Self-Perceived Psychophysical Well-Being of Young Competitive Swimmers With Physical or Intellectual Impairment

Regular practice of sport activities yields psychophysical benefits for both the general population and persons affected by physical or intellectual impairments. Practicing competitive sport may add further value to these benefits. The objective of this observational cross-sectional study was to investigate the role of competitive sport practice in enhancing self-perceived psychophysical well-being of some select participants, using the Psychological General Well-Being Index and the Short Form-12 indices. We recruited at national events 100 young Italian competitive swimmers affected by physical or intellectual impairment. These respondents' results were compared with those of a control group of 100 Italian participants who did not practice competitive sport but who were also affected by physical or intellectual impairment, randomly selected from rehabilitation clinics and communities of young people with disabilities. Scores of psychological and emotional well-being were higher by 40% or more for the practitioners of competitive sport (p < .0001; Cohen's effect size d ≥ 1.3). While our study's results suggest possible positive psychophysical benefits to competitive sport practice for young people affected by physical or intellectual impairment, longitudinal research is needed to be certain that our results are not due to self-selection into sports participation of those persons with disabilities who have a uniquely higher sense of well-being.


Introduction
The beneficial effects of regularly practiced amateur sport activities on general health, psychophysical well-being, and broad disease prevention are intensely promoted and advertised in modern society. At present, these benefits are so widely established and shared in public opinion that they have passed from scientific understanding (Eime, Young, Harvey, Charity, & Payne, 2013;Rasciute & Downward, 2010;Scully, Kremer, Meade, Graham, & Dudgeon, 1998) to general acceptance and undisputed lifestyle guidelines (Janssen, 2007;U.S. Department of Health and Human Services, 2008). Likewise, or perhaps to an even greater degree, people affected by physical or intellectual impairment can obtain psychological, social, and health benefits from participation in sport activities (Chawla, 1994;Giacobbi, Stancil, Hardin, & Bryant, 2008;Martin Ginis, Jetha, Mack, & Hetz, 2010;Rogers, Furler, Brinks, & Darrah, 2008;Verschuren et al., 2007).
Although physical well-being is strictly related to the type and severity of the individual's physical impairment, psychological well-being may not be so restricted, as the sport context may offer benefits that outweigh the psychological implications of physical impairment. Psychological well-being is defined as the dynamic balance between one's personal needs and potentialities on one hand and the characteristics of the external environment on the other hand (Dodge, Daly, Huyton, & Sanders, 2012;Felce & Perry, 1995;Garcia, Nima, & Kjell, 2014;Ryff & Keyes, 1995). If sport activities are practiced at a competitive level, the related emotional, motivational, and social characteristics of that environment may provide a precious contribution to the individual psychological well-being of people affected by physical or intellectual impairment, helping them shape their own lives through a steady realization of personal objectives, favoring an acceptance of their pathology and healing the natural imbalance between their needs/potentialities and environmental demands (Ryff & Keyes, 1995). According to prior literature, the strength of one's well-being, including personal growth, optimism, and strong social support networks, increases as athletes with impairments increase their level of competition (Macdougall, O'Halloran, Sherry, & Shields, 2016). This is particularly crucial in youthful life stages-adolescence and young adulthood-when major mental and physical changes occur and accentuate the effects of positive and negative external influences (Dykens, Rosner, & Butterbaugh, 1998;Pollard & Lee, 2003;Sherrill, Hinson, Gench, Kennedy, & Low, 1990;Wilson & Clayton, 2010).
