A Thematic Analysis of Social Identity and Injury in CrossFit®

The purpose of this study was to explore the viability of the social identity approach as a 19 theoretical framework for examining injury in the context of a group exercise program, 20 CrossFit®. Specifically, we sought to identify values of group exercise participants relevant to 21 overuse risk behaviors as well as participants’ responses to criticisms about injury. Via thematic 22 analysis, observations of a CrossFit® setting (N = 31) and interviews of members (N = 14) 23 yielded three social identity content (i.e., Being Hard Core, Achieving Results, Camaraderie). 24 Behaviors employed to enact these social identity content (e.g., engage in frequent, high- 25 intensity workouts; attend despite low-level pain; encourage others to continue despite pain; 26 withhold pain reports from group leaders) enabled members to obtain positive evaluations or 27 avert negative evaluations of group members yet also incurred higher overuse injury risk. We 28 also identified two prominent types of responses of CrossFit® members to criticisms about 29 injury in CrossFit® activity: Compare dimensions (e.g., how well members handled the injuries; 30 the effort they put into prevention; health benefits; strength gained) of the group which were 31 perceived as superior to other contexts, and denounce critics. These response types were 32 interpreted to reflect social creativity and polarization, respectively. Altogether, the findings 33 indicate that group-based psychological factors contribute to overuse injury, advancing previous 34 literature in which intra- and inter-personal factors were the primary focus. This study 35 contributed to the literature by identifying theory-based injury risk factors in group exercise 36 contexts which may inform future injury-prevention interventions. 37

SOCIAL IDENTITYAND INJURY IN CROSSFIT® 5 an exercise program/class) whose members perceive themselves to be similar to each other in 86 meaningful ways through shared values, beliefs, attitudes, and/or behaviors (Jetten et al., 2017). 87 Social identity content refers to shared values that underpin group membership (Evans et  group of university students, binge drinking may occur (Livingstone & McCafferty, 2015). In 99 this way-and similar to the social identity content underpinning our running group example-100 social identities may become a curse that threatens and potentially harms group members' health 101 and well-being (Jetten et al., 2017). Building on this, we proposed that negative social identity 102 content may influence engagement in injury-risk behaviors in group exercise contexts. 103 To examine psychological factors specific to injury in group exercise contexts, we chose were on par with that of sports participants (e.g., powerlifters, elite gymnasts; Montalvo et al.,112 2017). Some critics of Crossfit® have expressed concern about the amount of involved injury 113 risk (Diamond, 2015). In contrast, CrossFit® members find that the modalities of the CrossFit® 114 context, along with the atmosphere and connectedness, contribute to physical activity adherence 115 (Bailey et al., 2017). As such, CrossFit® members may perceive the criticisms of the injury rate 116 of CrossFit® to be threatening to their group. In other words, within the social identity 117 approach, social threats involve negative evaluations of a social identity group such that 118 members, feeling that a source of positive self-esteem is threatened, may be incited to defend 119 their group (Brown & Ross, 1982;Evans et al., 2016). 120 The overarching purpose of this study was to apply the social identity approach to the research questions, rather than philosophical concerns (e.g., nature of reality and knowledge), is 132 a focus. Thus, researchers are called upon to consider the information and beliefs that informed 133 their methodological choices, weigh the consequences, and adjust accordingly until they form a 134 warranted belief that the method is suited for answering the research questions. 135 The first research question required a means for identifying values of a CrossFit® group. 136 In previous studies, the values of groups-to infer social identity content-were pre-identified 137 by researchers, or statements of group leaders were used to identify meaningful social identity  Therefore, we adopted the recommendation of Evans et al. (2016) by employing qualitative 142 methods to elicit the group's social identity content. In line with the pragmatic paradigm, we 143 also opted to use two methods-observations and interviews-as multiple methods enhance the 144 ability to gain knowledge (Morgan, 2014). Observations are also relevant to social identity 145 content because they reveal which behaviors are used to enact the values of a social identity 146 group (Hogg & Reid, 2006 program. Additionally, there is evidence that there is wide variation between CrossFit® gyms 157 (e.g., management practices, injury rates) though they share the same brand name (Weisenthal et 158 al., 2014). Given these disparities between gyms, social identity content may also differ between 159 gyms; thus, we sought participants with membership at the same gym. Convenience sampling 160 was primarily used for both observations and interviews in order to be non-invasive and 161 emphasize anonymity. This decision reflected ethical consideration to avoid negatively affecting 162 the gym's business activity or the relationships between owners and members.

