Staying physically active after spinal cord injury: a qualitative exploration of barriers and facilitators to exercise participation

Background While enhancing physical activity has been an essential goal of public health officials, people with physical impairments such as spinal cord injury (SCI) are more likely to live a sedentary lifestyle. Exercise has been shown to decrease the risk for many of the secondary conditions associated with SCI, including osteoporosis, cardiovascular disease, pressure ulcers, urinary tract infections, diabetes and arthritis, yet this population is rarely a target for health promotion efforts. This paper examines the self-reported exercise experiences of people with SCI using a qualitative-exploratory design. Methods We enrolled 26 individuals with SCI (15 self-described 'exercisers' and 11 'non-exercisers') from a non-random pool of survey responders. Semi-structured phone interviews were conducted to record participants' experiences with exercise pre/post injury, barriers and facilitators to being active and perceived health impact. Results Regardless of exercise status, all participants reported physical activity prior to injury and expressed interest in becoming active or maintaining an active lifestyle. Participants identified a range of both motivational and socio-environmental factors that were either facilitating or constraining of such a lifestyle. Non-exercisers identified barriers to exercise, including a perceived low return on physical investment, lack of accessible facilities, unaffordable equipment, no personal assistance and fear of injury. Exercisers identified facilitators, including personal motivation, independence, availability of accessible facilities and personal assistants, fear of health complications, and weight management. Exercisers associated a greater range of specific health benefits with being active than non-exercisers. Conclusion Despite motivation and interest in being exercise active, people with SCI face many obstacles. Removal of barriers coupled with promotion of facilitating factors, is vital for enhancing opportunities for physical activity and reducing the risk of costly secondary conditions in this population.


Background
More than a decade ago, the United States National Institute of Health Consensus Conference on Physical Activity (1995) concluded that 'All Americans should engage in regular physical activity' [1]. Enhanced physical activity is often identified as a key public health objective and leading health indicator and yet research has shown that, upon returning to the community after rehabilitation, people with activity limitations, such as spinal cord injury (SCI), are less likely to be physically active when compared to the able-bodied population [2,3]. Trend data tracking physical activity among people with disabilities between 1997 and 2006 put forward by the U.S. Centers for Disease Control and Prevention show that Healthy People 2010 activity targets are largely missed as activity rates continue to stagnate [4]. Ranking at the lower end of the physical activity spectrum, people with SCI are at a heightened risk for ill health and costly secondary conditions [5][6][7]. As a group, people with disabilities rarely feature as a target for health promotion efforts [8].
Acute mortality rates for people with spinal cord injury (SCI) have declined over the past three decades [9] while clinical attention has increasingly focused on the prevention of secondary conditions. It has been established that people with SCI are highly susceptible to medical complications and secondary chronic conditions, such as pressure ulcers, urinary tract infections, diabetes, cardiovascular disease, obesity, osteoporosis and arthritis [10][11][12][13][14][15][16], and that physical activity can help prevent such conditions [17], enhance functional abilities [18][19][20] and increase quality of life and social integration [20][21][22].
The path towards a physically active lifestyle however, is fraught with obstacles for people with disability. Several barriers to exercise have been identified in prior studies, including accessibility, pain, costs, psychological barriers, a lack of motivation and energy, and a lack of logistical information [23][24][25][26]. Other determinants of physical activity identified include completeness and level of injury [2,27], intention and perceived behavioral control [28,29], and the presence of health complications [25]. Factors working against these barriers to facilitate an exercise active lifestyle have also been identified, including preparation during time in rehabilitation and the role of peer mentors [25,30]. An increase in quality of life, energy, self-confidence and body image were reported by individuals with SCI participating in a structured exercise training program [31] while health and fitness have been identified as reasons for continued participation in exercise [30].
While existing research on exercise after SCI provides valuable insight towards identifying determinants, much of this research has been done in a controlled setting, through a structured exercise program, or in relation to rehabilitation discharge. Studies have used a range of standardized means to measure physical activity, including a structured reporting scale [24,32], a validated selfreport instrument [24,25,28], an activity monitor [2] or by providing exercise parameters to use in self-identification (e.g. 30 minutes per day, etc.) [27,29]. In contrast, this study asked participants to self-identify as 'exercisers' or 'non exercisers', thus taking into account the perceived identity of individuals as 'physically active' individuals rather than professionally defined categories. It may be assumed that self-definition of physical activity may broaden the range of barriers and facilitators that could be identified. Additionally, studies have primarily focused on perceived benefits and barriers, with less attention on perceived facilitators. It is unclear why some people continue to be physically active and pursue exercise opportunities after their injury while others do not. While exercise is essential to enhancing health, the onset of a physically disabling condition can provide serious challenges towards an active lifestyle. Knowledge of factors that may enhance exercise participation after SCI is crucial, particularly in order for health professionals to adequately tailor their support and interventions to the needs and lifestyle of their patients, thus improving upon the long-term rehabilitative process. For this reason, this paper reports on findings from a qualitative study aimed at identifying the factors that people with SCI living in the community perceive as affecting their level of physical activity. More specifically, the study sought to identify barriers to and facilitators of exercise as well as any perceived links between physical activity and secondary health conditions.

