affective attitudes as target mediators). Our fourthresearch question was concerned with whether there aregender differences across the primary outcome (physicalactivity adherence behavior) by assigned condition.Consistent with our previous findings that womendemonstrate a slightly stronger preference for same-gender contexts [19] we expected that (in addition tomain effects for the SASG context when compared tothe mixed-gender settings), the effects for SASG versus theother two conditions (in relation to adherence) would bemore pronounced for women. The following hypotheseswill be tested:Hypothesis 1. Older adults in the SASG groupscondition will demonstrate improved adherence(i.e., attendance rates) over 3-months and 6-monthsthan older adults in the SAMG condition, who in turnwill demonstrate improved adherence to those olderadults in a standard (control) group-based exercisecondition (MAMG).Hypothesis 2. A greater proportion of older adults inthe SASG condition will re-enroll in the program (after3 months for an additional 3 months; 6 months total)than those in the SAMG and MAMG groups.Hypothesis 3. The covariance of the assigned conditions(SASG, SAMG, MAMG) on adherence (at 3 and6 months) will be explained (mediated) by changes inolder adults’ perceptions of group cohesion and affectiveattitudes (enjoyment).Hypothesis 4. Gender will moderate the effects of theintervention conditions in relation to adherence andre-enrollment.In addition to the above outcome assessment analyses,a process evaluation will be conducted to evaluate theprocedures embedded within the intervention. Thiswill involve qualitative (interview-based) methodologies.Although no a priori hypotheses will be tested, the processBeauchamp et al. BMC Public Health  (2015) 15:592 Page 4 of 11evaluation will provide important insight into both con-tent fidelity (“what is done”) and process fidelity (“how itis done”) of the trial [35].MethodStudy designThe GrOup-based physical Activity for oLder adults(GOAL) Trial is a 3-arm randomized controlled trial(RCT) developed in alignment with the tenets of self-categorization theory. The study has been approved bythe Behavioral Research Ethics Board at The University ofBritish Columbia, and is registered with ClinicalTrials.gov(# NCT02023632).The design, conduct, and reporting of this study willadhere to the Consolidated Standards of ReportingTrials (CONSORT) guidelines [36]. The pre-screeningprocess included the completion of the Physical ActivityReadiness Questionnaire for Everyone (PARQ+) andElectronic Physical Activity Readiness Medical Examination(ePARmed-X+) [37]. Pre-screening was conducted either bya trained research assistant or the project coordinator(authors SMH and SAW) using a pre-screening script forconsistency. If the ePAR-medX+ highlighted that the inter-ested older adult needed physician approval prior to joiningthe program, the research assistant informed the individualand indicated that subsequent physician approval was re-quired before s/he then could enroll in the study. Followingthe initial screening process for inclusion/exclusion,informed consent was obtained, and participants wererandomized to one of three study arms:1. Similar-Age Same-Gender (SASG)2. Similar-Age Mixed-Gender (SAMG)3. Mixed-Age Mixed-Gender (MAMG)ParticipantsWe aimed to recruit 540 older adults that would berandomized across the three conditions. To be eligibleparticipants needed to be 65 years of age or older(targeted recruitment of 50 % female and 50 % male)and not have contraindication which might prevent themfrom participating in moderate-intensity physical activity.To both effectively manage the trial and avoid unnecessaryburden on the respective YMCA centers, the trial wasdesigned to run in two (N = 270 in cohort 1, and N = 270in cohort 2) cohorts. The first cohort ran between Marchand August 2014 and the second cohort is runningbetween March and August 2015. Both cohorts are runwithin the same March-August window in order tominimize any seasonal effects between cohorts.Study interventionsThe trial is taking place at three different YMCA centersin the Lower Mainland of British Columbia, over 2 years.The group exercise classes take place in 3-month blockswith the opportunity to re-enroll in the same program foranother 3 months, thus lasting 6-months in total (see Fig. 1).Each program is run on the basis of classes taking placethree days per week, with classes lasting 50–60 min. Thisdosage (150–180 min/week) is consistent with Canada’scurrent physical activity guidelines for older adults engagingin ≥ 150 min of moderate-to-vigorous physical activityper week [38], as well as findings from the CanadianCommunity Health Survey [39] which found that 67 %of seniors who are active three or more times a weekare in good health, compared to 36 % who are infre-quently active.The design of the SASG physical activity conditionwas informed by the tenets of self-categorization theoryas well as the results of a recent case study of a highlyefficacious physical activity program entitled the LivelyLads (a pseudonym) for older adult males [22]. Althoughthe Lively Lads program was developed by-seniors-for-seniors, it made use of a number of salient theory-drivengroup dynamics strategies that appear to be implicatedin its success. Several of the Lively Lads strategies wereutilized in the design of the SASG intervention arm ofthe GOAL Trial. First, one of the core features of thiscondition is that it designed exclusively for those of asimilar age and same gender. Such an environment wasreported by Dunlop and Beauchamp [22] to provideopportunities for social connectedness, as well as personalcomfort (e.g., reduced likelihood of embarrassment anddisplays of physical incompetence in front of women).Second, the volunteer exercise class instructors are drawnfrom the ranks of older adults. Such