163
For observations, participants consisted of members who entered the gym during the 164 times when the first author conducted observations. Sex, role (e.g., trainer, member), physical 165 description, and behaviors were the only characteristics of observed members recorded. To 166 increase the number of members and types of behaviors observed, observations of 29 workouts 167 were made at multiple times of day (i.e., morning, n = 7; afternoon, n = 10; evening, n = 12). To 168 prevent observations from being biased by advance knowledge, participants were not notified in 169 advance about which workout periods would be observed. Also observed were one intra-gym 170 competition and one mandatory induction course for new members. Observations included 85 171 participants (44 male members, 32 female members, 6 male trainers, 1 female trainer, 2 gym 172 owners). For interviews, 10 members volunteered to be interviewed. Two members were 173 recruited when they initiated conversation with the first author, at which time the first author 174 invited them to participate as interviewees. Snowball sampling was also used in that 175 interviewees were asked to recommend other members for interviews. The first author 176 approached two recommended members, providing contact information in case they were willing Prior to data collection, five pilot interviews and three pilot observations were conducted.

188
A high quantity of data was rendered from each pilot interview and observation, highlighting the 189 need to narrow the scope of the study. Therefore, the research questions were limited to social 190 identity content and criticisms of CrossFit® rather than exploring more aspects. Further, we 191 learned that some members perceive researchers conducting research about injury in CrossFit® 192 settings to be critics, evoking a defensive posture. After piloting, the interview guide was 193 adjusted such that explicit questions about pain and injury were last. In this way, we were 194 careful to avoid asking leading questions about injury. Consequently, we found that 195 interviewees brought up the topics of pain and injury prior to being explicitly asked about these 196 topics. Pilot interviews also revealed that members were not familiar with overuse injury origins  This demonstrated the need for researchers to identify participants' descriptions of behaviors as 199 overuse risk behaviors when participants did not name them as such.  After receiving approval from an institutional ethics committee, informed consent was 218 sought from the gym owner. Given the public nature of the venue, the gym owner was identified 219 as the "gatekeeper" who was responsible for providing access and giving informed consent for 220 observations in these settings. Two weeks before observations started, flyers at the gyms and

238
Data were analyzed using NVivo software (v. 11). To start, the first author reviewed all 239 interview transcripts and field notes. Transcripts were sent to interviewees who were invited to 240 provide comments, clarifications, or changes in views. This was intended to check transcript 241 accuracy and generate additional data and insight, but interviewees did not provide new 242 information. Next, an inductive approach was used for a thematic analysis (Braun & Clarke,243 2006). The analysis consisted of descriptive coding used to identify simple, lower-order codes 244 across interviews followed by coding of observation data. Then, higher-order themes were 245 developed to represent relationships between lower-order codes across interviews and 246 observations. A focus of these steps was on internal homogeneity (i.e., each code/theme had 247 adequate evidence) and external homogeneity (i.e., no overlap between evidence supporting two 248 codes/themes). A final step, as employed by other sport/exercise psychology researchers (e.g.,   Throughout data collection and analysis, data saturation was considered to determine 264 whether additional interviews or observations were needed. We note that guidance regarding 265 data saturation and sample size typically pertains to analysis of one type of data such that little 266 guidance is given in assessing data saturation from multiple methods (i.e., observations and 267 interviews). Thus, we opted to assess data saturation after higher-order themes were identified.   Rigor can also be assessed specific to the methods used. Given our use of qualitative 287 methods, we considered markers of quality of qualitative research, including criteria (italicized 288 below) summarized by Tracy (2010). We believed the topic to be worthy given the harms of 289 injury. To achieve rich rigor, we considered theoretical constructs in relation to the topic; 290 captured extensive data from multiple sources; and presented original text samples such that 291 readers could determine plausibility of our interpretations. To contribute to transparency, we 292 provided details regarding our rationale for our choices (e.g., why we placed explicit questions 293 about injury last in the interview guide). Regarding self-reflexivity, we acknowledge the first 294 author was a member of this CrossFit® gym for a five-month period approximately two years 295 prior to conduct of this study. This membership resulted in a positive preconception of 296 CrossFit® as a program which enabled people to gain the physical and psychological benefits of 297 physical activity. Thus, it was of particular value to include research team members who had no 298 relationship with the gym. Though the first author's five-month membership at the gym was a 299 potential source of bias, prior knowledge of the gym's practices contributed to the study's 300 credibility. Credibility was also enhanced by ensuring findings included thick description (e.g., 301 concrete details) and dissenting views amongst participants. To enhance resonance, details of 302 participants' words and behaviors were presented such that readers with no exposure to 303 CrossFit® gyms or CrossFit® lexicon could understand within their own personal life 304 experiences, thus contributing to naturalistic generalizability. The study represents a significant 305 contribution, in that we give voice to a population who may be criticized by others, and we 306 advance the study of injury in exercise contexts. Ethical considerations included efforts to 307 ensure anonymity such that participants' characteristics were not detailed to a degree that would 308 enable them to be recognized by other members, trainers, or gym owners. Finally, we attempted 309 to achieve meaningful coherence by showing how our choices were supported by the pragmatic 310 paradigm and by focusing on psychological factors unified by theory.