Method
This study is based on a qualitative-exploratory design following a quantitative exploration of exercise and physical activity patterns.

Sampling
Twenty-six adults with spinal cord injury were recruited (15 self-reported 'exercisers' and 11 'non-exercisers'). This sample size was deemed sufficient to explore the topical scope addressed in this paper, and is consistent with similar studies reported in the literature [33,34].
Participants were selected from a pool of 592 nation-wide survey participants dependent upon whether they selfidentified as an 'exerciser' or 'non-exerciser'. The survey pool itself was non-random and consisted of self-selected participants who met the following criteria: -1 year post SCI -18+ years of age -Speak English -Consent to being interviewed Researchers randomly contacted eligible participants by telephone or email to verify eligibility, willingness to participate, status as an 'exerciser' or 'non-exerciser' and a convenient time for the phone interview.

Semi-structured interview guide
A semi-structured interview guide was developed for indepth exploration based on core topical areas of the survey questionnaire. The guide was pilot-tested with 2 postinjury exercisers and 2 post-injury non-exercisers.
Questions focused on: • Experiences with exercise before injury We adopted a pragmatic participant-centered operationalization of 'exercise'. The determination of exercise status was based on self-reported information in the survey. 'Exercisers' were identified as those who reported being engaged in exercise activities either at home, in a gym or in both places. 'Non-exercisers' were identified as neither exercising at home or in a gym.
Consequently, in this paper, we will use the term 'exercise' to reflect both formalized exercise activities as well as physical activity initiated and carried out by individuals without the explicit intent to exercise (e.g. yard work, wheelchair pushing, etc.). Individuals, however, must have identified the activity as 'exercise'. This 'naturalistic' understanding of exercise reflects the recommendation by the Centers for Disease Control and Prevention and the American College of Sports Medicine that both lowerlevel intensity activity, as well as vigorous exercise, is beneficial in reducing the risk of heart disease and enhancing general fitness levels [35].
Similarly, our definition of 'secondary conditions' was participant-centred. The scope of conditions was defined by study participants.

Procedure
All participants consented to the interviews before completing the first wave of the national survey. The study protocol, survey document, interview guide, informed consent and HIPAA documentation forms were approved by the MedStar Research Institute IRB in Hyattsville, MD. Essential demographic information was obtained at the time of the survey.
Each interview was conducted by telephone at a time convenient to participants. Interviews, conducted by an experienced interviewer, lasted between 20 and 30 minutes and were recorded with participant consent. Audio recordings were transcribed, and subsequently verified by two moderators.