an approach isconsistent with research from the perspective of socialcognitive theory, highlighting the value of ‘similar models’as sources of vicarious efficacy enhancement informationand verbal persuasion [40]. From the perspective ofprogram sustainability, volunteer exercise leaders enablethe program to keep its costs low (there are no costsinvolved in paying instructors, as they are volunteers).Consistent with social identity and self-categorizationperspectives, the program also makes use of a series ofstrategies to foster group identity (e.g., providing partici-pants with T-shirts to foster a sense of ‘distinctiveness’).Finally, although a major objective of the SASG interven-tion condition is to engage in physical activity, an importantstrategy (informed by both social identity theory and theLively Lads program) is to provide opportunities for theolder adults to connect with one another after the classeshave ended (e.g., post-workout coffee gatherings). Althoughthe Lively Lads program was developed with older adultmales, classes are also provided in the GOALTrial for olderadult women at each YMCA site.The SAMG physical activity condition mirrors the SASGgroup condition, insofar as the program was restricted toBeauchamp et al. BMC Public Health  (2015) 15:592 Page 5 of 11older adults (≥65 years), but was open to older adults fromboth genders. The same strategies to those usedwithin the SASG condition were also utilized (e.g., T-shirts,opportunities to socialize after the program), with classesalso offered on three days per week. Older adults were alsorecruited to be instructors for the group classes (≥65 years),with both males and females invited to occupy theseinstructional roles.The control condition operationalized within the RCTis designed to reflect ‘standard’ group-based exercises thatone sees in typical physical activity centers. Specifically,these classes (regular group-based physical activity classesrun by the respective YMCAs) are not restricted toparticipants on the basis of age or gender, and as such olderadults in this condition participate in groups comprised ofFig. 1 Flow of participants through the trialpeople younger than themselves as well as those of bothgenders. Specifically, participants randomized to this condi-tion were invited to select one of the standard group-basedexercise classes offered by the respective YMCA. Althoughthese classes involve both GOAL Trial participantsand regular YMCA members, only the older adultsrecruited to and consented in the GOAL Trial will beused in the analyses.Program structureThe intervention group classes were developed with theintent of fostering an engaging environment that reflectssound group exercise classes [41]. That is, the researchteam purchased music playlists that had the appropriatebeats per minute (BPM) to align with warm-up/cool-downBeauchamp et al. BMC Public Health  (2015) 15:592 Page 6 of 11(120–134 BPM) and moderate intensity physical activity(135–160 BPM). Further, a website was developed toprovide instructors (and class participants) with audio,written, and visual (i.e., videos) tutorials for completingexercises that were performed in the sessions. Thesetutorials made explicit use of both male and female olderadult models to demonstrate correct posture and form.As with the Lively Lads program, similar-aged volunteerinstructors lead the GOALTrial exercise classes. Instructorswere recruited via local media, flyers at the partneringYMCA, and word of mouth. In addition, for the secondyear of the trial we recruited former participants (from thefirst cohort) who had expressed an interest to continue asvolunteer instructors for the second cohort. Instructorswere provided training, through three modules, at one ofthe three participating YMCAs related to ‘ProgramBasics and Guidelines’, ‘Tailoring the Program’, and‘Instructor Mastery’. This training was guided by aManual of Procedures (available from the first authoron request). In the ‘Program Basics and Guidelines’module, instructors were informed of the study goalsand the underlying theoretical principles in lay terminology.They also completed a facility orientation (including FirstAid procedures). In the ‘Tailoring the Program’ module,volunteer instructors learned of the different experimentalconditions and how to modify the program specific to theSASG or SAMG composition of their group. Finally,instructors had the opportunity to practice leading a group(of the other volunteer instructors) in an exercise class inorder to develop their confidence and comfort with theexercises. The duration of volunteer training depended oninstructors’ level of certification and previous experience;however, 45 hours of instruction were available.The instructors were provided with guidance regardingthe structure of the program (i.e., to include the warm-up,moderate intensity exercises, and a cool-down). However,each instructor had autonomy on choosing the exercisesto be included in each class. Instructors were providedwith six unique sequences of recommended exercises withthemes of: a full-body, basic class (All Over Burn) aswell as classes emphasizing gluteal and back muscles(Spectacular Backsides), abdominal muscles (ABsolutelyIntense), upper body (Superhuman Strength), agility andbalance moves (Adios Arthritis and Balance Bodies).OutcomesAll pre-screening measures took place in January andFebruary (2014 for cohort 1 and 2015 for cohort 2).Baseline physical health and fitness assessments wereconducted in the last week of February, at which pointparticipants also completed questionnaires designed tomeasure demographic and background variables, physicalactivity, general health status, as well as a measure ofparticipant personality. In weeks 2, 7, 12, 14, 19, and 24 ofthe intervention (March to August) participants com-pleted questionnaires that included measures