312
In this section, the findings are divided into two parts reflecting the two research  Being Hard Core. "Hard core-ness" was a term used by MM44 to describe the type of 322 people who do CrossFit®, which tended to be people who "enjoy intense workouts" and were 323 "not afraid of discomfort". According to MM42, "People that voluntarily join CrossFit® are 324 people that want to sort of push themselves more or exert more effort." MM34 liked CrossFit® 325 because "it's something that pushes me really to the limit of what I can tolerate". He previously 326 experienced that feeling in cycling, but "still never anything quite as much as something that is 327 really a great CrossFit® session". One way that members enacted the value for being hard core 328 was by completing high-intensity, challenging workouts. A reason for completing high-intensity 329 workouts is explained by MM43 who said that members earn a "badge of honor". "Like, 'I'm 330 kind of a tough guy because I can do these CrossFit® workouts, and I push myself'". Advil® every few hours.

362
In one instance, the first author observed that being hard core in terms of attending "no matter 363 what" affected CrossFit® members' amount of rest in between workouts. On a morning in 364 which the gym was not open due to a scheduling glitch, members (e.g., MM20, FM12, MM39) 365 who usually attended the 5:30 a.m. sessions arrived, but, seeing the gym was closed, left. Later 366 that day, these members attended the 5:30 p.m. CrossFit® class. The next day, they attended the 367 CrossFit® workout at their normal 5:30 a.m. class time. Therefore, they attended two, high-368 intensity CrossFit® workouts in less than 12 hours rather than opting to miss a workout, yet 369 insufficient rest is an overuse risk behavior. 370 However, it was also observed that some members adjusted their attendance and intensity 371 at times. For examples, MM32 typically attended despite pain but did not attend "no matter 372 what". "I definitely come with aches and pains every day, don't get me wrong", but "one time 373 where I really felt like I hurt myself, I wasn't going to go in for a few days through that." When 374 FM31 struggled with an illness, she did not attend CrossFit® for a couple of weeks. During 375 FM31's absence, GO2 messaged her, "When are you going to be here? I miss you". FM31 376 perceived these actions by GO2 to be "really sweet". When FM31 returned to CrossFit® after 377 the absence, FM31 did the warm-up with the rest of the members, but then did a workout that 378 GO2 designed for FM31. The workout "was something to get me sweating a little bit, but it 379 wasn't too intense because I had been sick, and I didn't want to push myself too far." GO2 told results which, per the next section, was a basis for members continuing as paying gym members.

385
A third way in which members enacted being hard core was by withholding pain reports 386 from trainers. That is, they did not inform the trainers or others about pain. Instead, they 387 continued to exercise despite pain which is an overuse risk behavior. One reason for doing so 388 was an aversion to being perceived negatively, as shown by MM43: "especially when I first gave me more advice on how to strengthen those rotator cuff muscles.

407
After MT1 asked MM44 about the pain, MM44 became more comfortable reporting pain.

408
"Now, during the warm-ups, I will say 'MT6, hey, my shoulder is not feeling so hot today'".

409
Likewise, other members tended not to report pain until trainers directly solicited a pain report.

410
In one workout, a female member said, "My arms really hurt." After hearing her, MT1 asked,    Thus, in pursuit of desired results, some CrossFit® members engaged in overuse risk behaviours 479 (e.g., continuing exercise despite pain; doing more repetitions rather than resting).