Data analysis Quantitative
We conducted bi-variate, nonparametric analyses for differences in the demographic and clinical profiles between exercisers and non-exercisers. We computed Mann-Whitney U tests for continuous data (e.g. age, duration of injury), and χ 2 tests for independent samples for categorical data. The Fisher's Exact Test was used when cell sizes were smaller than 5.

Qualitative
We used an ethnographic approach and used a descriptive framework provided by the interview questions [36] to guide initial coding and content analysis. All transcripts were read by two analysts (authors of this manuscript) independently and initial thematic categories were recorded. Constant comparative coding [37] using TAMS Analyser for Mac OS X generated a total set of 55 codes, grouped as exercise type, barriers, facilitators, perceived benefits and secondary conditions. Thematic codes were discussed, negotiated and continuously refined between the two authors. Table 1 contains background characteristics of the 26 study participants.

Participant characteristics
There were no statistically significant differences between exercisers and non-exercisers with regard to age, gender, race, education level, employment status, marital status, injury level or completeness of injury. The only variable both groups differed on was 'duration of injury', with non-exercisers having a significantly longer median duration of injury. Despite the failed statistical significance between the groups in terms of reported household income, seven exercise-active participants indicated a household income above $60,000, while only two did so in the non-exercise group. The uneven group size may have masked group differences.

Pre-injury exercise experiences
Regardless of post-injury exercise status, most participants reported being exercise active pre-injury. Thus, not all preinjury exercisers continued being exercise active after injury and several pre-injury non-active participants became active post-injury. Our finding suggests that preinjury activity levels may prove to be a poor predictor of post-injury activity, a finding indicated elsewhere [38].
The conducted interviews revealed a wide range of factors that were identified as either facilitating or constraining of an individuals' level of activity. Each factor identified as a deterrent by one group, was often inversely identified as a facilitator by the other group, and vice versa. Understand-ing these psychosocial conditions and their elasticity can be essential to the success of promoting more active lifestyles post-injury. Often it is not one factor that facilitates or constrains, but a cluster of overlapping and intertwined factors, making unique each individuals experience while thus challenging our ability to identify consistent themes.
An important distinction among the facilitating and constraining factors reported by our participants, were whether their origin was socio-environmental or motivational. Participants identified motivation, or lack of motivation, as a major factor in the determination of their physical activity level. We present the varied roots of participants' motivation, or potential motivation. Secondly, participants identified socio-environmental barriers to or facilitators of an exercise active lifestyle. We present these exogenous factors and how they both helped or hindered participants' level of physical activity.

Motivational triggers/constraints
Both exercisers and non-exercisers identified 'motivation' as the most critical factor to being and staying exercise active. Just as exercisers cited motivation as a strong facilitator of their being active, non-exercisers regarded their lack of motivation as a major constraint to an exerciseactive lifestyle. Exercisers identified a variety of 'triggers' for their motivation while some non-exercisers suggested potential sources of motivation.

Perceived return on investment
One of the most commonly reported barriers to exercising among non-exercisers was the perceived 'limited return on investment' from aerobic exercise. The amount of time and energy needed to reach perceived beneficial levels of activity were identified as too demanding or unrealistic and failed to motivate some individuals. Although some non-exercisers did make statements that can be characterized as 'positive' and 'optimistic', collectively their statements were far less frequent, less specific and would be better described as 'hopeful'.
Another motivational factor was to stay as independent as possible and to reduce reliance on personal assistance. Several participants identified their excellent physical constitution before their injury as a major factor for good recovery. The value of maintaining health, strength, and fitness was perceived as an incentive for considering exercise post-injury.
'... being in great shape before injury is a lifesaver. ...They were saying due to the physical shape I was in... that was one of the main factors' ~Non-exerciser, male, 41, C3 complete.

Socio-environmental resources/barriers
All participants, regardless of exercise status and motivation level recognized the impact and influence of social and environmental factors on their level of physical activity. Non-exercisers cited various barriers that prevented even the most motivated among them from becoming exercise active. While exercisers were often able to identify socio-environmental facilitators, many did so with a caveat, suggesting needs for improvement or acknowledging some level of constraint.

