Disability and Health Journal 8 (2015) 216e222 www.disabilityandhealthjnl.com

Research Paper

Sport participation among individuals with acquired physical disabilities: Group differences on demographic, disability, and Health Action Process Approach constructs Marie-Josee Perrier, Ph.D.a,*, Celina H. Shirazipour, M.H.K.b, and Amy E. Latimer-Cheung, Ph.D.b a

Department of Kinesiology, McMaster University, 1280 Main Street West, Hamilton L8S 4L8, Canada b School of Kinesiology & Health Studies, Queen’s University, Kingston, Canada

Abstract Background: Despite numerous physical, social, and mental health benefits of engaging in moderate and vigorous intensity physical activities (e.g., sport), few individuals with acquired physical disabilities currently participate in adapted sport. Theory-based sport promotion interventions are one possible way to increase the amount of individuals who engage in sport. Objectives: The primary objective of this study was to examine the profiles of three different sport participation groups with respect to demographic, injury, and Health Action Process Approach (HAPA) constructs. Methods: ANOVAs and Chi-square tests were used to determine group differences on demographic and disability-related constructs. A MANCOVA was conducted to determine differences between three sport participation groups (non-intenders, intenders, and actors) with age, years post-injury, mode of mobility, and sex included as covariates. Results: A cohort of 201 individuals was recruited; 56 (27.9%) were non-intenders, 21 (10.4%) were intenders, and 124 (61.7%) were actors. The MANCOVA revealed significant differences between groups on the HAPA constructs, F(22,370) 5 9.02, p ! .0001, Pillai’s trace 5 .70, demonstrating that individuals with acquired physical disabilities will rate important health behavior constructs differently based on their sport intentions. Conclusion: These results provide an important framework that adapted sport organizations can use to tailor their sport promotion programs. Ó 2015 Elsevier Inc. All rights reserved. Keywords: Physical disability; Sport; Health behavior theory; Health Action Process Approach

An emerging body of literature suggests a number of unique physical and psychosocial health benefits of participating in adapted sport including better community integration, improved life satisfaction, the development of important friendships, and higher employment rates.1e4 Furthermore individuals with acquired physical disabilities, such as spinal cord injury (SCI), who engage in sport accrue more minutes of physical activity and work at higher intensities; therefore individuals who participate in sport, in comparison to other physical activities, may be more likely This abstract was presented at the Canadian Society for Psychomotor Learning and Sport Psychology’s annual meeting (SCAPPS) in November 2012. The lead author was supported by a Social Sciences and Humanities Research Council Doctoral Fellowship (Grant no: 767-2010-1882) during the data collection, analysis, and beginning of the writing phase. * Corresponding author. Tel.: þ1 416 890 5641. E-mail address: [email protected] or [email protected] (M.-J. Perrier). 1936-6574/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.dhjo.2014.09.009

to achieve the fitness benefits associated with a physically active lifestyle.5 Despite these benefits, an estimated 3% of individuals with acquired physical disabilities currently participate in sport.6 Despite this small figure, approximately 50% of individuals with physical disabilities have expressed an interest in exploring adapted sport options.7 Therefore, it is essential to explore how to promote sport participation within this population. Behavior change theories are a useful guide for understanding the necessary constructs to target in order to change behavior; to our knowledge, no behavioral theory has been applied to understand and promote sport among people with physical disabilities. However, several theories have been used to understand physical activity in this population. For example, both the Theory of Planned Behaviour8 and Social Cognitive Theory9 are commonly used to predict physical activity behaviors among people with acquired physical disabilities.10,11 While these theories are relatively successful at predicting intentions, they are