480
Camaraderie. Members indicated that they valued camaraderie which embodied social 481 aspects of CrossFit® such as "social interaction", "community", "like family", "encouraging", 482 "welcoming", and "inclusive". One way in which this value was enacted was conversations.  Another interviewee demonstrated that the shared, intense experience led to "people liking each 500 other". He initially disliked a new, "annoying" member, but "I love the guy now Aside from conversing and personal greetings, another way camaraderie was enacted was 517 via encouragement of other members (e.g., applauding, cheering other members). According to 518 FM12, members could be positively evaluated by other members when they encouraged others.

519
She was "very impressed by the good people who encourage the people who are struggling".

520
For some members, such as MM30, the outcome of encouragement was to increase effort: "I can  Members also indicated that CrossFit® was superior to other exercise contexts because members 574 tended to modify workouts around pain and resolve injury rather than giving up and ceasing 575 exercise due to injury. As MM1 stated, "CrossFit® will find your weakness, so a lot of people,  Some CrossFit® members criticized critics who demonized CrossFit® without taking 596 into account the health and fitness benefits of exercise adherence. Before starting CrossFit®, 597 MM1 was overweight and had not adhered to any physical activity consistently. Though he 598 nursed a sore shoulder for 10 months during CrossFit® workouts, he weighed the sore shoulder 599 against the benefits of CrossFit® membership which enabled him to adhere consistently so that 600 he lost weight and perceived himself to be healthier. MM32 had tried many other exercise/sport 601 programs but only sustained regular adherence in CrossFit®. Though he tweaked his back in 602 CrossFit®, CrossFit® was still worthwhile to him. As MM34 said, "If this is what I need to do 603 to get in shape and be the best person that I can be, more power to me. I'll work out my way.

604
You work out your way".
Interviewees also negated critics' who had no direct experience with CrossFit®. FM31's 606 boyfriend was "very worried about me doing it…he's afraid I'm going to hurt my back." 607 However, "He's never tried it [CrossFit®]." Rather than stopping CrossFit® due to his 608 concerns, she opted to not discuss CrossFit® with him: "I don't really talk about it with 609 him…because if we do bring it up, I don't really want to have an argument about it." When FM2 610 learned that students in exercise science programs at a nearby university were being taught that 611 CrossFit® was "bad", she said, "you need to try it before you say anything else… you don't 612 know what you're talking about…it's like trying to talk about cake when you've never tried 613 cake." MM36 also discredited critics who did not participate in CrossFit®: "[they] make it 614 sound like we do one-rep maxes 20 times…They don't know about scaling." In the CrossFit® 615 lexicon, scaling involves reducing workout quantities to amounts suited to the individual's 616 factors (e.g., ability level, injury). MM44 described his interactions with two physical therapists 617 who initially indicated disapproval of CrossFit®. One told him, "You're going to hurt yourself.

618
You're going to mess your shoulder up. I'd never let my kids do it". After interacting with him 619 more, they then told MM44, "You seem like the kind of guy who's going to take care of 620 yourself...if it hurts, stop. If you feel yourself going too far, take a break, but as long as you do 621 exercises…and rehab your shoulder on your own, you'll be fine". Thus, MM44 believed that 622 critics' negative perception of CrossFit® activities changed when they were exposed to an actual 623 CrossFit® member. Finally, interviewees emphasized that CrossFit® gyms differ on many 624 facets (e.g., trainer attentiveness/experience, workout programming). Thus, they discounted 625 general criticism of CrossFit® that was not specific to the context at this gym.

627
In this study, we sought to identify values of group exercise participants relevant to 628 overuse risk behaviors as well as their responses to criticisms about injury. Through thematic 629 analysis, we identified three values (i.e., being hard core, achieving results, camaraderie) that 630 were relevant to overuse risk behaviors. We also identified two prominent types of responses 631 (i.e., compare dimensions of CrossFit® to other physical activities, denounce critics) of 632 CrossFit® members to criticisms about injury in CrossFit® activity. Here, we discuss these 633 findings in relation to constructs of the social identity approach: Social identity content and 634 social threats.

