Study design and power calculation
All phases of this research were approved by The University of Southern Mississippi's Institutional Review Board, and informed consent and a medical disclaimer were obtained from all participants upon enrollment into the study. To provide benefit to all community participants and advance the scientific rigor of CBPR trials, this CBPR study utilized a 2-phase design approach. The first 6-month intervention phase was a quasi experimental design to assess the effectiveness of intervention treatment effects on SBP. During this phase, all enrolled walking participants were offered the same 6-month intervention which included social support provided by walking group volunteer leaders (designated "coaches"), motivational interviewing provided by intervention staff, pedometer diary self-monitoring, and five monthly education sessions. The second 12-month maintenance phase is a randomized controlled trial to assess the treatment effects of a low versus high dose of motivational interviewing delivered via telephone. Over the 12-month maintenance phase, the low treatment arm will receive four additional telephone motivational enhancement sessions while the high treatment arm will receive 10 additional motivational enhancement sessions. Between the two groups, the motivational enhancement sessions are designed to be of similar content, quality and session length.
This study was powered to test the treatment effects of the motivational interviewing dose at 18-months on SBP, while controlling for 6-month intervention treatment effects. Anticipating a 20% attrition rate at 18-months, 267 participants were enrolled. The projected sample at 18-months (N1 = N2 = 106) provides 80% power to detect a moderate effect size of 0.4 [difference of 6 (SD = 15) mmHg between groups] with an alpha of 0.05.
Hattiesburg community & targeted population
Hattiesburg, located in southeast Mississippi, has a population of approximately 45,000 residents. The median household income of Hattiesburg is $24,409, which is lower than state and national averages at $31,330 and $41,994, respectively . The city is approximately 47% African-American and 49%White. Twenty-four percent of deaths among non-whites in 2007 in Hattiesburg were attributable to heart disease, cerebrovascular disease, and stroke, but no incidence or prevalence data are available for these or other conditions at the city or county level . For non-whites in the nine county southeast Mississippi public health district in which Hattiesburg is located, the prevalence rate for hypertension in 2007 was 43%.
Recruitment efforts were primarily directed toward African American residents; however race/ethnicity was not an exclusion factor. Eligibility criteria included 18 years of age or older, English-speaking, non-institutionalized, and residing in the Hattiesburg area. Participants with blood pressure over 180/110 were directed to obtain immediate medical attention and were disqualified from participating; however, all other participants were eligible for study participation regardless of blood pressure status and medication regimen.
Engagement of community members and the community-based participatory process
HUB City Steps was planned in the context of a community-wide wellness initiative, Get Healthy Hattiesburg (GHH), led by the mayor's office. One of the priorities of the GHH steering committee was to increase physical activity in the community. The city of Hattiesburg and its GHH steering committee collaborated with the University of Southern Mississippi in preparing the grant application that funded HUB City Steps. University researchers shared with the GHH steering committee information about the design and outcomes of a feasibility study and pilot test of a community walking intervention successfully conducted in a rural Mississippi community [20–24]. Together, these collaborators adapted the 6-month quasi-experimental intervention design previously used by proposing in the grant application to target members of neighborhood associations in majority African American city wards, a recruitment approach that was already being used by GHH. When the HCS CBPR intervention project was funded, GHH steering committee members were invited to become part of a community advisory board (CAB) for HCS. In configuring the HCS CAB, HCS project staff also recruited as members other individuals who were engaged in community networking and advocacy/service provision within the targeted African American community in Hattiesburg, and/or in health promotion in the broader community. Invitations were issued to an initial organizational meeting that took place in November 2008, at which attendees were oriented to the concepts of CBPR and to the pilot intervention. Membership stabilized at 21 members, who are affiliated with local city and county government (n = 5), public and private health and medical clinics and agencies (n = 9), organizations with an educational mission, including the local school district and the Extension Service (n = 4), and private non-profit community organizations (n = 3). Twenty members are female, 14 African American (AA) and 7 White, and the median age range of members is 31-40.