Information access
A lack of knowledge and resources were a commonly cited obstacle towards being active. Some non-exercisers expressed interest in being more physically active but were unsure of where to look for tips or assistance. For those living in a rural community with no local rehabilitation facility, this may be particularly true.

Discussion
Our findings suggest that participants' physical activity levels are contingent on a combination of motivational and socio-environmental factors, varying from case to case, and making generic exercise prescription problematic. Reported barriers and facilitators to exercise may be differentiated by these motivational or socio-environmental origins. All participants identified multiple factors as having influence over their level of physical activity. In sorting the common themes of our interviews, we recognize and emphasize the interwoven relationship between motivational triggers/constraints and socio-environmental resources/barriers. An attempt to simplify or minimize the potential inter-dependence of these factors would be to distort the experiences reported.
In fact, what we may best discern from these shared experiences is that no one factor acts as a function of an individuals' level of physical activity; it is instead a cluster of varying factors, with shifting degrees of severity and influence from case to case. This paper seeks only to heighten and highlight individual factors within this more complex context.
The Theory of Planned Behavior [39] posits that physically active behavior is a function of intention, which is influenced by attitudes, subjective norms and perceived behavioral control. Other studies have successfully applied this theory to the SCI population [27,28] and our findings further validate its usefulness. Exercisers' attitudes were linked to the knowledge of the benefits of being physically active while non-exercisers considered the level of physi-cal investment required to reach the perceived degree of benefit as unrealistic, unattainable or simply too demanding. Exercisers and non-exercisers reported similar levels of pre-injury physical activity, suggesting a shared norm of exercise as an essential part of life, yet the motivation needed to comply with this standard was a point of divergence between the two groups post-injury. Finally, nonexercisers reported frustration with physical limitations and unhappiness with existing exercise options, while exercisers identified physical therapy and training received in rehabilitation as major facilitators. A positive experience with exercise options during rehabilitation may shape the perception of behavioral control in terms of one's ability to be exercise active post-injury and lay the foundation for exercise engagement. Participants in this study, and elsewhere [25,30] identified their time in rehabilitation and physical therapy as critical for their current level of exercise commitment, while several non-exercisers noted a lack of support or recommendation to exercise from their physician, a problem recorded elsewhere [26,32]. It is also worth noting that non-exercisers had a significantly longer median duration of injury, compared to exercisers. The possibility that duration of injury has a potentially negative effect on the intentions to exercise should be explored further.
Study participants saw multiple benefits of regular exercise in terms of maintaining physical health and preventing secondary conditions. Apart from general cardiovascular fitness, several participants were concerned about pressure ulcers, urinary tract infections and respiratory problems. Some perceived the threat of these conditions as a motivator for maintaining their exercise regimen. The risk of weight gain was also a contributing factor for several participants. The psychological benefits that many respondents derived from their exercise engagement are consistent with findings from the literature [40][41][42][43]. While these health related factors were reported as facilitators in this study, they have been reported as benefits of exercise elsewhere [23,31]. The difference between a facilitator and a benefit is important to note. While perceived health benefits may act as a facilitator of continued exercise, it would seem that only the anticipation of such benefits would facilitate initial engagement.
In theoretical terms it may be argued that people who participate in regular exercise show greater 'self-determination'. Self-determination theory [44,45] posits that individuals need to develop a sense of autonomy and competence, which is essential for a process of internalization and integration of health behaviors (i.e. exercise). Autonomy implies that individuals value and prioritize behaviors and make them integral to their life-style. Additionally, people may feel more inclined to be physically active if they perceive themselves as confident and compe-tent. People who experience a greater degree of autonomy may also be more likely to learn new behaviors and feel competent. In our study several exercisers acknowledged that their exercise routines and strategies had changed post injury but that the reasons for being physically active had not. It may be argued that they managed to integrate physical activity effectively in their post-injury life and felt competent in exploring alternative means of achieving the expected exercise benefits. Other participants who had been physically active before their injury however, did not perceive these benefits (e.g. aerobic gains) and did not make exercise an integral part of their lives post-injury.
There is some indication that individuals with long-standing exercise routines may see greater personal benefits than external incentives (e.g. weight loss; appearance) to pursue regular physical activities [46]. Another element of the self-determination theory is 'relatedness', referring to the quality of the patient-practitioner relationship and how it may shape individuals' motivation to engage in behavior change. As noted, several respondents in this study felt that health professionals did not provide sufficient support and recommendations for exercise postinjury. Self-determination theory however, has not been studied in people with SCI. Similarly, a better understanding of psychological constructs, such as 'optimism' [47], with regard to influencing outcome expectancies related to exercise may further complement the picture.
The identification of socio-environmental barriers, especially those focused on accessibility and disability-specific knowledge of health providers, are not unique to the problem of exercise, and have been reported in conjunction with access to health services [26,32,[48][49][50][51]. A greater proportion of participants expressed interest in being physically active than actually were, a finding reflected elsewhere [24,26]. The removal of socio-environmental barriers could prove to be one of the most effective facilitators, thus allowing a motivated individual access to choosing an exercise active lifestyle.
It is important to emphasize that the socio-environmental barriers identified by non-exercisers were often, in their inverse form, factors that facilitated a physically active lifestyle for exercisers. Motivation aside, exercisers were able to be active because of the availability of an accessible community-based facility, or because of their capability to maintain home equipment. Non-exercisers often regarded the lack of such facilities and/or equipment as a barrier, a finding reflected in prior studies [26,52]. The role of personal assistance is equally noteworthy as many exercisers, particularly those relying on home equipment were able to do so only with the assistance of another. The absence of such help was identified as a major constraint by participants in both groups.
Even those for whom cost was not a barrier acknowledged its pervasive role in determining access. Facilities, personal assistance and home equipment are often available with a financial cost; one that several exercisers admitted at being able to meet. For those relying on insurance coverage, particularly Medicaid and Medicare, limitations of coverage emerged as an often impenetrable barrier, potentially denying individuals access to continuous and longterm physical activity.
More comprehensive and multi-level efforts are needed to address the physical health promotion needs of individuals with spinal cord injuries and other physical disabilities as Healthy People 2010 goals are far from being met.