636
The three values identified in this studybeing hard core, achieving results, camaraderie 637 were interpreted to represent the social identity content of the group. That is, members 638 perceived these values to be reasons for being members of this group instead of other physical 639 activity groups; commonly endorsed by members; exemplified by highly-regarded members; and 640 a means for being perceived more positively by other members. This is not to say that these 641 values and associated behaviors are unique to this group; it might be that members of other social 642 identity groups endorse similar values (e.g., military members endorse camaraderie). Nor do we 643 imply that these members did not experience these values in other contexts (e.g., camaraderie felt 644 in previous sport participation). Rather, these CrossFit® members indicated that these values, 645 and the behaviors they used to enact the social identity content, were unique to their membership 646 in CrossFit®. This is demonstrated by CrossFit® members who expressed that they had not 647 engaged in some behaviors to the same degree in other physical activity contexts (e.g., a member 648 who had not engaged in the same intensity in biking; a member's attendance in previous 649 sport/exercise contexts was sporadic; a member did not know the names of people in other, non-650 CrossFit® fitness classes). Only when they became members of this social identity group-this 651 CrossFit gym-were these values central to shared social identity such that members engaged in 652 associated behaviors to a higher intensity, frequency, or level not experienced previously. When 653 behaviors stemming from the group's social identity content constituted overuse risk behaviors, 654 this group-based psychological factor was shown to be relevant to overuse injury. This finding is 655 unique given that individual (e.g., Type A personality) and inter-personal (e.g., pressure from

658
The value for being hard core was enacted, in part, by members who attended high-659 intensity workouts more than three times per week and, in some instances, with less than 24 660 hours between workouts, which puts members at risk for overuse, orthopedic injuries (ACSM,

682
In this study, we observed a desire for performance-and/or appearance-related results.  Some findings pertaining to CrossFit® members' camaraderie was not exclusive to this 694 study. Other researchers (e.g., Bailey et al., 2017) have also found that the shared experience of 695 high-intensity workouts is viewed as a source of CrossFit® members' camaraderie, and that 696 encouragement between members is a common behaviour in the CrossFit® context. However, a 697 novel finding was that a way in which camaraderie is enacted-through verbal encouragement-698 may induce higher effort. These findings in a naturalistic setting augment those of laboratory 699 settings in which researchers provided verbal encouragement to participants who then tended to 700 respond with increased effort (e.g., Moffatt et al., 1994). Together, these findings are suggestive 701 that verbal encouragement used to enact camaraderie may inadvertently be relevant to overuse 702 injury when members respond to verbal encouragement with excessive effort.

703
Throughout the findings related to social identity content, members were able to obtain 704 positive evaluations or avert negative evaluations of group members and/or leaders when 705 behaviors were aligned with social identity content. As illustrated by the member who initially 706 found another member annoying, completing high-intensity workouts enabled the 'annoying 707 member' to eventually be liked and accepted. Moreover, participants admired-or were admired 708 by-fellow Crossfit® members who enacted the social identity content via other behaviors such 709 as regular attendance, attendance despite adversity (e.g., recovering from illness), performing 710 well on a specific activity even if they were not typically one of the best performers, and 711 encouraging a struggling member. Thus, the behaviors used to enact social identity content gave concern is this: The behaviors that group members used to enact social identity content may 719 enable them to derive enjoyment and self-esteem from group membership, yet the same 720 behaviors put members at higher risk for overuse injury occurrence and severity. in that it strengthened members so that their injury susceptibility decreased.

738
According to the social identity approach, through positively distinguishing one's group 739 from other groups, being a member of a group increases positive evaluations of one's own worth 740 (i.e., self-esteem; Jetten et al., 2017). When an aspect of a group is negatively evaluated by others, the valued source of self-esteem is threatened (i.e., social threat). In response, group 742 members may engage in social creativity (Haslam & Reicher, 2006). Social creativity involves 743 maintaining a positive social identity through developing the group's social identity content such 744 that the group is seen as superior to other groups (i.e., achieves positive distinctiveness). For 745 example, a sport team on a losing streak cannot achieve positive distinctiveness on the dimension 746 of winning (outcome). Therefore, members may assert the teams' superiority on a dimension 747 other than outcome, such as sportsmanship or creativity. They may claim, for example, 'that 748 winning isn't everything; more important is how you play the game and playing fairly.' In this 749 sense, CrossFit® members' responses to injury criticisms resembled social creativity such that 750 injury occurrence wasn't everything; more important was how well members handled the 751 injuries, the effort they put into prevention, the health benefits, or the strength they gained.