Research studies have investigated the effect of practicing competitive sport on self-concept (Dykens & Cohen, 1996;Weiss, Diamond, Demark, & Lovald, 2003) and social competence (Dykens & Cohen, 1996) among individuals affected by intellectual impairment. Interviews conducted with five British paralympic swimmers aged 20-24 years with congenital impairments revealed that competitive swimming facilitates self-and social-acceptance, identity development, and a sense of normalization (Brewer, Boin, Petitpas, Van Raalte, & Mahar, 1994). A study of 26 basketball players who used wheelchairs showed that competitive physical activity seemed to support these players' self-efficacy beliefs, feelings of empowerment, and motivation for continued involvement (Giacobbi et al., 2008). A number of studies have addressed sport orientation and athletic identity in athletes with impairments (Kokaridas, Perkos, Harbalis, & Koltsidas, 2009;Martin, Adams-Mushett, & Smith, 1995;Martin, Eklund, & Mushet, 1997;Pack, Kelly, & Arvinen-Barrow, 2017). Athletic identity is the extent to which individuals identify with the role of an athlete, and, based on this prior research, it has been associated with the individual's self-perception, behavior, and commitment to ideological and occupational options. Martin et al. (1995Martin et al. ( , 1997 pointed out that athletic identity involves not only positive aspects such as motivation, goal orientation, and sense of empowerment but also negative aspects such as exclusivity (i.e., inability to identify with other roles) and negative affectivity (i.e., negative emotional responses to injury, retirement, or other sources). An investigation of psychophysical well-being in participants affected by either physical or intellectual impairment through the practice of competitive-level sport activities should be extended into different contexts and target populations (e.g., country, type of practiced sport, age range, type of impairment, method of investigation, etc.). The interest and potential impact of this issue go well beyond the medical field and also concern ethical, social, and economic aspects of our society globally. Indeed, on one hand, the well-being of individuals affected by impairments is a measure of a country's civility and inclusivity, and, on the other hand, more pragmatically but just as important, the well-being of this population may yield a significant societal reduction in health-care costs, to which rehabilitation programs contribute remarkably.
In studies addressing current theories on well-being (Felce & Perry, 1995;Garcia et al., 2014;Pollard & Lee, 2003;Ryff & Keyes, 1995), psychological well-being has been multidetermined. Among the factors contributing to wellbeing are self-acceptance, strong and lasting affection bonds, independence, aspiration to realize objectives, and balance between self and environment. The existing literature (Brewer et al., 1994;Dykens & Cohen, 1996;Giacobbi et al., 2008;Weiss et al., 2003) is limited, but reports of results on specific dimensions of well-being in impaired competitive athletes reveal an overall trend toward beneficial effects of engaging in competitive sport. Our objective in the current study was to assess the specific hypothesis that the practice of competitive swimming would play a positive role in the self-perceived psychophysical and emotional well-being of young participants affected by physical or intellectual impairment. In this article, we present the results of a cross-sectional survey of the self-perceived psychophysical, physical, emotional well-being of 100 young age paralympic athletes 1 who took part in two events of the Italian calendar of paralympic swimming competitions, namely the promotional Paralympic Swimming Italian Youth Paralympic Championship, April 29-30, 2017, Fabriano (An), Italy, and the Italian Paralympic Swimming Championship of sport clubs, May 20-21, 2017, Lignano Sabbiadoro (Ud), Italy. We compared these survey results with those of a control group comprised of 100 young participants who were also affected by physical or intellectual impairment but who did not practice competitive-level sport. In this effort, we used two widespread and well-documented survey indices, namely the Psychological General Well-Being Index (Dupuy, 1984) and the 12-Item Short-Form Health Survey (SF12; Ware, Kosinski, & Keller, 1996), to provide a quantitative probe of self-perceived psychophysical well-being in terms of different well-being domains. Our main independent variable was the practice of competitive sport, but we also examined gender, age, the acquired or congenital nature of the participants' impairment, the type of participant disorder, and the mode of survey administration.

Participants and Survey Administration Procedures
We adopted the following participant recruitment criteria: (a) diagnosis of physical or intellectual disability, (b) 11-23 years of age, (c) self-awareness ability for responding to questionnaires with or without assistance, and (d) holding an Athlete's license issued by either the International Paralympic Committee (IPC) or the Italian Federation for Paralympic Swimming (FINP), valid the for the relevant season. We chose criterion (b) as representative of the youth age range, according to the European Para Youth Games participation criteria and in compliance with the World Health Organization age classification groups. We evaluated criterion (c) with help from relatives and coaches, who assisted the participants in completing the questionnaires. We applied criterion (d) to the population sample under study, and an opposite criterion to the control sample, namely no practice of sport at a competitive level.