M.-J. Perrier et al. / Disability and Health Journal 8 (2015) 216e222

not as successful at accounting for variance in behavior. Given this ‘‘intention-behavior gap,’’12 there has been a call to use stage models that include a post-intentional phase. These models suggest that people can be classified into groups based on their readiness for behavior change. As such, more effective interventions will be tailored to individuals’ stages in stage-matched interventions.13,14 The Health Action Process Approach (HAPA) is one example of a stage model that has been used within disability and clinical populations.15,16 There are two distinct phases in the HAPA model, the motivational phase and the volitional phase.15 In the motivational phase, individuals set intentions to engage in a specific behavior.15 To develop intentions, individuals must have high risk perceptions, positive outcome expectancies, and task selfefficacy. Risk perceptions refer to individuals’ beliefs that they are at risk if they do not do the behavior, such as the risk of cardiovascular disease if they remain sedentary.15 Outcome expectancies refer to individuals’ beliefs about the possible outcomes of a behavior; positive outcome expectancies refer to beliefs such as the health benefits of engaging in physical activities while negative outcome expectancies refer to the negative outcomes of the behavior, such as pain or injury as a result of physical activity.15e17 Task self-efficacy refers to individuals’ confidence in their ability to perform a specific behavior, such as their confidence in playing wheelchair rugby.15,17 Once individuals have set intentions, they must translate these intentions into behavior during the volitional phase. In order to translate intentions into behavior, individuals need high maintenance and recovery self-efficacy and will also need to develop plans.15 Maintenance self-efficacy is the confidence individuals have in their ability to maintain a behavior, such as regularly participating in physical activity, even in the presence of barriers.15 Recovery selfefficacy is the confidence individuals have in their ability to return to a behavior after an absence, such as returning to an exercise program after a lapse.15 Strong plans detail what the individual will do, when, where, for how long, and with whom.16 With respect to the volitional phase, individuals can be further divided based on their behavior; intenders refer to individuals who intend to, but are not currently, engaging in a behavior while actors are those who are currently engaging in the behavior.16 As such, differences may emerge with respect to the HAPA constructs for non-intenders, intenders, and actors.16 Though the HAPA model has yet to be applied to staging individuals with respect to sport participation, it has recently been applied to explore group differences in physical activity among people with acquired physical disabilities.18,19 Martin Ginis and colleagues18 explored the differences between non-intenders, intenders, and actors with SCI for HAPA constructs with respect to physical activity. They found significantly higher scores among actors when compared to both non-intenders and intenders. Furthermore, intenders had greater scores on the HAPA

217

constructs when compared to non-intenders. Similarly, Chiu and colleagues19 explored how individuals with Multiple Sclerosis could be differentiated based on HAPA constructs. They found groups could be differentiated based on two mean centroids: motivation and volition. Precontemplators (i.e., non-intenders) scored low on both motivational and volitional centroids, the contemplation group (i.e., intenders) scored high on the motivational centroid but not the volitional centroid, and the final group (i.e., actors) scored high on both the motivational and volitional centroid. Thus among those with acquired physical disabilities, there could be distinct group differences with respect to the HAPA constructs. Given the paucity of literature on sport participation for people with acquired physical disabilities, more research is needed to understand how to tailor information based on HAPA stage. Therefore, the objective of this study was to examine the profiles of three different sport participation groups with respect to the HAPA constructs. We hypothesized that the three sport participation groups would significantly differ on all of the HAPA constructs. Specifically, non-intenders would score lowest and actors would score highest on all constructs. Based on past findings,18,19 we also hypothesized that intenders would be significantly different than both non-intenders and actors.

Methods This study is a secondary analysis in which the full methodology and participant inclusion criteria are previously reported.17 The methods are briefly summarized below. Participants A convenience sample of adults with acquired physical disabilities was recruited. Inclusion criteria included: a permanent physical disability acquired at the age of 16 or older; completed inpatient rehabilitation; and self-report to have no cognitive or memory impairments. There was no criterion regarding current participation in sport; however, to ensure variance in sport participation for the first analysis athletes were over recruited.17 Data collection and measures Eligible participants were invited to complete a questionnaire via telephone with the lead author or a trained research assistant or online through survey software. The same instructions were made available in both formats; Chi-square and ANOVAs revealed no statistically significant differences between participants who chose the online format and those who chose to complete the questionnaire by telephone. No personal information, such as name or address, was recorded on the questionnaire or included in the data file. Rather, participants were assigned a numeric

218

M.-J. Perrier et al. / Disability and Health Journal 8 (2015) 216e222

ID (e.g., P1, P57). Information necessary for sending compensation was recorded on a separate sheet and locked in a file drawer at the lead author’s institution. The questionnaire took approximately 30 min to complete. The General Research Ethics Board at the authors’ institution approved all procedures and materials prior to study commencement. Staging sport Sport was defined as: a structured physical activity between two or more people in a competitive event where a winner can be determined.17 Individuals selected the statement that best represented their current level of sport participation and were staged accordingly. Non-intenders were participants who were not engaging in sport, nor thinking about it. Intenders were participants who were considering engaging in sport within the next six months or were actively making plans for sport. Actors were currently involved in an adapted sport. This type of staging has demonstrated validity in physical activity behavior.20 Outcome expectancies Four affective, five instrumental, and eight negative outcome expectancy items were used to measure beliefs about the outcome of sport participation. These items were informed through previous research, thus ensuring construct validity21e23; the item stems were based on recommendations by Schwarzer and colleagues.13 Affective outcome expectancies included items such as ‘‘participating in sport would be enjoyable.’’ Instrumental outcome expectancies included items such as ‘‘participating in sport would be good for my health.’’ Negative outcome expectancies included items such as ‘‘participating in sport would lead to pain.’’ Items were rated on a scale from 1 5 not at all to 7 5 definitely. All subscales were internally consistent24 (aaffective 5 .94; ainstrumental 5 .75; anegative 5 .77). Risk perceptions Four items measured perceived risk for chronic disease, such as heart disease and Type II Diabetes.13 To ascertain a level of construct validity, these items were selected given that one of the most cited reasons for participating sport is to achieve physical health benefits.22,23 Items were rated on a scale from 1 5 no chance to 7 5 certain to happen (a 5 .75). Task self-efficacy Six items were used to assess confidence to engage in sports (e.g., how confident are you in your ability to play wheelchair basketball?). To specifically target task self-efficacy, as opposed to maintenance self-efficacy, participants were told to assume they had all the resources necessary, such as equipment and time, to do the behavior.25 Items were rated on a scale from 1 5 not at all confident to 10 5 completely confident (a 5 .88).

Intentions Four items measured commitment to sport participation (e.g., I will participate in sport over the next two weeks). Sport was specifically defined at the beginning of the questionnaire and a time period was attached to the items,13 which have demonstrated predictive validity in past research on physical activity within this population.26,27 Items were rated on a scale from 1 5 not at all to 7 5 definitely (a 5 .97). Maintenance self-efficacy Maintenance self-efficacy was composed of scheduling and barrier self-efficacy. Scheduling self-efficacy was measured using three items that assessed individuals’ confidence in their ability to schedule sport into their week (e.g., how confident are you in your ability to schedule sport at least one day per week?). These items have been previously validated for use with people with acquired physical disabilities, such as SCI.26 Barrier self-efficacy was measured using seven items that assessed individuals’ confidence to overcome common barriers to sport (e.g., how confident are you in your ability to participate in sport even when you have limited free time?). These items were based on barriers previously reported in the literature with individuals with SCI as well as amputations.28,29 Both scales’ items were rated on a scale from 1 5 not at all confident to 10 5 completely confident (ascheduling 5 .92; abarrier 5 .89). Planning Planning was composed of action and coping plans. Based on recommendations by Schwarzer and colleagues, four items assessed the presence of action plans, including items to assess when, where, what, and with whom.13 This format of questioning has been recommended for use with disability populations and demonstrated predictive validity in the SCI population.18,26 Similarly, two items measured coping plans (e.g., I have developed plans for dealing with barriers to my sport participation over the next two weeks). Items were rated on a scale from 1 5 not at all to 7 5 definitely (aaction 5 .97; rcoping 5 .68). Recovery self-efficacy One item measured confidence in one’s ability to return to sport after having stopped sport participation (e.g., how confident are you in your ability to go back to practice after your absence?). Responses were rated on a scale from 1 5 not at all confident to 10 5 completely confident. Data analysis A MANCOVA was used to determine differences between the three sport participation groups on HAPA constructs. Covariates included age, years post-injury, mode of mobility, and sex given that previous research has demonstrated relationships between these variables and sport.5,30

M.-J. Perrier et al. / Disability and Health Journal 8 (2015) 216e222 Table 1 Participant demographics Variable Age, mean (SD) Sex, n (%) Women Men Education, n (%) High school or less College/university Postgraduate Marital status, n (%) Single Married/common law Divorced/widowed Ethnicity, n (%) Caucasian Other Mode of mobility, n (%) Walk, assistive device Manual chair Power chair Years post-injury, mean (SD)

219

Non-intenders (n 5 56)

Intenders (n 5 21)

Actors (n 5 124)

Statistical test

51.6 (11.2)a

42.9 (11.4)b

41.2 (12.6)

F(2,198) 5 14.4, p ! .001 c2(2) 5 12.0, p 5 .002

b

26 (46.4%) 30 (53.6%)

15 (71.4%) 6 (28.6%)

41 (33.1%) 83 (66.9%)

13 (23.2%) 30 (53.4%) 13 (23.2%)

4 (19.0%) 9 (42.9%) 8 (38.1%)

20 (16.1%) 80 (64.5%) 24 (19.4%)

Fisher’s exact 5 5.7, ns

c2(3) 5 6.6, ns 14 (25.0%) 33 (58.9%) 9 (16.1%)

7 (33.3%) 9 (42.9%) 5 (23.8%)

43 (34.7%) 65 (52.4%) 10 (8.1%)

51 (91.1%) 5 (8.9%)

18 (85.7%) 3 (14.3%)

105 (84.7%) 19 (15.3%)

Fisher’s exact 5 1.4, ns Fisher’s exact 5 46.6, p ! .001 26 27 3 20.5

(46.4%) (48.2%) (5.4%) (13.5)a

8 8 5 11.3

(38.1%) (38.1%) (23.8%) (7.6)b

9 86 29 15.5

(7.3%) (69.4%) (23.4%) (10.8)b

F(2,198) 5 6.3, p 5 .002

SD 5 standard deviation. A different subscript denotes significant differences between groups on continuous variables.

Preliminary chi-square, Fisher’s Exact test, and ANOVAs demonstrated significant differences in these hypothesized covariates among the sport participation groups (Table 1), confirming that these constructs must be included in the final model. With unequal group sizes, the assumption of homoscedasticity will likely be violated; Box’s M was used to assess this assumption.31 The Box’s M test was significant ( p ! .0001), indicating a violation of this assumption, therefore Pillai’s trace was used to assess the results given that it is robust to violations of homoscedasticity.31 Significant effects in the MANCOVA were followed up with ANCOVAs (bonferroni adjusted a 5 .006). Tanhame’s 2 post-hoc tests were performed on significant ANCOVAs. This test was selected given that Levene’s test was significant for the majority of ANOVAs and therefore equal variances could not be assumed.31 Partial eta squared and Cohen’s d were used to measure effect size for specific analyses; a Cohen’s d of .20 represents a small effect, a d of .50 represents a medium effect, and a d of .80 represents a large effect.32 Power calculation A power calculation in was completed for ANCOVAs with the following details: 3 groups, 4 covariates, and a medium effect size (partial eta squared 5 .30). Assuming these parameters, we were adequately powered (.803).

Results A total of 216 participants agreed to participate in the study. Of these, 201 (93.1%) completed questionnaires (Table 1); 15 participants could not be contacted by the

phone or email they had given. There were no demographic differences between responders and non-responders. Of those who did respond, 56 participants (27.86%) were non-intenders, 21 participants (10.45%) were intenders, and 124 participants (61.69%) were actors. Intenders and actors were approximately 10 years younger than nonintenders. With respect to gender, there were a greater proportion of men in the actor group and women in the intender group. Furthermore, both intenders and actors were fewer years post-injury than non-intenders; though intenders were approximately 4 years younger than actors, this difference was only marginally significant ( p 5 .09). Finally, a greater proportion of actors were manual chair users than both the non-intenders and intenders. There were no differences between groups with respect to education, ethnicity, or marital status. The MANCOVA revealed significant differences between groups on the HAPA constructs, F(26,364) 5 7.82, p ! .001, Pillai’s trace 5 .72. Follow-up ANCOVAs revealed a main effect of group on eight constructs: affective outcome expectancies, task self-efficacy, intentions, action planning, scheduling self-efficacy, barrier self-efficacy, and recovery self-efficacy. Post-hoc analyses revealed three patterns (Table 2). First, actors had significantly higher scores than both intenders and non-intenders, with no differences between intenders and non-intenders, on task self-efficacy, intentions, and scheduling self-efficacy. Calculation of Cohen’s d revealed large differences (d O .80) between the nonintenders and actors, as well as the intenders and actors. However with the exception of task self-efficacy (d 5 .25), the differences between non-intenders and intenders were moderate-sized (d 5 .50) for both intentions and

220

M.-J. Perrier et al. / Disability and Health Journal 8 (2015) 216e222

Table 2 Group differences on Health Action Process Approach (HAPA) constructs Non-intenders Construct ANCOVA (n 5 56)

Intenders (n 5 21)

Actors (n 5 124)

Pattern 1, mean (standard deviation) Task self-efficacy F(2,194) 5 16.03, p ! .001, h2 5 .14

21.30 (15.06)a

24.90 (13.96)a

40.80 (13.82)b

Intentions

F(2,194) 5 78.09, p ! .001, h2 5 .45

9.98 (8.27)a

15.15 (9.93)a

25.49 (4.73)b

Scheduling self-efficacy

F(2,194) 5 23.66, p ! .001, h2 5 .20

14.48 (9.34)a

19.52 (10.18)a

25.91 (5.70)b

Pattern 2, mean (standard deviation) Affective outcome F(2,194) 5 21.01, p ! .001, h2 5 .18 expectancies

18.82 (7.13)a

25.14 (4.25)b

25.11 (3.16)b

F(2,194) 5 32.75, p ! .001, h2 5 .25

5.59 (3.24)a

8.86 (1.96)b

9.18 (1.40)b

24.89 (13.08)a

36.19 (11.51)b

46.58 (14.60)c

9.55 (7.62)a

16.75 (9.35)b

24.48 (4.33)c

24.79 (6.80)

27.43 (5.08)

27.46 (4.63)

Recovery self-efficacy

Pattern 3, mean (standard deviation) Barrier self-efficacy F(2,194) 5 24.95, p ! .001, h2 5 .21

Action plans

F(2,194) 5 77.58, p ! .001, h2 5 .44

Non-significantb, mean (standard deviation) Instrumental outcome F(2,194) 5 2.05, p 5 .13, h2 5 .021 expectancies Negative outcome expectancies

F(2,194) 5 5.20, p 5 .006, h2 5 .051

26.96 (9.04)

25.57 (9.60)

23.42 (7.08)

Risk perceptions

F(2,194) 5 1.60, p 5 .21, h2 5 .016

14.45 (6.14)

15.43 (4.79)

12.11 (5.00)

Coping plans

F(2,194) 5 3.65, p 5 .028, h2 5 .036

5.39 (3.99)

6.61 (4.62)

8.23 (3.96)

Effect size (Cohen’s d )a dNon-intenders, Intenders 5 .25 dNon-intenders, Actors 5 1.34 dIntenders, Actors 5 1.15 dNon-intenders, Intenders 5 .57 dNon-intenders, Actors 5 2.29 dIntenders, Actors 5 1.32 dNon-intenders, Intenders 5 .51 dNon-intenders, Actors 5 1.49 dIntenders, Actors 5 .76 dNon-intenders, Intenders 5 1.10 dNon-intenders, Actors 5 1.20 dIntenders, Actors 5 .0080 dNon-intenders, Intenders 5 .76 dNon-intenders, Actors 5 1.46 dIntenders, Actors 5 .17 dNon-intenders, Intenders 5 .95 dNon-intenders, Actors 5 1.58 dIntenders, Actors 5 .78 dNon-intenders, Intenders 5 .92 dNon-intenders, Actors 5 2.56 dIntenders, Actors 5 1.06 dNon-intenders, Intenders 5 .43 dNon-intenders, Actors 5 .46 dIntenders, Actors 5 .0061 dNon-intenders, Intenders 5 .15 dNon-intenders, Actors 5 .44 dIntenders, Actors 5 .25 dNon-intenders, Intenders 5 .18 dNon-intenders, Actors 5 .42 dIntenders, Actors 5 .68 dNon-intenders, Intenders 5 .28 dNon-intenders, Actors 5 .71 dIntenders, Actors 5 .38

A different subscript denotes significant differences between groups. a Cohen’s d: .2 is a small effect size, .5 is a medium effect, and .8 is a large effect size.32 b Bonferroni-adjusted a 5 .006.

scheduling self-efficacy. While the differences between non-intenders and intenders did not reach statistical significance, it may still be a notable one. In the second pattern, both actors and intenders had higher scores than non-intenders on affective outcome expectancies and recovery self-efficacy. Differences between the non-intenders and intenders, as well as non-intenders and actors, were large (d O .80). In comparison the difference between intenders and actors on these two constructs were small (d ! .20). Finally, there was a third pattern where each group was significantly different, such that intenders had lower scores than actors, but higher scores than non-intenders, on barrier self-efficacy and action planning. The differences between non-intenders and intenders, as well as non-intenders and actors were large (d O .80). Similarly, the differences between intenders and actors were also large (d O .80).

Although there were no statistically significant differences between groups on instrumental outcome expectancies, negative outcome expectancies, risk perceptions, and coping planning, effect sizes were calculated to explore the differences between group means. Most effect sizes were between small (d 5 .20) to moderate (d 5 .50) indicating few differences between groups. With respect to instrumental outcome expectancies, there were moderate sized differences in the means between non-intenders and intenders (d 5 .43), as well as non-intenders and actors (d 5 .46) while there was no difference in means for intenders and actors (d 5 .0061). This is similar to Pattern 2. There was a moderate sized difference in negative outcome expectancies between non-intenders and actors (d 5 .44), and small differences between non-intenders and intenders (d 5 .15) as well as intenders and actors (d 5 .25), similar to Pattern 3. Actors scored lowest on risk

M.-J. Perrier et al. / Disability and Health Journal 8 (2015) 216e222

perceptions, with differences between non-intenders and actors, as well as intenders and actors, being moderate to large (d 5 .42, d 5 .68), and differences between nonintenders and intenders being small (d 5 .18), similar to Pattern 1. Lastly, differences in coping planning were moderate to large for the non-intender and actor group (d 5 .71), and small to moderate for the non-intender and intender group (d 5 .28), as well as the intender and actor comparisons (d 5 .38), similar to Pattern 3.

Discussion As hypothesized, there were significant group differences between non-intenders, intenders and actors on demographic, disability, and HAPA constructs. These results provide an important framework that adapted sport organizations can use to tailor their sport promotion programs as well as for the creation of theory-based interventions. Nonintenders differed from both intenders and actors in several ways with respect to these theoretical determinants of sport. Non-intenders scored lower on the motivational constructs of the HAPA model than actors; this finding is in line with research among individuals with multiple sclerosis19 and individuals with SCI.18 Theoretical interventions that target intentions to get involved in sport among non-intenders will therefore need to target constructs such as affective outcome expectancies to first increase interest in adapted sport, as well as task self-efficacy. Interventions to build self-efficacy and support planning are necessary for those who are interested in sport, but are not yet participating. In comparison to non-intenders, intenders had similar levels of affective outcome expectancies as actors. Yet intenders had lower scores than actors on volitional constructs of the HAPA model. Given that intenders have the desire to participate in sport, interventions for intenders should focus on supporting these intentions by building task self-efficacy, maintenance selfefficacy, and action planning. To do so could help these individuals become actors given that past research has demonstrated the utility of planning on maintained physical activity.26 Significant group differences were not present for four constructs: risk perceptions, instrumental and negative outcome expectancies, and coping planning. Given that past research has demonstrated a relationship between instrumental and negative outcome expectancies for sport intentions,17 interventions that highlight the physical health benefits and mitigate the physical risks of sport participation may be useful for all groups. Past research has demonstrated that health risk perceptions have no significant relationship with intentions to engage in sport17; therefore this construct is not essential for sport promotion interventions. While action planning was significantly different between groups, coping planning was not. This was an

221

unexpected finding given that previous research has demonstrated the additional importance of coping planning in a physical activity and disability setting.26 Examination of the raw means demonstrates small differences between the three groups; inspection of the effect sizes reveals a moderate to large difference between the nonintenders and actors (d 5 .71), and a small to moderate sized difference between the intenders and actors (d 5 .48) indicating stronger coping plans among actors. Thus, while the differences did not reach statistical significance, there may be the same pattern for group differences on coping planning as there is for action planning. Given past research, as well as this trend, further exploration of the efficacy coping planning, as well as combined action and coping planning interventions, on sport participation is warranted. Limitations This study is not without its limitations. It is important to note that solely sport was measured, so any other activity done to support sport participation, such as strength training, is unaccounted for. For the first analysis (reference removed for blind review), athletes were over recruited to ensure variation in the initial study’s outcome; however, this may have an effect on group representativeness with respect to those who are not engaged in sport. To assess possible bias, one sample t-tests and Chi-square tests comparing our sample to demographics for the Canadian and American SCI populations5,33; we noted several demographic differences between our sample and these populations, though these were not always consistent. On average, our sample was approximately three years younger than both country’s demographics (44.0 years vs. 47.1 years and 48 years). Similarly, we also noted differences in the race, such that our sample contained fewer Caucasian participants than the Canadian sample (80.6% vs. 88.6%) and fewer African American participants than the American sample (1% vs. 15%). In comparison to the Canadian sample, we had fewer men (59.7% vs. 76.4%); however, this was not a significant difference compared to the proportion of men in the American sample (59.7% vs. 61%). We also had more participants with postgraduate education (19.9% vs. 6.5%). Given these differences, we may not be able to generalize these findings to specific subpopulations of people with disabilities such as those with a high school education or less and African Americans. Further research to explore predictors of sport participation in these subpopulations is warranted. Furthermore, a one-item measure for recovery self-efficacy may not be the most appropriate way to assess this construct in non-intenders and intenders. All data were measured, cross-sectionally, by self-report. Further research that uses objective measures in experimental studies to determine the effectiveness of targeting these constructs is warranted.

222

M.-J. Perrier et al. / Disability and Health Journal 8 (2015) 216e222

Conclusion The results of this study suggest that there are important differences with respect to demographic, disability, and HAPA constructs between three groups staged by sport participation. In comparison to those engaging in sport, non-intenders and intenders score lower on a number of modifiable, HAPA constructs. These constructs may be used to tailor theory-based sport promotion interventions among individuals with acquired, physical disabilities. Future research exploring sport promotion interventions that to target these constructs is warranted; furthermore, future research that explores the presence of these three sport intentions groups in key subpopulations, including people with less education and African Americans, is necessary.

References 1. McVeigh SA, Hitzig SL, Craven BC. Influence of sport participation on community integration and quality of life: a comparison between sport participants and non-sport participants with spinal cord injury. J Spinal Cord Med. 2009;32(2):115e124. 2. Chun S, Lee Y, Lundberg NR, McCormick B, Heo J. Contribution of community integration to quality of life for participants of community-based adaptive sport programs. Ther Recreation J. 2008;42(4):217. 3. Slater D, Meade MA. Participation in recreation and sports for persons with spinal cord injury: review and recommendations. NeuroRehabilitation. 2004;19:121e129. 4. Blauwet C, Sudhakar S, Doherty AL, Garshick E, Zafonte R, Morse LR. Participation in organized sports is positively associated with employment in adults with spinal cord injury. Am J Phys Med Rehabil. 2013;92(5):393e401. 5. Martin Ginis KA, Latimer AE, Arbour-Nicitopoulos KP, et al. Leisure time physical activity in a population-based sample of people with spinal cord injury Part I: demographic and injury-related correlates. Arch Phys Med Rehabil. 2010;91(5):722e728. 6. ParaSport Ontario. ParaSport Ontario Resouce Directory. Toronto; 2009. 7. Sport England. Adults With a Disability and Sport National Survey 2000e2001: Headline Findings. London; 2002. 8. Ajzen I. The theory of planned behavior. Organ Behav Hum Decis Process. 1991;50(2):179e211. 9. Bandura A. Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall; 1986. 10. Martin Ginis KA, Latimer AE, Arbour-Nicitopoulos KP, Bassett RL, Wolfe DL, Hanna SE. Determinants of physical activity among people with spinal cord injury: a test of social cognitive theory. Ann Behav Med. 2011;42(1):127e133. 11. Latimer AE, Martin Ginis KA. The theory of planned behavior in prediction of leisure time physical activity among individuals with spinal cord injury. Rehabil Psychol. 2005;50(4):389e396. 12. Webb TL, Sheeran P. Does changing behavioral intentions engender behavior change? A meta-analysis of the experimental evidence. Psychol Bull. 2006;132(2):249e268. 13. Schwarzer R, Lippke S, Luszczynska A. Mechanisms of health behavior change in persons with chronic illness or disability: the Health Action Process Approach (HAPA). Rehabil Psychol. 2011;56(3): 161e170.

14. Abraham C. Beyond stages of change: multi-determinant continuum models of action readiness and menu-based interventions. Appl Psychol. 2008;57(1):30e41. 15. Schwarzer R. Self-efficacy in the adoption and maintenance of health behaviors: theoretical approaches and a new model. In: Schwarzer R, ed. Self-efficacy: Thought Control of Action. Washington, D.C.: Hemisphere; 1992:217e242. 16. Schwarzer R. Modeling health behavior change: how to predict and modify the adoption and maintenance of health behaviors. Appl Psychol. 2008;57(1):1. 17. Perrier MJ, Sweet SN, Strachan SM, Latimer-Cheung AE. I act, therefore I am: Athletic identity and the health action process approach predict sport participation among individuals with acquired physical disabilities. Psychology of Sport and Exercise. 2012;13(6):713e720. 18. Martin Ginis KA, Tomasone JR, Latimer-Cheung AE, ArbourNicitopoulos KP, Bassett-Gunter RL, Wolfe DL. Developing physical activity interventions for adults with spinal cord injury. Part 1: a comparison of social cognitions across actors, intenders, and nonintenders. Rehabil Psychol. 2013;58(3):299e306. 19. Chiu C, Fitzgerald SD, Strand DM, Muller V, Brooks J, Chan F. Motivational and volitional variables associated with stages of change for exercise in multiple sclerosis: a multiple discriminant analysis. Rehabil Couns Bull. 2012;56(1):23e33. 20. Dishman RK, Vandenberg RJ, Motl RW, Nigg CR. Using constructs of the transtheoretical model to predict classes of change in regular physical activity: a multi-ethnic longitudinal cohort study. Ann Behav Med. 2010;40(2):150e163. 21. French DP, Sutton S, Hennings SJ, et al. The importance of affective beliefs and attitudes in the theory of planned behavior: predicting intention to increase physical activity. J Appl Soc Psychol. 2005;35(9):1824e1848. 22. Tasiemski T, Kennedy P, Gardner BP, Blaikley RA. Athletic identity and sports participation in people with spinal cord injury. Adapt Phys Activ Q. 2004;21:364e378. 23. Scelza WM, Kalpakjian CZ, Zemper ED, Tate DG. Perceived barriers to exercise in people with spinal cord injury. Am J Phys Med Rehabil. 2005;84(8):576e583. 24. Nunnally JC. Psychometric Theory. New York: McGraw Hill; 1978. 25. Foulon BL, Martin Ginis KA, Benedict C, Latimer AE. A little bit goes a long way: the effects of a wheelchair sport session on social cognitive theory constructs and leisure time physical activity among people with physical disabilities. J Sport Exerc Psychol. 2010;32:S163eS164. 26. Arbour-Nicitopoulos KP, Martin Ginis KA, Latimer AE. Planning, leisure-time physical activity, and coping self-efficacy in persons with spinal cord injury: a randomized controlled trial. Arch Phys Med Rehabil. 2009;90(12):2003e2011. 27. Latimer AE, Martin Ginis KA, Arbour KP. The efficacy of an implementation intention intervention for promoting physical activity among individuals with spinal cord injury: a randomized controlled trial. Rehabil Psychol. 2006;51(4):273e280. 28. Bragaru M, Dekker R, Geertzen JHB, Dijkstra PU. Amputees and sports: a systematic review. Sports Med. 2011;41(9):721e740. 29. Wu SK, Williams T. Factors influencing sport participation among athletes with spinal cord injury. Med Sci Sports Exerc. 2001;33(2):177e182. 30. Perrier MJ, Latimer-Cheung AE, Martin Ginis KA; The SHAPE-SCI Research Group. An investigation of seasonal variation in leisure-time physical activity in persons with spinal cord injury. Spinal Cord. 2012;50(7):507e511. 31. Tabachnick BG, Fidell LS. Using Multivariate Statistics. Boston, MA: Pearson/Allyn & Bacon; 2007. 32. Cohen J. Statistical Power Analysis for the Behavioral Sciences. Hillsdale, NJ: Lawrence Earlbaum Associates; 1988. 33. Christopher and Dana Reeve Foundation. One Degree of Separation: Paralysis and Spinal Cord Injury in the United States. Short Hills, NJ; 2010.

Sport participation among individuals with acquired physical disabilities: group differences on demographic, disability, and Health Action Process Approach constructs.

Despite numerous physical, social, and mental health benefits of engaging in moderate and vigorous intensity physical activities (e.g., sport), few in...
218KB Sizes 0 Downloads 8 Views