During its first seven months of monthly meetings, the CAB created a community capacity building action plan for physical activity, and as part of that plan, completed such tasks as creating a vision and mission statement, naming the intervention, and designing a logo. The community capacity building action plan, described elsewhere, utilized Lempa and colleagues' framework for community capacity and applied lessons learned related to community capacity through the prior pilot intervention. The CAB also took on the responsibilities of providing oversight and guidance related to intervention implementation, evaluation, and planning for sustainability. As part of the action plan, the CAB identified community gatekeepers related to the HUB City Steps intervention. These were individuals who were perceived by the CAB as having the potential to influence the development or success of the intervention, by virtue of their holding positions of respect or influence within health, government, education, neighborhood, social service and church settings in the African American community or the larger geographic community of Hattiesburg. For the purpose of intervention adaptation and tailoring, community gatekeepers were invited to attend one of three community conversation workshops held during July-August 2009, and a similar workshop was held with members of the Community Advisory Board. The workshops were facilitated by a research staff scientist with extensive experience in community health planning, and used a focus group type methodology (unpublished data). The community conversations elicited input on a variety of aspects of intervention planning, including recruitment of walking coaches and participants, retention of participants, intervention design, scheduling and format of education sessions, and data collection procedures. Participants in the four community conversation sessions numbered 25, of which 20 were African American and 23 were female. Results summaries were provided to the research staff and recommendations were incorporated into protocols and manuals of procedures. For example, the maintenance phase design was modified to randomize participants to a low or high dose motivational interviewing-based telephone follow-up, rather than randomizing half the participants to a reminder-only maintenance group to receive quarterly calls about health fair opportunities.
In addition to CBPR and community capacity, social support and motivational interviewing frameworks were used to undergird this project. These frameworks share something of a common orientation in that each focuses away from the trained professional and towards the individual or community owning the problem or challenge. For examples, social support emphasizes the role of family, friends, and community to provide comfort and encouragement[18, 26], and MI calls forth the individual's problem-solving skills to address behavior change. Research indicates that social support predicts increased physical activity either directly or indirectly through its impact on self-regulation and self-efficacy[27–31]. Including social support as a component in community physical activity interventions has been strongly recommended by the Task Force on Community Preventive Services based on a systematic review of physical activity interventions[32, 33]. Social support has also been shown to be an indirect predictor of maintaining long-term physical activity and has been found to be an important psychosocial construct in behavioral intervention research in African American populations[34–36].
Motivational interviewing (MI) is a collaborative, participant centered counseling approach aimed at strengthening internal motivation to change a behavior[19, 37]. Developed as a non-confrontational style, MI represents an important shift in addressing behavioral change by focusing on evoking an individual's reasons, needs, and desires for change while recognizing that ambivalence about change is natural and must be resolved by the individual. MI has been studied extensively and shows promise as an efficacious intervention in diverse settings and with an assortment of health issues. Although these studies generally support the use of MI as an approach for improving dietary and physical activity behaviors, there are some limitations to the existing body of research. For instance, more examination of MI's efficacy relating to health behavior promotion among low income, minority, and rural participants is warranted. Furthermore, little is known about what MI dose is needed to promote and sustain health outcomes, and there are no known empirical studies that have examined the effectiveness of different MI doses.
Researchers, intervention staff, and delivery agents
The research staff included contributions from eight doctoral level researchers (KY, JZ, CC, BW, MM, EM, CL, MW) and three master's level researchers (DC, AS, KZ). The community intervention staff included three full-time master's level African American women including an intervention coordinator who served as the primary liaison between the CAB and academic team, and also supervised the coach recruitment and retention coordinator and participant recruitment and retention coordinator.
Outcome data collections were staffed by a variety of individuals from both community and academic settings. Seven AA and one White woman were employed as community research assistants, serving as greeters, check-in personnel and escorts during data collection sessions, and providing child care as needed. They underwent a two-hour training session which included content on intervention goals, methods, and schedule, as well as participant confidentiality.
Anthropometric, biological, and fitness variables were assessed by trained exercise science and kinesiotherapy students (n = 18; 9 male; 2 AA, 15 White, 1 Hispanic). All students had completed a credit course in Exercise Testing and Prescription which included skill development, practice and evaluation of competency by the course instructor for all anthropometric measures, blood pressure, and the 6-minute walk test. All data collectors were specifically trained on study assessment protocols and required to demonstrate proficiency to the data collection coordinator for this area. Total training time was approximately three hours. Four of the 18 students were also trained to assess blood parameters using the Cholestech analyzer. In a 60-minute session, they were instructed on use of the Cholestech and in sterile procedures in the collection of capillary blood and handling of hazardous waste. Each student was required to perform 2-3 analyses on other individuals before being certified to perform this assessment during study data collection.
Six psychology graduate students (5 female; 1 AA, 5 White) served as interviewers, along with project research staff. All interviewers received six hours of classroom training and interviewing practice with other interviewers as well as engaging in home study. Training content included general interviewing skills as well as instrument-specific training, with a question by question review and use of response hand cards, on all psycho-social and demographic questionnaires and the food frequency questionnaire. Each interviewer was observed by the trainer while completing a minimum of three practice interviews. A booster session was held prior to each subsequent data collection.
Five doctoral level psychology graduate students and three master's level registered dietitians provided the MI. Seven were female, two AA and six White, with an average age of 28.1 (SD = 4.7). The counselors had 2.1 (SD = 1.9) years experience providing general clinical services. Training of MI counselors followed suggestions offered by Madson, Loignon and Lane and consisted of 16 hours of didactic and experiential activities aimed at building skill in MI. Finally, the MI counselors conducted three audiotaped mock sessions following the study protocols, which were reviewed by fellow MI counselors and by the MI Intervention Director (a member of the Motivational Interviewing Network of Trainers) to assess counselor fidelity to the protocols and to provide formative feedback. Throughout the intervention fidelity was monitored through multiple means. First, participants completed the Client Evaluation of MI which assesses the client perception of a counselor's MI behavior. Next, MI counselors completed a session specific checklist of activities they were to implement in each meeting with a participant. Finally, MI counselors completed the Therapist Evaluation of Motivational Interviewing which aims to assess the counselors' perception of MI consistent behavior during each counseling session. Feedback was also provided throughout the intervention about fidelity to the protocol based on review of these measures.
Education sessions were led by four community health professionals (all female, two AA, two White) trained to conduct the nutrition education component of each session, assisted by two fitness instructors (both AA) who led the exercise component of each education session. The nutrition education session leaders, who were professional health educators or registered dietitians, participated in five hours of training conducted by research staff which focused on session content and its conceptual underpinnings, session format, and the rationale and procedures for maintaining fidelity to the session content. Each session leader practiced conducting an education session while adhering to the fidelity monitoring guide for the session and being observed and critiqued by her peer session leaders and training staff. Because each education session was delivered three times, research staff fidelity monitors provided formative feedback after the first of the three sessions.
Coach recruitment & training
The community intervention staff recruited walking group leaders using flyers and word-of-mouth, assisted by the CAB, City of Hattiesburg collaborators, and community agency contacts. Beginning August 1, 2009, an online application was posted through the university Human Resources website to allow all individuals interested in becoming walking coaches an opportunity to apply for the position. Initially, 28 coaches were hired. Of those, 26 completed the orientation session. At the end of the intervention participant enrollment, 21 of the originally hired coaches remained and three replacement coaches were added, yielding 24 walking coaches (23 female; 23 AA, 1 White) when the intervention began. Walking coaches participated in 16 hours of training, which included content on program goals, intervention design, protocols and procedures, walking coach responsibilities and compensation, group leadership and motivation, and CPR and first aid.
HUB City Steps was publicized broadly to the target community over the six months preceding the intervention kick-off. Community buy-in and awareness was aided through efforts of community intervention staff, city of Hattiesburg partners, CAB members, walking coaches, and other community stakeholders. Information about HCS was shared with community members in a number of ways, which included word of mouth and distribution of brochures and flyers by personal contact and through participation of HCS staff in other scheduled community events such as Night Out Against Crime and community health fairs. A HCS flyer provided to all potential participants during community events and recruitment activities conducted by staff and coaches used a Frequently Asked Questions format, describing each component and expectations for participation. The two phase approach was described, including the 6 month intervention phase and the assignment to one of two follow-up approaches during the 12 month maintenance phase.
As part of the walking coach training, each coach received training on participant eligibility and recruitment strategies, and was directed to recruit 10-12 persons to serve as walking participants on his/her team. Research staff and CAB members helped to identify potential participants. During the recruitment process, research staff compiled a membership list for each walking group, consisting of potential participants who had been nominated by a coach, staff member, or stakeholder, or who had self-identified as interested in participating. Final eligibility was determined using a pre-enrollment screening form, which was collected from each potential participant by her/his coach, and reviewed by project staff to determine eligibility and need to obtain a medical release to participate, required of anyone who answered yes to one or more of five screening questions.
Coaches group leadership and social support
During the recruitment phase, walking coaches identified and contacted potential group members, informed them, as part of the informed consent process, of the responsibilities of joining HUB City Steps, provided and collected pre-enrollment screening forms, and advised and reminded group members regarding making enrollment appointments. Once participants were enrolled, coaches performed the following functions for their walking groups: contacted each group member weekly to encourage routine walking; worked with each participant to set individual weekly walking goals; encouraged each group member to walk throughout each week; arranged group walking and health-related activities at their discretion (minimum of two per month required); monitored each group member in completing and submitting walking records; notified group members about upcoming educational sessions and data collection health assessments; served as a liaison between walking group members and project staff by conveying any questions or problems to the staff. The coaches were also required to perform all the responsibilities expected of participants, including submitting walking logs and attending data collection health assessments and education sessions.
Coaches were paid at a rate of $15.00 per hour for training, recruitment of group members up to a total of 15 hours, and for walking group leadership activities up to 8 hours per week during the intervention phase. Coaches maintained participant contact logs, and submitted biweekly time sheets to the community coordinator. Coaches were provided with biweekly progress reports on their group members' activity which were generated from the submitted walking logs.
Pedometer diary self-monitoring
Each participant received a Yamax pedometer (Yamax model SW-701, Yamax corporation, Tokyo, Japan), and the intervention staff provided face to face instruction on pedometer use and individualized stride calibration. Participants were instructed to wear the pedometer on their waist during waking hours, to remove only upon showering, bathing or swimming, and to reset the pedometer to zero each morning. Participants had the option of recording their daily steps on weekly pedometer diary postcards or recording daily steps through logging on to the intervention's website. Coaches were responsible for encouraging their team members to record their step counts and submit their weekly diaries. Monthly step count reports that illustrated each team members' step count, as well as the total team's step count, was generated and distributed to each group at the monthly education sessions.
The monthly education sessions lasted approximately 90 minutes and included a group physical activity, sharing time for successes and challenges, education focused on the principles of the DASH diet, and time for questions and evaluations. Each monthly session was offered three times at varying times, days of the week, and location to accommodate participants' schedules and encourage broader participation. The Transtheoretical Model of Change, and most notably the processes of change, were used to guide development, execution, and evaluation of the education sessions. In brief, the monthly education sessions focused on: 1) the negative side effects of high blood pressure and the benefits of exercise in controlling BP; 2) the DASH diet and community options available for exercise and healthy eating; 3) problem solving techniques related to cooking and the role of sodium in controlling blood pressure; 4) empowering the participant to implement healthy lifestyle changes; and 5) healthy choices when eating out and identifying triggers that may influence unhealthy menu choices. A sixth monthly session was a celebration event. Attendance was recorded at each session and session evaluations were completed by attendees.
The outcome measures described below were framed as a health assessment for the participants. Immediately upon completing the health assessment, a 'Know Your Numbers' card was provided to each participant which detailed individual blood pressure, anthropometric and biological results, and dietary intake. This card served as a central point of discussion for MI. Participants engaged in a one-on-one MI session aimed at building participant internal motivation to adopt health eating and exercise behaviors and to develop an individualized change plan at baseline, which was reviewed and updated at subsequent data collection time points.
All outcome measures are scheduled at baseline, 3 months, 6 months and 18 months (or 12 months post-intervention). A data collection manual of procedure was developed to detail standardized guidelines for assessing all outcome measures. Data collection occurred at a local community center, conveniently located for study participants.
Medical history, background, and dietary intake
A medical history questionnaire was used to assess diagnosis and prescribed medications related to blood pressure, blood sugar and cholesterol. A medication adherence questionnaire was also added at the 3-month time point. Additional background information included fasting and smoking questions along with demographics. The National Cancer Institute's (NCI) Five-Factor Screener was used to monitor dietary intake. This valid 18-item screener approximates intake of fruits and vegetables servings, grams of fiber, teaspoons of added sugar, milligrams of calcium, and dairy servings[47, 48].
Anthropometric, biological, and fitness variables
Blood pressure, height, weight, body composition, lipids, glucose, and the 6-minute walk test were assessed. Systolic and diastolic blood pressure was assessed using an OMRON HEM-907XL automatic inflation sphygmomanometer. Height was measured using a portable stadiometer and a Tanita Body fat analyzer model TBF-310T was used to measure weight and percent body fat and to calculate body mass index (BMI). Non-fasting cholesterol, triglycerides and glucose were assessed using the Cholestech LDX Lipid Analyzer. As a measure of fitness, the 6-minute walk test was completed. The 6-minute walk test (6MWT) is an objective, simple, inexpensive, and relatively safe exercise test and has been identified as the test of choice when using a functional walk test for research purposes[49, 50]. The self-paced 6MWT assesses the submaximal level of functional capacity by having participants cover as much distance as possible on a hard, flat, 120 meter circuit within six minutes without running. The 6MWT has been shown to have high reliability and high ability to discriminate between functional levels in a high-function population.
Previously validated psychosocial instruments were used and adapted for this study and included social support for physical activity, treatment self-regulation for physical activity and treatment self-regulation for diet, and processes of change for physical activity. The adapted instruments included physical activity social support from coaches (11 items, 3-sub-scales: guidance, reliable alliance, reassurance of worth); physical activity social support from walking group members (12 items, 3-sub-scales: guidance, reliable alliance, social integration); treatment self-regulation for diet and physical activity (15 items each, 4-sub-scales each: amotivation, external, introjection, identification and integration); and processes of changes (30 items, 10 subscales: stimulus control, social liberation, reinforcement management, helping relationships, counter conditioning, self-liberation, self-reevaluation, environmental reevaluation, dramatic relief, and consciousness raising). Two phases of formative testing were conducted with individual's representative of the target population on the adapted psychosocial instruments, including cognitive testing with five individuals and pilot testing with 20 individuals. Feedback resulted in wording, question structure, and format changes and promoted cultural appropriateness of survey items for use with the target population. Cronbach's alphas and item analysis statistics were used to evaluate the internal consistency of the psychosocial instruments on the baseline data. The majority (24 of 29) of scales and subscales demonstrated strong internal consistency with Cronhbach's alpha ≥0.70 when item composition was maintained from the intended item clustering. Moderate internal consistency ranging with Cronhbach's alpha from 0.44-0.63 was noted for five scales including amotivation for physical activity, amotivation for diet, dramatic relief, self-liberation, and social liberation.
Descriptive statistics including means, standard deviations, frequencies, and percents were used to summarize demographics, participation and compliance rates, and outcome variables. Paired t tests were used to examine the difference in blood pressure between baseline and 3-month follow-up among all study participants. A further subgroup analysis was conducted to examine the difference in blood pressure among participants with normal blood pressure, participants diagnosed with high blood pressure and taking medication, and participants diagnosed with high blood pressure but not taking any medication in the past 30 days. All statistical analyses were performed using the SAS 9.1.3 statistical software package (SAS Institute Inc., Cary, NC). A critical value of .05 was adopted for significance testing in bivariate comparisons.