752
The second pattern observed in CrossFit® members' responses to injury criticisms 753 involved denouncing features of those who criticize CrossFit®. This was done by dismissing 754 critics whose criticisms were products of bias from extreme examples of injury, incomplete 755 information, lack of personal experience with CrossFit®, or lack of specificity to individual 756 CrossFit® contexts. A possible interpretation of this pattern is another type of response to social 757 threats referred to as polarization (Brown & Ross, 1992). Polarization involves members' 758 defense of a social identity group by discounting the information critics provide. Of note, 759 instead of agreeing with critics, or adhering to advice and recommendations of critics, members 760 tend to react to criticisms by becoming more ensconced in their beliefs as well as a decreased 761 desire to leave the group and an increased antipathy towards other groups (Brown & Ross, 1982; 762 Hogg & Reid, 2006). Altogether, these findings demonstrated that criticisms about injury-even 763 when the critics were exercise and medical experts-did not induce members to perceive injury 764 as a problem, reflect on how to prevent injury, or change their injury-related behaviors because 765 these criticisms did not come from members of their own group.

766
Having identified some underlying values associated with a CrossFit group together with 767 associated (negative) behaviors, future research might examine how social identity content can 768 be modified by group leaders to change resultant negative behaviors (Haslam et al., 2011). group values from, for example, "We are hard core" to "We are smart about injury prevention".

772
Likewise, the basis for positive evaluations could be changed. For example, CrossFit® members 773 may be more apt to work out at a more moderate intensity, rest more, or decrease 774 effort/participation/report pain when they feel pain if they are praised for being injury-free for 20 775 months instead of only being praised for attending 20 months or for visible results. The findings 776 about social threats suggest that injury-prevention recommendations may be more effective when 777 implemented or communicated by CrossFit® leaders or members rather than experts who are not 778 members. For example, rather than experts critiquing the form of CrossFit® members, group 779 leaders may teach members to word verbal encouragement to emphasize technique (e.g., "Keep 780 good form!") instead of excessive effort (e.g., "Keep going!").

781
Despite the value of these practical implications, we acknowledge the study's limitations. 782 We limited the scope of psychological factors to identification of group values. Other factors 783 may have greater bearing on overuse injury occurrence in this context. Also, we opted to focus 784 on the utility of the social identity approach which led to us interpret data in relation to social 785 identity constructs (e.g., social identity content, social creativity, polarization). Other theoretical 786 approaches may reveal different, viable interpretations of participants' experiences and data. For 787 example, impression management theory could yield insight into findings pertaining to fear of 788 negative evaluation beyond negative evaluation by members of one's social identity group. 789 Further, our use of qualitative methodology and sampling method limited the generalizability in 790 that these findings are specific to one CrossFit® gym. 791 However, we considered the results in terms of other forms of generalizability applicable 792 to qualitative research methods (Smith, 2018), which could be viewed as a strength of this 793 project. Naturalistic generalizability involved presenting details of participants' words and 794 behaviors such that readers with no exposure to CrossFit® gyms, CrossFit® lexicon, social 795 identity, or injury could understand these results within their own personal life experiences (e.g., 796 being amazed upon learning one can complete a difficult task; a gym where patrons do not talk 797 to each other or know each other's names). Via inferential transferability, people not involved in 798 this specific CrossFit® setting may consider adopting a new practice due to what was learned in 799 this project (e.g., other exercise group leaders may guide exercisers to temper effort when 800 encouraged by others to try harder or solicit pain reports). Analytical generalization was achieved by 801 generalizing results to an established concept or theory (e.g., discussing results in relation to 802 social identity constructs of social creativity and polarization).

803
This study is one of the first to examine social identity constructs in relation to injury, 804 psychological factors of overuse injury in exercise contexts, and psychological factors 805 underlying injury in a CrossFit® context. It provided empirical support for the proposition that 806 the social identity approach is an applicable theoretical framework for examination of injury.

807
Overall, this study is critical in understanding why exercisers engage in injury-inducing 808 behaviors and how membership in social identity groups plays a role. Notes. Additional information about participants is not presented to preserve anonymity. Ability level refers to participants' ability to 927 meet assigned quantities in workouts (e.g., amount of weight or repetitions).