of cohesion,instrumental and affective attitudes, task and self-regulatoryself-efficacy, stigma consciousness and psychologicalflourishing. At week 2 and 14 assessments the question-naire battery also included measures of intra-group percep-tual similarity, commute time, and commute mode. Atweek 7 and 9, the questionnaire battery also included mea-sures of intra-group communication and group interactionprocesses. In weeks 12 and 24, the questionnaire batteryalso included measures of participants’ physical activitybehavior outside of their respective YMCA programs.Measures of program adherence were obtained throughoutthe course of the respective six-month programs. At theend of the respective six-month programs, participantscompleted the same physical health and fitness assessmentsas those completed at baseline, at which point the processevaluation interviews were also conducted.MeasuresDemographic and background measures Data relatedto a number of background and demographic vari-ables were collected in relation to participants’ age,sex, country of birth, dwelling arrangements [42], postcode (as a measure of socio economic position) [42], andemployment/retirement status [43] as well as CanadianCensus questions [44] for marital status, ethnicity, level ofeducation, household income. We also collected measuresrelated to participants’ commute time, and mode of trans-port, to the respective YMCA program.General health status Data were collected in relation toparticipants’ smoking status, general health status, previoushistory of illness, and current use of medication.Physical activity adherence behaviors Class attendancewas objectively measured via reports generated by theuse of participants’ YMCA access cards. These data willbe used to determine attendance throughout the trial.With regard to the research question concerning theextent to which participants choose to re-enroll after theinitial 3-month program has ended, program enrolmentrecords were used (dichotomous: 1 = yes, 0 = no). Wecollected data at baseline related to participants’ physicalactivity behavior using Godin’s Leisure Time ExerciseQuestionnaire (LTEQ) [45], and at weeks 12 and 24 wecollected measures related to participants’ physical activitybehaviors outside of the program, using proceduresdescribed by Wilcox and colleagues [46].Physical health and fitness measures We collected datarelated to participants’ height, weight, body composition,blood pressure, functional fitness, and mobility. Specificallyheight was measured using a standiometer. Both weightBeauchamp et al. BMC Public Health  (2015) 15:592 Page 7 of 11(kgs) and body composition (percentage of body fat, asassessed through bio-electrical impedance) were measuredusing a commercially available portable body compositionanalyzer (Tanita Model TBF 300 GS, Tanita ManufacturingCo., Tokyo, Japan). Blood pressure was assessed usingautomatic blood pressure monitors (Life source UA-767Plus, A&D Medical, USA). These monitors use the oscillo-metric method to simultaneously provide recordingsof systolic (SBP) and diastolic blood pressure (DBP).Participants were required to remain seated for atleast 5 min prior to all assessments. Three recordingswere made, with an average taken for SBP and DBP.Mean Arterial Pressure (MAP) is calculated accordingthe following equation: MAP = DBP + 1/3(SBP - DBP).Finally, participants completed the functional fitness testfor older adults developed by Rikli and Jones [47]. Thisincludes a battery of six tests that assess upper and lowerbody flexibility and strength as well as aerobic fitness via a2-min step test. The activities performed during thesetests are designed to reflect “the physiologic attributes thatsupport the behavioral functions necessary to performactivities of daily living” (p. 133, [47]).Psychological variables The primary psychologicalcognitions targeted in the intervention related to groupcohesion and affective attitudes (enjoyment). Classcohesion was assessed using the Physical Activity GroupEnvironment Questionnaire (PAGEQ) [48]. The PAGEQ isa 21-item self-report questionnaire designed to assess fourdimensions of cohesion within exercise classes; namely,attraction to the group’s task (ATG-T), and social (ATG-S)activities, as well as perceptions of group integrationaround the group’s task (GI-T), and social (GI-S) activities.The PAGEQ was developed specifically for older adultstaking part in physical activity classes, with scores derivedfrom this instrument found to demonstrate good reliability(α ≥ .72 [48]), factorial validity, and predictive utility.Affective attitudes (enjoyment) towards physical activitywere assessed using the procedures described by Rhodesand Matheson [49]. Specifically, a 7-point semanticdifferential scale was used, with anchors including“Enjoyable—Unenjoyable”, “Pleasant—Unpleasant”, “Inter-esting—Boring.” Previous research with older adults hasfound support for both the internal consistency and pre-dictive utility of scores derived from this instrument [49].In addition to the two primary psychological variablestargeted in the intervention (i.e., cohesion and affectiveattitudes), data were also collected on a secondary set ofpsychological variables related to participants’ (a) personal-ity [50], (b) task self-efficacy [51], (c) barriers self-efficacy[52], (d) instrumental attitudes [49], (e) stigma conscious-ness [53], (f) intra-group communication and group inter-action processes [54], (g) psychological flourishing [55], and(h) within class perceptual similarity [17, 20].Process evaluation A process evaluation was conductedto provide insights into both content fidelity (“what isdone”) and process fidelity (“how it is done”) with regardto intervention delivery, as well as the extent to whichthe intervention meets with the needs of those involved(i.e., older adults) in the program [35]. Without anyassessment of intervention fidelity internal validity ispotentially compromised [35]. Furthermore, as Plummerand colleagues [56] suggest, process evaluations “canhelp explain the program’s outcomes and identifyways to improve and/or replicate it. For example, ifthere are unsatisfactory outcomes, it is important tounderstand whether this could be due to poor programdesign, inadequate implementation or special contextualfactors.” (p. 500).In the GOAL Trial, interviews with program participantswill enable us to appraise the specific subcomponents ofthe program and, where appropriate, further modify thesefor future initiatives [57]. Semi-structured interviews wereused that allow us to examine each of the underlyingprinciples of the program (e.g., effects of intra-group ageand gender composition, perceptions of class instructor).One of the project coordinators (author SAW) conductedthe interviews with participants, and although qualitativedata analysis will be overseen by the principal investigator,the coding will be performed by research assistants(i.e., unconnected with the intervention activities) [58].Sample sizeWe powered our study to detect significant differencesin individuals’ physical activity adherence (over 3 and6 months) in the SASG groups when compared tothe SAMG and MAMG standard care control condi-tion. In order to detect a medium effect size f = .25(difference between the SASG and both SAMG andMAMG conditions) based on a 2 (Gender) x 3 (Conditions)ANCOVA with the percentage of classes attended overthree and six months specified as dependent variable(while controlling for baseline levels of physical activity)with power (1 - β) set at .80, and alpha set at .05, 211participants were required across the 3 centers [59].In order to conduct a logistic regression analysisbased on individuals’ re-enrollment across the threegroup-based programs, based on power at .80, alpha at .05,an anticipated medium effect size (odds ratio = 2.5), and abalanced design, 167 participants were required [59].To test for mediation through use of a cross-laggedpanel model based on a structural equation modeling(SEM) approach, while modeling gender invariance, wedrew from three broad criteria. First, based on recommen-dations provided by MacCallum et al. [60], and usingPreacher and Coffman’s [61] R-code for assessing RMSEA(< .05), based on power set at .80, a minimum sample sizeof 163 was identified as being necessary for conducting theBeauchamp et al. BMC Public Health  (2015) 15:592 Page 8 of 11cross-lagged panel model (without considering genderdifferences). This calculation approximates with recom-mendations provided by Garver and Mentzer [62] that a‘critical sample size’ of 200 is required as a general rule ofthumb for providing sufficient statistical power for SEManalysis. In addition to these two sets of considerations, inlight of the fact that the panel model will include amulti-group (males, females) component in order toexamine gender invariance, this requires twice thesample size (i.e., n = 326, cf. MacCallum et al.; n = 400,[60]). Thus, in order to account for an attrition rate ashigh as 25 % (over the course of the program) a samplesize of 540 would be sufficient to examine the mediationalmodel proposed in this trial. In sum, we determined thatan overall sample size of 540 older adults across the threearms of the trial would be sufficient for examining each ofthe study hypotheses.For the process evaluation component of the trial 15participants per experimental condition (n = 45 in total)will be invited to participate in a semi-structured interviewdesigned to evaluate each of the three experimentalconditions (group-based programs) embedded withinour trial. It has been suggested that such a samplesize is generally sufficient to ensure data saturationwith qualitative interview-based data [63].RecruitmentParticipants were recruited via advertisements placedthrough the local media, recreation centers, health carecenters, hospitals, physician general practices, shoppingmalls, golf courses, and online interest sites within theLower Mainland of British Columbia. Eligibility criteriawere such that participants must be 65 years of age orolder (both males and females) and did not have anycontraindications which might prevent that person fromparticipating in moderate-intensity physical activity. Weintended to recruit an equal proportion of males andfemales. Interested persons were asked to call the trialhotline to inquire about program details. The GOALTrial hotline was used for a pre-screening procedure.Proposed outcome analysesPreliminary analyses will be conducted to examinewhether any patterns of missing data exist (e.g., missingat random, missing completely at random, etc.) for eachof the psychosocial variables (cohesion, affective attitudes)using the Missing Value Analysis (MVA; examination ofLittle’s chi square test) on SPSS Version 20. The data willalso be examined for multivariate and univariate outliers,as well as for violations of normality, with the appropriatetransformation procedures utilized. Prior to the mainanalyses, we will also examine invariance in the primaryoutcome variable (adherence over 3 and 6 months) acrossthe two cohorts (cohort 1 – March to August 2014; cohort2 – March to August 2015). In light of the fact thetwo cohorts will be examined at exactly the same time ofyear, using exactly the same experimental procedures, wewould expect invariance in patterns of adherence acrossthe three conditions in each year, thus supporting thepooling of data from both cohorts. Two 2 (gender) x 3(conditions) ANCOVAs with baseline measures of physicalactivity (LTEQ scores) entered as a covariate, andadherence to the program over 3-months and 6-monthsas the dependent variable.A logistic regression analysis will be conducted toexamine the likelihood of older adults randomized to theSASG condition re-enrolling in the same SASG conditionafter the initial 3-month program, when compared to there-enrollment of older adults in the SAMG and MAMGconditions. In the logistic regression model, gender will beadded as an independent variable, with the regressionmodel explained by π ¼ 11−e−x where x = b0 + b1G1 + b2G2 + b3 gender + b4G1*gender + b5G2*gender (where b0 isthe intercept and b1-b5 are the slopes for predictors). Forhypotheses 1, 2, and 4, SPSS Version 20 will be used toanalyze the data.Mediation will be tested through use of a multi-group(to examine whether the effects are invariant across gender)cross-lagged panel model using a structural equationmodeling (SEM) framework. Cross-lagged panel modelsare a type of auto-regressive modeling. This approach wasproposed by Cole and Maxwell [64] for longitudinal dataand will be adapted for the purpose of the present study.This approach has two advantages. Specifically, it allows usto examine the reciprocal relations between the mediators(cohesion and affective attitudes), and the outcome(adherence). In a recent paper we indicated the importanceof assessing group cohesion throughout the lifespan of aphysical activity group, and not just through a singletime-point early-program measure (this has been themost commonly used method of assessing cohesionas a mediator within group dynamics physical activityresearch [65]). This panel modeling approach, inwhich all the variables are measured at multiple timepoints, can rigorously test the prospective relationsbetween predictors (assignment to experimental condition),mediators, and outcomes (i.e., predictors prior to mediatorsand mediators prior to outcomes). The cross-lagged panelmodel will be estimated using Mplus 7.2 with a fullinformation maximum likelihood (FIML) estimatorused to handle missing data [66]. This procedure willuse all available data points for parameter estimation underthe assumption that the data are missing at random. FIMLestimation tends to produce less biased estimates thandeletion or simple missing data imputation techniques(e.g., EM algorithm, regression, listwise/pairwise deletion,mean replacement) even when data are not missing atrandom [67]. By examining multi-group models, this willBeauchamp et al. BMC Public Health  (2015) 15:592 Page 9 of 11allow us to examine whether the structural pathways areinvariant across genders.Process evaluation analysisOlder adults purposively selected to participate in theprocess evaluation component of the trial, will be invitedto participate in interviews designed to elicit in-depthinformation about the quality of processes embeddedwithin the programs [68]. This component of interventionevaluation will draw from a qualitative social constructionistperspective [69] to understand in the older adults’ ownwords, the beneficial features and any problematic compo-nents of the exercise program. Social constructionism isconcerned with understanding the manner in which peoplereflect on and interpret their own and others’ behaviors,and the meanings and values that they ascribe to thoseinteractions. Data collected via the semi-structured inter-views will be analyzed through use of inductive contentanalytic procedures [70], and themes will be identifiedthat correspond to the strengths and limitations of therespective programs.Trial statusIn accordance with the proposed time-line, the fullprogram (e.g., recruitment, randomization, baseline testing,six-month physical activity programs, post testing) hasbeen completed for the first cohort of the GOALTrial. Forthe second cohort, participants have also been recruited,randomized, and completed baseline measures. Currently,the second cohort of the trial is underway, but (at the timeof this protocol manuscript submission) has not yet beencompleted). No data from either cohort have been subjectto any form of data analysis.DiscussionUnderstanding the predictors of physical activity adherenceis an important research endeavor within the fields ofpreventive medicine and health psychology. A growingbody of epidemiologic evidence now exists in supportof the ongoing involvement in active lifestyles amongolder adults, and indeed the benefits of regular activityamong this population have been well-established [71].Our proposed study will examine whether a theory-driven, evidence-based intervention has the capacity tosupport the sustained involvement in physical activityamong older adults. The proposed research will also pro-vide the most rigorous test to date of the efficacy of SASGphysical activity settings for sustaining the physical activitybehaviors of older adults. Although our previous researchon exercise preferences [16, 19] as well as the predictiveutility of intra-group similarity [16, 18, 20] points to theimportance and viability of SASG group contexts for sus-taining physical activity adherence behaviors, it shouldalso be noted that these studies utilized non-experimentalobservational designs. Thus, the causal link between intra-group similarity and exercise adherence has yet to beexamined.In the current study, an experimental (RCT) design isused to examine the efficacy of SASG (and SAMG)settings, the findings of which have the potential toinform the delivery of effective health-enhancing physicalactivity interventions that are likely to sustain theadherence behaviors of older adults in those programs. Ifeither the SASG or SAMG conditions have significantlyhigher adherence rates, when compared to the MAMGgroup, we will have preliminary evidence to support asmall change with substantive impact. From a knowledgetranslation perspective, attending to group compositionconsiderations is an easy, sustainable, and low-cost way toinfluence physical activity behaviors for older adults. Thisprogram would easily translate to a variety of physicalactivity settings including YMCAs, other community-centers, retirement communities, among others. The datacollected in this trial have the potential to inform nextsteps for large-scale implementation and understandingpotential mechanisms related to the efficacy of group-based interventions.There are a few potential limitations to address at theonset. First, the MAMG is unique in that participantscan choose any 3-day per week class offered at the YMCA.From a design perspective, there may be a limitation ingiving MAMG participants choice in terms of whichYMCA course they wished to participate, since partici-pants in the SASG and SAMG conditions were not givenany such choice. However, from a pragmatic perspec-tive, we believe that the MAMG group represents themost appropriate type of control condition as it reflectsthe typical type of exercise class available within commu-nity exercise settings. Thus, MAMG provides a strongpoint of comparison.There may also be critique of variation related to classleaders. Measures were taken in the extensive trainingof the volunteer, similar-aged instructors. All volunteerleaders were trained at the same time, through the threemodules outlined above. In this way, we sought toincrease the likelihood of treatment fidelity across allconditions at each site. However, there may be someinherent differences based on instructor personalities,level of exercise instructor experience, and so forth. Toaccount for these potential differences across conditionsand sites, the process evaluation includes queries aboutperceptions of the class leaders (e.g., leadership andcommunication style).AbbreviationsGOAL: GrOup-based physical Activity for oLder adults; MAMG: Mixed-AgeMixed-Gender; SAMG: Similar-Age Mixed-Gender; SASG: Similar-Age Same-Gender; PAR-Q+: The Physical Activity Readiness Questionnaire for Everyone;ePARmed-X: The Electronic Physical Activity Readiness Medical Examination.Beauchamp et al. BMC Public Health  (2015) 15:592 Page 10 of 11Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsAuthors SMH and SAW served as program coordinators for this protocol.All authors, with the exception SMH and SAW contributed to the designof the trial. MRB, RER, PAE, YL, WD are listed as applicants on the CanadianInstitutes of Health Research grant that is supporting this trial. All authorscontributed to, reviewed, and approved the final submitted manuscript.AcknowledgementsFunding for this research was provided by the Canadian Institutes ofHealth Research (Grant # MOP-125911). We would like to acknowledgethe volunteer, peer instructors, without whom this work would not bepossible impossible. We would also like to thank all of our participants.Finally, we would like to thank the volunteer research assistants for theirefforts with functional fitness testing and data collection.Author details1School of Kinesiology, University of British Columbia, Vancouver, BC, Canada.2Department of Human Nutrition, Foods and Exercise, Virginia Tech,Blacksburg, VA, USA. 3Fralin Translational Obesity Research Center,Blacksburg, VA, USA. 4Behavioural Medicine Laboratory, School of ExerciseScience, Health and Physical Education, University of Victoria, Victoria, BC,Canada. 5Department of Cell Biology, Harvard Medical School, HarvardUniversity, Boston, MA, USA. 6Department of Psychology, University ofCalifornia-Riverside, Riverside, CA, USA. 7Department of Psychology,University of British Columbia, Vancouver, BC, Canada. 8Faculty of Education,University of British Columbia, Vancouver, BC, Canada. 9Department of FamilyMedicine, Carilion Clinic, Roanoke, VA, USA.Received: 1 June 2015 Accepted: 4 June 2015References1. Denys K, Cankurtaran M, Janssens W, Petrovic M. Metabolic syndrome in theelderly: an overview of the evidence. Acta Clin Belg. 2009;64(1):23–34.doi:10.1179/acb.2009.006.2. Yazdanyar A, Newman AB. The burden of cardiovascular disease in theelderly: morbidity, mortality, and costs. Clin Geriatr Med. 2009;25(4):563–77.doi:10.1016/j.cger.2009.07.007.3. Paterson DH, Warburton DER. Physical activity and functional limitations in olderadults: a systematic review related to Canada’s Physical Activity Guidelines. Int JBehav Nutr Phys Act. 2010;7(8):1–22. doi:10.1186/1479-5868-7-38.4. Marsh AP, Lovato LC, Glynn NW, et al. Lifestyle interventions andindependence for elders study: recruitment and baseline characteristics. JGerontol A Biol Sci Med Sci. 2013;68(12):1549–58. doi:10.1093/gerona/glt064.5. Colley RC, Garriguet D, Janssen I, Craig C, Clarke J, Tremblay MS. Physicalactivity of Canadian adults: accelerometer results from the 2007-2009 CanadianHealth Measures Survey. Health Reports, Statistics Canada. 2011; Catalogue82-003-XPE. http://www.statcan.gc.ca/pub/82-003-x/2011001/article/11396-eng.pdf. Accessed May 11, 2015.6. Kern ML, Reynolds CA, Friedman HS. Predictors of physical activity patternsacross adulthood: a growth curve analysis. Pers Soc Psychol Bull.2010;36(8):1058–72. doi:10.1177/0146167210374834.7. Canadian Institute for Health Information (CIHI). National Health ExpenditureTrends, 1975–2011. Ottawa, ON; 2011. https://secure.cihi.ca/free_products/nhex_trends_report_2011_en.pdf. Accessed May 11, 2015.8. 2006 to 2101 Population Projections, Australia. Australian Bureau of Statistics.2008; Catalogue 3222.0. http://www.ausstats.abs.gov.au/Ausstats/subscriber.nsf/0/0E09CCC14E4C94F6CA2574B9001626FE/$File/32220_2006%20to%202101.pdf.Accessed May 11, 2015.9. Rogerson PA, Kim D. Population distribution and redistribution of thebaby-boom cohort in the United States: recent trends and implications.Proc Natl Acad Sci U S A. 2005;102(43):15319–24. doi:10.1073/pnas.0507318102.10. Population Projections for Canada, Provinces and Territories, 2009 to 2036.Statistics Canada. 2010; Catalogue 91-520-X. http://www.statcan.gc.ca/pub/91-520-x/91-520-x2010001-eng.pdf. Accessed May 11, 2015.11. Carron AV, Hausenblas HA, Mack D. Social influence and exercise:a meta-analysis. J Sport Exerc Psychol. 1996;18(1):1–16.12. Dishman RK, Buckworth J. Increasing physical activity: a quantitativesynthesis. Med Sci Sports Exerc. 1996;28(6):706–19. doi:10.1097/00005768-199606000-00010.13. Estabrooks PA, Carron AV. The influence of the group with elderly exercisers.Small Gr Res. 1999;30(4):438–52. doi:10.1177/104649649903000403.14. Estabrooks PA, Carron AV. Group cohesion is older adult exercisers:prediction and intervention effects. J Behav Med. 1999;22(6):575–88.doi:10.1023/A:1018741712755.15. Dickens AP, Richards SH, Greaves CJ, Campbell JL. Interventions targetingsocial isolation in older people: a systematic review. BMC Public Health.2011;11(647):1–22. doi:10.1186/1471-2458-11-647.16. Beauchamp MR, Carron AV, McCutcheon S, Harper O. Older adults’preferences for exercising alone versus in groups: consideringcontextual congruence. Ann Behav Med. 2007;33(2):200–6.doi:10.1007/BF02879901.17. Beauchamp MR, Dunlop WL, Downey SM, Estabrooks PA. First impressionscount: perceptions of surface-level and deep-level similarity within postnatalexercise classes and implications for program adherence. J Health Psychol.2012;17(1):68–76. doi:10.1177/1359105311408156.18. Dunlop WL, Beauchamp MR. Examining the relationship between intra-groupage similarity and adherence within group-based exercise programs. Am J PrevMed. 2012;42(1):53–5. doi:10.1016/j.amepre.2011.08.018.19. Dunlop WL, Beauchamp MR. En-gendering choice: preferences for exercisingin gender-integrated and gender-segregated groups and considerationof overweight status. Int J Behav Med. 2011;18(3):216–20.doi:10.1007/s12529-010-9125-6.20. Dunlop WL, Beauchamp MR. When similarity makes a difference: predictingadherence behaviors and cohesion within exercise group settings.Group Dyn. 2011;15(3):258–66. doi:10.1037/a0023642.21. Burton NW, Khan A, Brown WJ. How, where and with whom? physicalactivity context preferences of three adult groups at risk of inactivity.Br J Sports Med. 2012;46(16):1125–31. doi:10.1136/bjsports-2011-090554.22. Dunlop WL, Beauchamp MR. Birds of a feather stay active together: a casestudy of an all-male older adult exercise program. J Aging Phys Act.2013;21(2):222–32. doi:10.1016/j.jhealeco.2011.11.003.23. Turner JC. Rediscovering the social group: a self-categorization theory.New York, NY: Basil Blackford; 1987.24. Hogg MA. Subjective uncertainty reduction through self-categorization:a motivational theory of social identity processes. Eur Rev Soc Psychol.2000;11(1):223–55. doi:10.1080/14792772043000040.25. Tajfel H. Social identity and intergroup behavior. Soc Sci Inform.1974;13(1):65–93.26. Hogg MA. Social identity and the psychology of groups. In: Leary MR,Tangney JP, editors. Handbook of self and identity. 2nd ed. New York,NY: Guilford; 2012. p. 502–19.27. Baranowski T, Anderson C, Carmack C. Mediating variable framework inphysical activity interventions: how are we doing? how might we dobetter? Am J Prev Med. 1998;15(4):266–97. doi:10.1016/S0749-3797(98)00080-4.28. Carron AV, Brawley LR, Widmeyer WN. The measurement of cohesiveness insport groups. In: Duda JL, editor. Advances in sport and exercise psychologymeasurement. Morgantown, WV: Fitness Information Technology;1998. p. 213–26.29. Lawton R, Conner M, McEachan R. Desire or reason: predicting healthbehaviors from affective and cognitive attitudes. Health Psychol.2009;28(1):56–65. doi:10.1037/a0013424.30. Gellert P, Ziegelmann JP, Schwarzer R. Affective and health-related outcomeexpectancies for physical activity in older adults. Psychol Health.2012;27(7):816–28. doi:10.1080/08870446.2011.607236.31. Deci EL, Ryan RM. The “what” and “why” of goal pursuits: human needs andthe self-determination of behavior. Psychol Inq. 2000;11(4):227–68.doi:10.1207/S15327965PLI1104_01.32. Ajzen I. The theory of planned behavior. Organ Behav Hum Decis Process.1991;50(2):179–211. doi:10.1016/0749-5978(91)90020-T.33. Rhodes RE, Fiala B, Conner M. A review and meta-analysis of affectivejudgments and physical activity in adult populations. Ann Behav Med.2009;38(3):180–204. doi:10.1007/s12160-009-9147-y.34. Emmons RA, Diener E. A goal-affect analysis of everyday situational choices.J Res Pers. 1986;20(3):309–26. doi:10.1016/0092-6566(86)90137-6.35. Dumas JE, Lynch AM, Laughlin JE, Smith EP, Prinz RJ. Promotingintervention fidelity: conceptual issues, methods, preliminary results from64. Cole DA, Maxwell SE. Testing mediational models with longitudinal data:questions and tips in the use of structural equation modeling. J AbnormPsychol. 2003;112(4):558–77. doi:10.1037/0021-843X.112.4.558.65. Dunlop WL, Falk C, Beauchamp MR. How dynamic are exercise group dynamics?examining changes in cohesion within class-based exercise programs.Health Psychol. 2013;32(12):1240–3. doi:10.1037/a0030412.66. Muthén LK, Muthén BO. Mplus User’s Guide. 6th ed. Los Angeles, CA:Muthén & Muthén; 2010.67. Enders CK, Bandalos DL. The relative performance of full information maximumlikelihood estimation for missing data in structural equation models. Struct EquModel. 2001;8(3):430–57. doi:10.1207/s15328007sem0803_5.68. Creswell JW. Qualitative inquiry and research design: choosing among thefive traditions. Thousand Oaks, CA: Sage; 1998.69. Schwandt TA. Three epistemological stances for qualitative enquiry:interpretivism, hermeneutics, and social constructionism. In: Denzin NK,Lincoln YS, editors. Handbook of qualitative research. 2nd ed. ThousandOaks, CA: Sage; 2000. p. 189–213.70. Maykut P, Morehouse R. Beginning qualitative research: a philosophic andpractical guide. Lewes, UK: Falmer; 1994.Beauchamp et al. BMC Public Health  (2015) 15:592 Page 11 of 11the early alliance prevention trial. Am J Prev Med. 2001;20(1):38–47.doi:10.1016/S0749-3797(00)00272-5.36. Moher D, Hopewell S, Schulz KF, et al. CONSORT 2010 explanation andelaboration: updated guidelines for reporting parallel group randomisedtrials. Brit Med J. 2010;340:869. doi:10.1136/bmj.c869.37. Warburton DER, Jamnik VK, Bredin SSD, Gledhill N. The Physical ActivityReadiness Questionnaire for Everyone (PAR-Q+) and Electronic PhysicalActivity Readiness Medical Examination (ePARmed-X+): summary ofconsensus panel recommendations. The Health & Fitness Journal of Canada.2011;4(2):3–23.38. Canadian Society for Exercise Physiology (CSEP). Canadian physical activityguidelines and Canadian sedentary behaviour guidelines. http://www.csep.ca/english/view.asp?x=949. Accessed May 11, 2015.39. Shields M, Martel L. Healthy living among seniors. Statistics Canada, Supplementto Health Reports. 2006; Catalogue 82-003. http://www.statcan.gc.ca/pub/82-003-s/2005000/pdf/9086-eng.pdf. Accessed May 11, 2015.40. Bandura A. Self-efficacy: the exercise of control. New York, NY: Freeman; 1997.41. American Council on Exercise. Ace group fitness instructor manual: a guidefor fitness professionals. 2nd ed. Monterey, CA: Healthy Learning; 2007.42. McNaughton SA, Crawford D, Ball K, Salmon J. Understanding determinantsof nutrition, physical activity and quality of life among older adults: theWellbeing, Eating and Exercise for a Long Life (WELL) study. Health QualLife Outcomes. 2012;10(109):1–7. doi:10.1186/1477-7525-10-109.43. King AC, Castro C, Wilcox S, Eyler AA, Sallis JF, Brownson RC. Personal andenvironmental factors associated with physical inactivity among differentracial-ethnic groups of U.S. middle-aged and older-aged women. HealthPsychol. 2000;19(4):354–64. doi:10.1037/0278-6133.19.4.354.44. Statistics Canada. Census 2006 – 2B (long form). http://www23.statcan.gc.ca/imdb-bmdi/instrument/3901_Q2_V3-eng.pdf. Accessed May 11, 2015.45. Godin G. The Godin-Shephard Leisure-Time Physical Activity Questionnaire.The Health Fit J Can. 2011;4(1):18–22. http://www.healthandfitnessjournalof-canada.com/index.php/html/article/view/82. Accessed May 11, 2015.46. Wilcox S, King AC, Brassignton GS, Ahn DK. Physical activity preferences ofmiddle-aged and older adults: a community analysis. J Aging Phys Activ.1999;7(4):386–99.47. Rikli RE, Jones CJ. Development and validation of a functional fitness test forcommunity-residing older adults. J Aging Phys Activ. 1999;7(2):129–61.48. Estabrooks PA, Carron AV. The Physical Activity Group EnvironmentQuestionnaire: an instrument for the assessment of cohesion in exerciseclasses. Group Dyn. 2000;4(3):230–43. doi:10.1037/1089-2699.4.3.230.49. Rhodes RE, Matheson DH. Discrepancies in exercise intention andexpectation: theoretical and applied issues. Psychol Health. 2005;20(1):63–78.doi:10.1080/08870440412331296071.50. Gosling SD, Rentfrow PJ, Swann Jr WB. A very brief measure of the Big-Fivepersonality domains. J Res Pers. 2003;37(6):504–28. doi:10.1016/S0092-6566(03)00046-1.51. Poag-Ducharme KA, Brawley LR. Self-efficacy theory: use in the prediction ofexercise behavior in the community setting. J Appl Sport Psychol.1993;5(2):178–94. doi:10.1080/10413209308411313.52. Bray SR, Gyurcsik NC, Culos-Reed SN, Dawson KA, Martin KA. An exploratoryinvestigation of the relationship between proxy efficacy, self-efficacyand exercise attendance. J Health Psychol. 2001;6(4):425–34.doi:10.1177/135910530100600405.53. Pinel EC. Stigma consciousness: the psychological legacy of social stereotypes.J Pers Soc Psychol. 1999;76(1):114–28. doi:10.1037/0022-3514.76.1.114.54. Harden SM, Estabrooks PA, Mama SK, Lee RE. Longitudinal analysis ofminority women’s perceptions of cohesion: the role of cooperation,communication, and competition. Int J Behav Nutr Phys Act. 2014;11(57):1–8.doi:10.1186/1479-5868-11-57.55. Diener E, Wirtz D, Tov W, Kim-Prieto C, Choi D, Oishi S, et al. New well-beingmeasures: short scales to assess flourishing and positive and negative feelings.Soc Indic Res. 2010;97(2):143–56. doi:10.1007/s11205-009-9493-y.56. Plummer ML, Wight D, Obasi AIN, et al. A process evaluation of a school-basedadolescent sexual health intervention in rural Tanzania: the MEMA kwa Vijanaprogramme. Health Educ Res. 2007;22(4):500–12. doi:10.1093/her/cyl103.57. Collins LM, Murphy SA, Nair VN, Strecher VJ. A strategy for optimizing andevaluating behavioral interventions. Ann Behav Med. 2005;30(1):65–73.doi:10.1207/s15324796abm3001_8.58. Merrick E. An exploration of quality in qualitative research: are “reliability”and “validity” relevant? In: Kopala M, Suzuki LA, editors. Using qualitativemethods in psychology. Thousand Oaks, CA: Sage; 1999. p. 25–36.71. Nelson ME, Rejeski WJ, Blair SN, et al. Physical activity and public health inolder adults: recommendation from the American College of Sports Medicineand the American Heart Association. Circulation. 2007;116:1094–105.doi:10.1249/mss.0b013e3180616aa2.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistribution59. Faul F, Erdfelder E. GPOWER: a priori, post-hoc, and compromise poweranalyses for MS-DOS [Computer Program]. Bonn, GER: University of Bonn; 1992.60. MacCallum RC, Browne MW, Sugawara HM. Power analysis anddetermination of sample size for covariance structure modeling. PsycholMethods. 1996;1(2):130–49. doi:10.1037/1082-989X.1.2.130.61. Preacher KJ, Coffman DL. Computing power and minimum sample size forRMSEA [Computer software]. http://quantpsy.org/. Published November 5,2006. Accessed May 11, 2015.62. Garver MS, Mentzer JT. Logistics research methods: employing structuralequation modeling to test for construct validity. J Bus Logist. 1999;20(1):33–57.63. Guest G, Bunce A, Johnson L. How many interviews are enough? anexperiment with data saturation and variability. Field Methods.2006;18(1):59–82. doi:10.1177/1525822X05279903.Submit your manuscript at www.biomedcentral.com/submit

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.52383.1-0223733/manifest

Comment

Related Items