Sport practitioner participants were the young paralympic athletes who took part in the promotional Italian Youth Paralympic Swimming Championship (Campionati Italiani Promozionali Giovanili), April 29-30, 2017, Fabriano (AN), Italy, and in the Italian Paralympic Swimming Championship of sport clubs (Campionato Italiano di Societa`), May 20-21, 2017, Lignano Sabbiadoro (UD), Italy. At both events, in briefings prior to the event onset, we described this study and gave detailed instructions for the administration of the questionnaires, according to related recommendations . The briefings were attended by FINP executives, coaches, athletes, and relatives of the athletes. During the competitions, we distributed the questionnaires in paper form to all the participants. In addition, we edited the same questionnaires and related informative notes through Google Drive software and made them available online, through a link posted on the official FINP website (http://www. finp.it/), emphasizing the importance of the project to foster participation. We collected 195 questionnaires from participants in the competitions; out of these, 92 were rejected because they did not fulfill the 11-23 age range criterion, and seven were rejected because participants affected by severe forms of autism were not able to answer properly, despite assistance from their coaches. Finally, we recruited four other participants who were practicing competitive triathlon and affected by multiple sclerosis (MS) to get specific comparative information on participants who were affected by severe progressive physical conditions. This process yielded 100 completed questionnaires, 58 in paper form and 42 online for data analysis.
We also recruited an equal number (N ¼ 100) of young participants affected by the same impairments who did not practice competitive sport to use as a control group. These participants were distributed within the same age range and levels of education as the sample of paralympic athletes. Forty-seven control participants were randomly selected from local rehabilitation clinics, while the remaining 53 were randomly recruited among attendees of a promotional workshop entitled ''Sport and disability in children and adolescents'' that took place June 10, 2017, at the Sport Festival, Genova, Italy.

Instruments
For this research, we used two scientific survey indices-the Psychological General Well-Being Index (PGWBI; Dupuy, 1984) and the SF12 index (Ware et al., 1996;Ware, Kosinski, & Keller, 1998)-whose validity and reliability were demonstrated in literature (Gandek et al., 1998;Lundgren-Nilsson, Jonsdottir, Ahlborg, & Tennant, 2013;Ware et al., 1996). Each of these tools has a generic (not syndrome-specific) character and focuses on the respondents' psychological and emotional well-being rather than on their specific physical well-being. This is a key requirement for instruments that probe a significantly heterogeneous sample of participants who have different types of health conditions or impairments. Both indices are divided into multiple subtests or domains that address different aspects of psychological well-being, allowing a more delineated data analysis for describing a person's psychophysical profile. The two questionnaires required no more than 30 minutes to complete and could be self-administered or administered in a face-to-face interview (best for participants with either visual or intellectual impairments). Widespread use of these and other similar indices has been well documented in prior literature (Becker et al., 1997;Boman, Bryman, & Mo¨ller, 2004;Wiklund & Karlberg, 1991).
The PGWBI questionnaire (Lundgren- Nilsson et al., 2013) consists of 22 items, each with six possible answers, whose score is proportional to the level of psychological and general well-being, rated by respondents on a 6-point scale from 0 to 5. The 22 items are sorted into six different domains: anxiety, depressed mood, positive well-being, self-control, general health, and vitality. Each domain includes 3-5 items.
The SF12 index (Gandek et al., 1998) has a twofold character, in that it addresses both physical and mental health measured by two separate synthetic indices, the physical component summary (PCS-12) and mental component summary . This measure probes physical and mental functions and wellbeing, the degree of impairment on both a social and individual basis, and personally perceived general health. Through this instrument, we were primarily interested in the respondent's mental health, whereas we treated physical health data as ancillary information in our data analysis. This questionnaire is made up of 12 items and assesses eight domains: physical functioning, role limitations due to physical health, role limitations due to mental health, emotional well-being, bodily pain, vitality, social functioning, and general health. The combined use of PGWBI and SF12 indices has been previously recommended (Grossi, Mosconi, Groth, Niero, & Apolone, 2005).
Together with the earlier two questionnaires, we administered a further questionnaire asking the participants to provide additional personal data and details on the administration conditions of the PGWBI and SF12 questionnaires. Specifically, we sought to know the participants' age, gender, level of education, kind of impairment, method of administration of the questionnaires (self-administered, administered with the assistance of someone else, administered as a face-to-face interview), and place of administration of the questionnaires (e.g., at home, in a clinic, at the venue of a sport competition even, etc.). Finally, we provided participants an explanatory note on the project and its scientific aim. In a separate protocol paper (Puce, Marinelli, Mori, Pallecchi, & Trompetto, 2017), we reported further details on the tools used in this study.

Statistical Analysis
For the statistical analysis, we used the Shapiro-Wilk test to probe the normality of the data distributions of all the parameters (age, PGWBI22, SF12-MCS, SF12-PCS). In case of a nonsignificant Shapiro-Wilk test, we assumed a normal data distribution. We used the Mann-Whitney test to compare the parameters between the two populations and calculated corresponding U values and effect size r¼Z/ ffiffiffiffiffiffiffiffi N tot p (where N tot is the total number of participants). We also calculated Cohen's effect size d values for the average differences between the two populations, in the approximation of normal distributions. We set the level of statistical significance to p < .05 for all the tests. Table 1 illustrates participant scores on the instruments administered, grouped into the two population samples of paralympic athletes and control participants and differentiated by gender, age, mode of survey completion (either with or without assistance), and by acquired versus congenital impairment. The same results are also displayed in Figure 1, for easy visual comparison. In histograms of collected individual scores (not shown, available upon request), PGWBI scores were discretized in intervals of three units, whereas SF12 scores were demarked in intervals of two units, as a compromise between score resolution and sufficient number of data in each interval. On the Shapiro-Wilk test, we found that the only data sets exhibiting normal distributions were the PGWBI scores for both the Paralympic and control groups. Otherwise, a larger data sample would be necessary to obtain a normal distribution, given the large range of possible scores. Of note, the number of participants in our study was similar to past studies (Dykens & Cohen, 1996;Lankhorst et al., 2015;Verschuren et al., 2007;Weiss et al., 2003;Wely et al., 2010;Zwinkels et al., 2015). The histograms clearly showed that the data collected from the control sample exhibited a larger variance than those of the paralympic sample. Indeed, in Table 1, the standard deviations are almost twice as large as for the control sample. However, none of the data deviated by more than three standard deviations from the mean value in either population.

Results
For data collected from the additional questionnaires, we examined the possibility that our results could be biased due to uneven distributions of the respondents' personal data or mode of administration in the two samples, and we verified that the two populations were similar with regard to gender (49 females and 51 males in the paralympic sample, compared with 58 females and 42 males in the control sample), age (M age [years] ¼ 17.4, SD ¼ 3.7 in the paralympic sample; M age ¼ 17.0, SD ¼ 3.7 in the control sample), level of education, and health condition. We registered an imbalance for the mode of questionnaire administration (with 70% self-administered and 30% administered with assistance in the paralympic sample and 35% self-administered and 65% administered with assistance in the control sample) and for the setting in which participants lived (79% urban and 21% rural in the paralympic sample, compared with 55% urban and 45% rural in the control sample), the latter outcome possibly due to the fact that practicing sport on a daily basis is logistically easier in urban settings. Importantly, the two populations were homogeneously distributed by type of disorder, as shown in Figure 2. There was no statistical difference in the number of participants for any of the subgroups.
Our main finding was that the paralympic sample scores on the PGWBI and SF12 were higher (in the direction of greater respondent well-being) than were the scores on these instruments from the control sample: We calculated the relative percent differences between the paralympic and control groups and subgroups with respect to the values of the control group/ subgroup, whereas for subgroups of the same sample (e.g., paralympic male vs. paralympic female or paralympic aged 11-17 years vs. paralympic aged 18-23 years), we calculated the relative percent differences with respect to the intermediate values among the two subgroups. The scores of the paralympic group were higher than those of the control group by 45% for PGWBI, 40% for SF12-MCS, and 9% for SF12-PCS. Hence, our results indicate that for the PGWBI and SF12-MCS indices that assess the respondents' psychological and emotional well-being, the athlete sample obtained larger scores with statistical significance (p < .0001). Of note, for the PGWBI and SF12-MCS indices, the corresponding Cohen's effect size d values were very large (d > 1.2), according to Sawilowsky (2009) and the Mann-Whitney effect size r values were large (r > .05).
Regarding the analysis of homologous subgroups of the two samples (males, females, aged 11-17 years, aged 18-23 years, with acquired impairment, with congenital impairment, who filled questionnaires with assistance, who filled questionnaires without assistance), the PGWBI and SF12-MCS score differences between all paralympic and respective control subgroups reached statistical significance (p < .0001 for the subgroups with acquired impairment and p < .001 for all the other subgroups). On the contrary, the SF12-PCS score differences between paralympic and respective control subgroups were not statistically significant. The Cohen's d values were very large for the PGWBI score differences between all paralympic and respective control subgroups and even huge (d > 2.0, Sawilowsky, 2009) in the case of paralympic and control subgroups with acquired impairment. The Mann-Whitney effect size r values for the differences in PGWBI and SF12-MCS scores were large (r > .5) between all subgroups, except for the medium r effect size of the subgroups who filled the questionnaires with assistance (.3 < r < .5).
Regarding the differences between subgroups of the paralympic sample, we found statistical significance only for the PGWBI scores of males versus females subgroups, aged 11-17 years versus aged 18-23 years subgroups and subgroups of athletes who filled with assistance versus those who filled without assistance, but in all these cases with small r effect size (.1 < r < .3). In the case of the subgroups of the control sample, we found statistical significance only for the PGWBI and SF12-MCS scores of aged 11-17 years versus aged 18-23 years participants and for SF12-MCS scores of participants who filled with assistance versus those who filled without assistance, with medium r effect size (.3 < r < .5) in all these cases.
We now address quantitative comparisons of survey scores of different intrasample and intersample subgroups: 1. For both populations, males reported 5-10% higher well-being scores than females. Specifically, in the paralympic sample, male scores (relative to female scores) were higher on the PGWBI by 10%, the SF12-PCS by 5%, and the SF12-MCS by 6%; in the control sample, male scores (relative to female scores) were higher on the PGWBI by 6%, the SF12-PCS by 4%, and the SF12-MCS by 10%. However, statistical significance between gender subgroups within the same sample was found only for the PGWBI index in the paralympic sample (U ¼ 832, p ¼ .0040, r ¼ .29, d ¼ 0.62). 2. For both populations, younger participants, aged 11-17 years, reported higher psychological and emotional well-being than older participants, aged 18-23 years. For the paralympic sample, younger participants scored higher than older participants by 7% on the PGWBI, and by 7% on the SF12-MCS, but were not different (À0.4%) on the SF12-PCS. For the control sample, younger participants scored higher than older participants by 28% on the PGWBI, 23% on the SF12-MCS, and 7% on the SF12-PCS. Statistical significance between age subgroups within the same sample was found only for the PGWBI index in the paralympic sample (U ¼ 879, p ¼ .0183, r ¼ .24, d ¼ 0.46) and for the PGWBI (U ¼ 735.5, p ¼ .00076, r ¼ .34, d ¼ 0.79) and SF12-MCS (U ¼ 786.5, p ¼ .0026, r ¼ .30, d ¼ 0.66) indices of the control sample. Among the comparison between the homologous age subgroups of the two samples, the largest difference was found in the PGWBI scores of the participants aged 18-23 years that were 64% higher in the paralympic group than in the control group (U ¼ 175, p < .001, r ¼ .69, d ¼ 1.74). 3. Regarding assistance in completing questionnaires, for participants in the paralympic sample, those using assistance (relative to those relying on selfadministered questionnaires) were larger by 7% for the PGWBI, 6% for the SF12-MCS, and 5% for the SF12-PICS, with statistical significance only for the PGWBI index (U ¼ 783.5, p ¼ 0.046, r ¼ .20, d ¼ 0.45). For control participants, the opposite was true, in that those relying on self-administered questionnaires were a higher number on the PGWBI by 12%, the SF12-MCS by 23%, and were equivalent on the SF12-PCS (higher by only 0.7%), with statistical significance only for the SF12-MCS index (U ¼ 675, p ¼ .00084, r ¼ .33, d ¼ 0.72). 4. Regarding whether the nature of the respondents' impairment (i.e., acquired or congenital) affected survey results, there was a much greater influence among control participants, while paralympic participants with either congenital or acquired impairments responded similarly. Control participants who reported acquired impairment (vs. congenital impairments) scored higher by 96% on PGWBI, 75% on SF12-MCS, and 27% on SF12-PCS. However, these intrasample subgroup differences were not statistically significant. Among intersample comparison of homologous impairment type subgroups, the largest difference was in the PGWBI scores of the participants with acquired impairment that were higher in the paralympic group by 96%, with statistical significance (U ¼ 121.5, p ¼ .00014, r ¼ .69, d ¼ 2.25). 5. Partial scores on the survey instruments for selected PGWBI domains are shown in Figure 3. The values for the paralympic sample (compared with the control sample) were all in the direction of statistically significant better physical and psychological well-being with specific subtest scores as follows: positive well-being of 13.54/20 higher by 71% (U ¼ 1581, p < .00001, r ¼ .  , 9, 15, and 20), anxiety (PGWBI items 5,8,17,19,and 22), depressed mood (PGWBI items 3, 7, and 11), and vitality (PGWBI items 6, 12, 16, and 21) were higher by 30% to 70%. Among participants in the paralympic sample, there was no evidence of any specific health condition for which systematically higher or lower scores were registered. On the other hand, in the control population, the scores of the six participants who were affected by MS were systematically lowest, especially PGWBI scores, but also SF12-MCS scores. While the number of participants in each impairment subsample (i.e., 10) was too small to test whether this finding was statistically significant, these very low scores merit specific mention. For the control sample, the average PGWBI score of MS participants was five times lower than the average PGWBI score of the whole population (À413%), and the average SF12-PCS and SF12-MCS scores were lower by as much as À53% and À98%, respectively. Within the paralympic sample, the scores of the MS population were also lower than average, though they were significantly higher than the scores of MS participants in the control group, with the PGWBI score higher among Paralympic MS respondents by 84%, the SF12-MCS score higher by 65%, and the SF12-PCS score higher by 18%.

Discussion
The key result of this investigation was that physical and psychological wellbeing indices, particularly the PGWBI and the SF12-MCS addressing psychological well-being, were larger by 40% or more for participants with impairments who practice competitive sport relative to participants with impairments who do not practice competitive sport. This finding supported our original hypothesis. Focusing particularly on domains that assessed positive well-being, anxiety, depressed mood, and vitality, we found the scores of paralympic participants to be higher by 30-70%, relative to those of other young people with physical and intellectual impairment who were not practicing competitive sport. On the contrary, the difference on the SF12-PCS scores between the two samples was not statistically significant. We attribute the latter finding to the large heterogeneity of the impairment types in the two populations, which makes comparison of self-perceived physical well-being less meaningful. Indeed, this work is focused on the psychological and emotional well-being described by the PGWBI and SF12-MCS indices, yet reporting all results so as to be complete.
There are a number of possible social and psychological mechanisms triggered by the practice of competitive sport (we consider competitive swimming representative of competitive sport generally) that may contribute to an improved self-perceived well-being, a motivated attitude, and reduced anxiety and depressive symptoms. Some mechanisms are strictly related to the physical benefits of sport activities. Sport that is adapted to a physical condition is an actual therapy that can be prescribed on a daily basis (Puce et al., 2018). Indeed, daily training sessions carried out to improve one's performance are aimed toward not only winning medals but also medical treatment, either direct (e.g., intended to increase proportions of connective tissue in targeted muscles) or indirect (e.g., intended to inhibit negative secondary effects of the pathology; Jarvinen, Jozsa, Kannus, Jarvinen, & Jarvinen, 2002;Trompetto et al., 2014). In addition, intensive sport training maximizes residual motor skills, develops new skills, and improves individual kinesthetic perception, all of which may be helpful in the everyday activities of people affected by various impairments. There are also purely psychological, social, and motivational mechanisms unique to competitive sport practice that may be beneficial to self-perceived well-being. It is well known that regular physical activity in adults, children, and adolescents improves health and reduces mood disorders (Eime et al., 2013). Also, in people affected by impairments, physical activity reduces anxiety and depression and increases self-esteem and self-efficacy (Shephard, 1991). However, competitive physical activity can lead to a slight increase in stress and anxiety if there is excessive environmental pressure (Nieman, 2002).
Mental well-being includes several factors, such as optimism, self-esteem, having a life purpose, a sense of belonging, and building strong and positive relationships with others. Self-esteem is a key factor in psychological well-being, and it reflects a belief and confidence in one's own ability and value. This concept is fundamental for sports players. Indeed, Fox and Corbin (1989) described self-esteem in sport as a multidimensional concept consisting of strength, conditioning, bodily attractiveness, and sporting competence. However, some researchers suggest that people affected by impairments may have lower selfesteem than able-bodied people, and Bredahl (2013) found that 75% of adults with impairment had negative experiences in sport. By contrast, adolescents with intellectual impairment taking part in swimming programs showed significant improvements in their sporting self-esteem when they were competing against others in their own impairment group (Wright & Cowden, 1986). Similarly, in our sample, increased sports self-esteem may have been due to the fact that these individuals were competing against others in their own impairment group, highlighting the importance of belonging to a peer group that shares one's experiences.
In sport events, a peculiar feeling is created in which the competitive spirit overcomes any tendency to commiserate over the physical or intellectual conditions of the athletes. This agonistic attitude may cancel out stereotypical responses to the pathology (Pack et al., 2017) and encourages overcoming one's limits. Indeed, athletes with impairments have comparable levels of athletic identity to athletes who do not have any impairment, and they compete in sport at similar levels of intensity (Groff & Zabriskie, 2006;Martin et al., 1995). The practice of competitive sport shifts the focus from impairment to competency and proficiency at a given activity (Pack et al., 2017). Within paralympics, any losing athlete is beaten by a peer who is affected by similar impairment; hence, any excuse for defeat is unacceptable and is rather replaced by self-direction toward stronger dedication and effort and a willing motivation to improve (Giacobbi et al., 2008;Kokaridas et al., 2009;Martin et al., 1995). On the opposite side, among participants who do not practice competitive sport, impairments are more likely to cause a depressed mood. Among our participants, neuroprogressive syndromes such as MS were associated with strongly depressed psychological conditions.
Regarding possible medical and economic implications of our findings of improved psychophysical fitness among competitive sport athletes affected by impairments, rehabilitation effectiveness of sport-related training sessions may be superior to typical weekly rehabilitation training, not only physically (Runciman, Derman, Ferreira, Albertus-Kajee, & Tucker, 2015;Runciman, Tucker, Ferreira, Albertus-Kajee, & Derman, 2016) but also emotionally and psychologically. For example, structuring rehabilitation programs according to periods targeted at improving medium-and long-term physical performances and scheduling collective performance tests, as in the case of competitive training, could have a huge motivational role. Effectiveness of such novel rehabilitation programs would leverage on the ability of the impaired participants to take advantage of the specificities of their body and get motivated by pursuing and sharing personal goals, with benefits in different aspects of life. Moreover, transferring knowledge about training load parameters such as volume, intensity, and recovery times, properly adapted to the type and severity of individual impairments, from the competitive sport field to the rehabilitation protocols could yield significant benefit to psychophysical well-being. As a consequence, the health-care system may benefit from promoting paralympic sport participation not only from a medical point of view but also from an economic perspective, as significant rehabilitative cost savings are likely. Indeed, enhanced psychophysical condition favors improved autonomy in daily activities and helps tackling negative effects of progressive and nonprogressive disorders, both of which could help saving costs of the health-care system.
Considering secondary outcomes of our data, as shown in Table 1, the much larger variance of survey responses among participants in our control versus paralympic population suggests that there was a common mind-set among paralympic athletes, possibly related to the fact that they share the same community, lifestyles, and ambitions. Perhaps this occurred because 22 of the 100 paralympic participants completed the survey at the venue of a competitive event and were surrounded by peers at the time. In contrast, all the control participants completed questionnaires online, probably alone. Past researchers showed that the mode of questionnaire administration can affect the obtained scores (Bowling, 2005;Grossi et al., 2005;Gundy & Aaronson, 2010;Hood et al., 2012;Weinberger, Oddone, Samsa, & Landsman, 1996). Furthermore, as approximately 20% of the of the paralympic group completed the questionnaire during the event, their responses may have been influenced by higher than normal levels of endorphins that are known to improve mood during or after acute exercise (Dishman & O'Connor, 2009). Endorphins modulate dopamine neurons involved in motivation, pleasure, and the reward system to enhance moods indirectly.
We found that assistance in the administration of the questionnaires played only a minor role, if any, in the respondents' answers, consistent with findings in other studies (Puhan, Ahuja, Van Natta, Ackatz, & Meinert, 2011;Ramos-Jorge et al., 2012;Rutherford, Nixon, