“It was just for us”: qualitative evaluation of an exercise intervention for African-American couples

Building upon previous couples-based intervention work in husbands and wives [6], the interdisciplinary (exercise science and family science), community-engaged pilot exercise intervention study recruited AA male-female couples who were married and/or cohabiting. Participants were recruited in the greater Knoxville, TN area via radio and newspaper advertisements, flyer distribution, speaking engagements in groups that served AA adults, and endorsements from existing participants. Briefly, the pilot intervention study aimed to use the established relationship between partners to support PA participation. The study examined the effects of prescribed walking for exercise (≥3 days/week, ≥30 min/day) plus supervised total body resistance training (2 days/week) in two randomly assigned groups (ET versus ES, defined above in the Introduction). Although both groups were prescribed the walking plus resistance training program, only the ET group received education during their resistance training sessions on how to facilitate partner support and enhance receptivity to partner support. This element of the program was adapted from the educational component of the Exercising Together® program [21], an existing partnered resistance training program designed to improve teamwork in couples during exercise. The ES group did not receive an educational intervention. All participants wore a wrist-mounted activity monitor and recorded the frequency, duration, and partner status (walked with or without the partner) of their prescribed walking bouts onto log forms daily for the entire 12-week study. Measured outcomes were exercise adherence, the provision of partner support and receptivity to partner health influence, and several cardiometabolic risk factors. Additional pilot study details have been published previously [7, 22].

The ten eligible couples (five couples in each intervention group; n = 10/group) were contacted via e-mail or telephone and invited to participate in the focus group study. Four couples (n = 8) from the ET group participated, while one failed to respond to the invitation after multiple attempts. Four couples were also represented in the ES group; however, this included both partners from only two couples (n = 4), as two individuals (n = 2) attended even though their respective partners had time conflicts. Neither partner from one ES couple attended due to a time conflict. To note, this was also the only couple who dropped out of the intervention due to one of the partner’s non-study-related injury. Focus groups occurred no more than 4 months after couples completed the 12-week study.

The lead investigator (LMH) conducted two focus group meetings – one for each intervention group. To promote a comfortable environment, both focus groups were held in the banquet room of a conveniently-located local restaurant. The focus group procedures and confidentiality were discussed with the group, then participants reviewed and signed an informed consent form. All participants received lunch and a $30 gift card from a national retailer for participation in this evaluation component of the pilot study. All study procedures were reviewed and approved by the University of Tennessee, Knoxville Institutional Review Board.

All participants completed a brief demographic questionnaire and were informed that this information, combined with select data collected during the pilot intervention, would be used as group descriptors. Investigators developed open-ended questions with probes based upon literature related to process evaluation [14, 23] and components of the pilot exercise intervention itself in order to assess several aspects of intervention implementation. A formative approach to the process evaluation was utilized with the goal of using the data to revise and improve the intervention for a subsequent launch on a larger scale [23]. As such, focus group guide questions concentrated on facilitators and barriers (i.e., strengths and challenges), as well as interactions with the research staff and investigators. Investigators then reviewed the questions for face validity (subjective assessment of whether the developed questions would provide the data desired). Discussion between investigators resulted in a 12- (ET group) and an 11-item (ES group) open-ended semi-structured focus group guide (Table 1). Although the context of some questions varied based upon intervention group assignment, the focus groups were run identically. Part of each meeting included a separation of female and male participants to provide individuals the opportunity to provide feedback about their experience in the pilot intervention study without the concern of censoring information due to the presence of their partner. These brief break-out sessions were facilitated by gender-matched study staff.

As recommended by Krueger and Casey [24], two research staff trained in focus group methodology recorded observations and took notes during each focus group. The lead investigator initiated the focus groups with a description of the focus group process and the purpose of the qualitative evaluation. Both focus groups were digitally audio recorded and lasted approximately 90 min. Multiple research staff members transcribed the digital audio recordings verbatim and double-checked for accuracy. The lead investigator then conducted a final accuracy check of both transcripts compared to the audio recordings.

Based on the semi-structured focus group guide, two investigators (LMH and CSB) used NVivo 12 software (QSR International; Doncaster, Australia) to analyze the qualitative data using an open coding approach based on thematic analysis [25]. The two investigators independently assigned codes to concepts, opinions, or ideas that emerged from the data [25]. The coding process was iterative and collaborative, and started with a planning meeting to review procedures related to thematic analysis and determine a timeline for the analysis. The investigators held four consensus meetings. At the first two meetings, investigators presented the codes, categories, sub-categories, and the corresponding rationale that they independently identified during their review and abstraction of the de-identified focus group transcripts. Throughout the process, the investigators independently developed provisional codebooks, which contained code names and descriptions, categories, subcategories, and examples for each. The provisional codebooks were used to discuss agreements and discrepancies between coding. During the third meeting, the investigators continued to independently review, abstract, and code both focus groups. At the final meeting, the researchers finalized their rationale and merged their provisional codebooks. The final codebook contained themes identified by both investigators.

All participants who attended the focus groups completed the 12-week exercise intervention in its entirety. In aggregate prior to entering the intervention, participants were 64.7 ± 6.8 years of age, classified as obese (body mass index 30.4 ± 4.8 kg/m²), and low active based on an average of 6292 ± 1615 steps/day [26]. The majority of participants (86%) reported some college education, most had part-time or full-time employment (64%), and all reported a gross annual household income of $50,000 or greater. All participants were married. Table 2 shows descriptive characteristics of participants by group.

Data analyses revealed three key themes: 1) value and benefits of intervention participation, 2) difficulties with intervention participation, and 3) suggestions to improve the intervention. Table 3 presents the identified themes with subsequent categories and sub-categories by focus group.

Participants in both groups discussed how the pilot study helped encourage them and to prepare mentally for engagement in exercise. The participants also expressed appreciation for the interactions they had with the study staff while engaging in the exercise activities because they were courteous, supportive, and “refreshing.”

Participants also mentioned how sharing the study experience with their significant other helped facilitate discussions about exercise with their partners.

Only the ET group discussed competition between partners. Participants in the ET group reported that exercising together spurred friendly rivalry in completing the prescribed exercise activities and/or re-ignited their competitiveness.

Only the ET participants discussed new knowledge about and experiences within their Neighborhood and Community and highlighted improvements in their exercise Self-Efficacy.

The groups differed in this category as the ES group discussed self-motivation to attend trainings and inconsistent training instructions as perceived challenges during their participation.

Participants in both focus groups were greatly invested in the pilot intervention and therefore, were candid when making suggestions on how to improve it. Categories identified in both groups under this theme were Programming, including calls for more flexibility training and a longer program duration, and Enhanced Transition Support as participants graduated from supervised training to independent exercise at their respective fitness facilities.

Both intervention groups identified the accountability and structure, encouragement to start a fitness program, and the shared experience with their significant other as key strengths of the pilot study. Research examining PA barriers in AA women and men has cited lack of time, lack of motivation, and lack of a PA partner/social support as issues [27, 28]. The current participants’ specific examples of how study participation required them to prioritize exercise over a myriad of other daily activities, even when they were not intrinsically motivated to do so, is important as it addresses these barriers. Cultural relevance and study staff were strengths identified by participants that addressed additional cited barriers to PA including a lack of physically active role models and lack of knowledge about how to use exercise equipment [28]. Based on the responses of the focus group participants, there appeared to be a level of overlap between the cultural relevance and the study staff categories, as the majority of the study staff and all of the exercise trainers were AA. These findings will inform the refinement of this intervention, in particular, the importance of having cultural relevance woven throughout the intervention and the use of AA study staff. This aligns with current literature that speaks to the importance of cultural relevance in research [12, 13].

One less expected theme to emerge was that the ET group identified competition between the partners as a study strength. The authors speculate that ET was the only group that mentioned competition between partners given the inherent dynamic that included direct and intentional interpersonal interactions during exercise. This finding suggests that for self-proclaimed competitive personality types, an underlying sense of rivalry may be elicited despite education during their supervised resistance training sessions on how to encourage each other and work as a team. That said, the competitive dynamic noted by some appeared to be affable overall and therefore, did not seem to negate the positive relational benefits of the study. Future research may wish to examine the effect of a competitive relationship and exercise adherence between partners. What may be key is whether spouses share a similar competitive approach to exercise, as discrepant views in this regard may diminish the benefits of joint exercise [29].

Although not the focus of the current evaluation, investigators note that participants provided limited qualitative impact evaluation data related to the study’s influence on the relationship dynamics. Both groups noted feelings of having achieved a closer relationship with their partner after participating in the intervention. However, the investigators found it interesting that the dynamic between the two intervention groups appeared to influence participants’ perceptions of the intervention’s benefits. The ET group acknowledged the study as both exercise and an opportunity to regularly connect with their partner. This connection was described both as time spent together completing the prescribed exercise and time spent participating in typical activities of daily living. It was apparent that this time spent together was valued and special to the ET group, particularly the women. All four of these women identified exercising together during the intervention as one of the few activities, if not the only one, that the couple did together just for themselves. The finding that women, in particular, noted the relationship benefits of exercising together is consistent with previous research [30] that examined the relational outcomes of couples who participated in the Exercising Together® program [21, 31]. Partners in this randomized controlled exercise trial, comprised of males recovering from prostate cancer and their female spouses, appeared to derive different benefits from exercising together. Only the females in the study showed improved relational functioning (e.g., increased affection) as a result of exercising together, whereas there were no changes for the male partners. The authors concluded that women may benefit from collaboration and be particularly attuned to their relationship functioning. Thus, women in the Exercising Together® study and the current one, may have appreciated the intervention as an opportunity to dedicate time and energy to renewing their bond with their spouse.

Participants in the ET group also disclosed feelings of long-term marriages having “died down,” but that the exercise study introduced a new and positive element into the relationship. This line of conversation did not surface in the ES group. In contrast to the ET group, the ES group appeared to view the study largely as an exercise intervention and mainly focused on physical health benefits, despite having referenced some tangential relationship benefits. Although it is plausible that the longer relationship duration of the ET group compared to the ES group (44.3 vs. 32.5 years, respectively) may explain why the former was reporting more of a relationship lull, recent longitudinal work on couples suggests that there is little evidence of deterioration in marital quality over the marital life course [32]. Thus, we suggest that it is unlikely that there were significant a priori differences in the relationships of couples assigned to the two groups, but instead speculate that the experience of exercising together may have drawn those partners’ attention to their relationship in a reflective and seemingly beneficial manner.

The ET group also noted opportunities to explore their neighborhoods and the surrounding community, a heightened awareness of nutrition habits, and self-efficacy as benefits of participation in the pilot study. Investigators reference the discussion above regarding the ES group viewing the pilot study with a more straight-forward lens (i.e., exercise intervention) to offer speculation as to why these benefits did not surface in the ES group. It may be that in being directed to focus not only on their own progress but their partner’s as well, the ET participants expanded their focus beyond exercise to think about health and well-being more holistically (e.g., nutrition) and more collectively (e.g., spouse, the community).

As wearing a PA monitoring device has been linked to improvements in PA [33] and self-monitoring has been clearly shown to promote behavior awareness (and likely to contribute to successful behavior change) [34], investigators are somewhat discouraged that the two monitoring strategies employed by the pilot intervention were viewed by participants as challenging. The functional issues experienced when using the consumer-grade, objective activity monitoring devices chosen for the study were unexpected, but still provoked a negative study-related experience for participants and staff. The device was chosen for its wear ease (wrist-mounted) and validity [35]. However, most participants recommended that model and/or the manufacturer not be used again. It is noteworthy that ES participants reported their desire to help the lead investigator’s research agenda as a motivator to record their walking activities, even though they found it to be a cumbersome study-related task. This demonstrates that community buy-in includes PA research conducted by AA researchers. Investigators are considering more convenient ways to acquire reports of daily walking that could replace the written log in the future.

Although relatively minor physical challenges were noted in both groups and are not uncommon in aging individuals, two ES participants perceived the exercise activities as something that they needed to push themselves to do. Even so, those participants and others expressed gratitude to be able to rely on the structure of the study to facilitate exercise participation and eliminate excuses. Diminished motivation did not emerge in the ET group. These accounts may be supported by data that link exercise enjoyment and group dynamics (between the participant, the exercise leader, and other participants), showing that enjoyment and intention for continued participation is optimized in the presence of a supportive group environment that is directed by a positive, supportive leader [36]. The ES group only interacted one-on-one with their exercise trainer during resistance training sessions and were prescribed to complete their walking alone, whereas ET always had a partner with them. That said, it could be argued that social support was also provided to the ES participants given partners entered the pilot study together. The current investigators will continue to explore how different configurations of dyadic engagement can affect exercise participation and adherence.

The suggested intervention improvement that was clearly the most urgent to participants was the call for more support when transitioning from the study to independent exercise. One participant candidly expressed feeling dropped “like a hot potato.” Although multiple mechanisms were in place that aimed to assist participants in their transition (e.g., 12 weeks of guidance to build exercise knowledge and confidence, paid commercial fitness facility memberships that provided cost-free small group classes and personal training sessions, keeping the study-distributed activity monitors), participants would have likely benefitted from a more gradual/guided change to their new exercise venues and more pointed discussions about transitioning to independent regular exercise. A common sentiment among participants was the desire to be held accountable for exercising after the 12-week program ended. Investigators agreed with participants that this could be achieved relatively easily by sending texts or e-mail reminders to exercise, and/or having study staff continue to remotely monitor activity through their activity monitors. Evidence related to exercise interventions in older adults suggests that these and other strategies can be successful in enhancing exercise sustainability, however more long-term research is needed in this area [37]. Additionally, participants expressed a desire to receive more flexibility training as a part of the study. As it stood, participants completed total body stretching with the assistance of their exercise trainer (ES) or their exercise trainer and their partner (ET) for 5–10 min immediately following their resistance training sessions. They were also encouraged to stretch independently after their walks, although no formal accountability was employed. As flexibility training is unquestionably beneficial to maintain range of motion as individuals age [38], investigators will prioritize more intense flexibility training during supervised sessions and accountability for home-based flexibility training in future studies in this population.

Community engagement and education were also prominent suggestions for program improvement posed by the ET group, yet were not mentioned in the ES group. Again, this may link to the ES participants’ more narrow perception of the pilot study solely as an exercise intervention, versus a more holistic perception of the study experience that seemed to persist in the ET group. As investigators developed the study using a community-engaged approach by consulting with a community advisory board [22], participant-generated ideas to broaden the local reach of the intervention are instrumental and highly valued to help maintain the intervention’s community ties. In addition, discussions about adapting the intervention to take place in the community (versus the laboratory gym) occurred in the ET group. Participants also expressed interest in additional background knowledge about the study methods as a means to better motivate themselves to comply with study procedures, namely, completing and submitting the walking logs. Nutrition education was also requested in terms of making healthy choices in general, as well as when and what types of foods to eat to fuel workouts. The investigators note that the addition of an educational component (outside of the ET group’s team-building education during workouts) was also suggested by our community advisory board, yet could not be executed in the pilot intervention study due to limited staff and resources.

The investigators acknowledge this study had limitations. First, given the overarching goal of seeking information about a specific pilot intervention study, the sampling strategy that targeted the 10 couples who entered the pilot study presented an inherent drawback of a small target participant pool of 20 potential participants (i.e., 10 participants per intervention arm). Investigators intended to invite all participants who entered the study to participate in subsequent focus groups; however, all couples who entered the pilot study after those invited to the current focus groups had their pilot study participation disrupted due to the COVID-19 pandemic. Consequently, investigators conducted only one focus group per intervention arm and the sample size was small. This hindered investigators’ ability to make comparisons within each intervention group and could have affected the results obtained. Investigators also noted that the ES group was less represented than the ET group. Even so, investigators were satisfied that saturation was achieved as similar themes surfaced between the two groups.

The investigators also recognize that more reserved participants may be hesitant to contribute to a focus group discussion. As such, these participants may not fully engage in the discussion and allow others to control the conversation. To minimize this, the focus group moderator monitored participant contributions and involved all participants. That said, the moderator observed that the small sample size appeared to organically enable a more intimate discussion during which all individuals willingly participated. Finally, investigators’ experiences can bias the analysis and interpretation of qualitative data. To prevent potential bias, notes taken at the focus groups by trained note takers supplemented the transcriptions ensured the context and accuracy of the data. Additionally, holding multiple consensus meetings with study staff that both attended and did not attend the focus group meetings allowed for the confirmation of data accuracy and analysis. Examination of data accuracy and confirmation of data interpretation helped to establish the credibility of the data and the study’s trustworthiness [39]. Although these data are not generalizable to all AAs or older adults, these findings may be transferable to comparable older AA adults living in the Southeastern region of the U.S. Despite these limitations, this study is valuable as it begins to establish a line of research that examines a novel, dyadic approach to exercise promotion in the AA community.