Limitations
The study has several limitations. Its qualitative nature does not necessarily allow findings to be generalized to a larger population. The selection of subgroup participants (exercise vs. non-exercise) was based on a pool of people with SCI who had been previously included in a non-random sample. One of the principal limitations of both the larger survey study, from which the sub-sample was drawn, and the resulting subsets was the limited representation of ethnic minority groups. Even though there were no statistically significant differences between exercisers and non-exercisers in our selected subgroup, more individuals in the exercise group had higher disposable incomes and reported having more personal exercise equipment and assistance. This may be a finding in itself. Economic inequities among people with SCI may drive differential access to exercise information and support.

Conclusion
Most people with SCI are principally motivated to engage in exercise so to maintain health and prevent secondary conditions. Labeling individuals simply as 'non-compliant' without a full appreciation of their motivational constraints and socio-environmental barriers to exercise is not helpful. Successful behavior modification requires the consideration of both these types of factors. While improvements in providing better access to affordable facilities and personal assistance, especially in rural communities, is essential in developing accessible, inclusive, and equitable exercise support and health promotion programs, equal consideration must be given to the unique socio-environmental realities faced by individuals if such programs are to reach their effective potential. These broadened considerations allow clinicians and public health professionals to arrive at an understanding of physical activity that is not solely focused on restoring, improving or maintaining function, but enhancing general health and wellbeing. Closing education gaps and addressing professional 'blind-spots' among health care professionals may additionally help abate some of the obstacles that currently deny motivated individuals with SCI participation in healthy living activities.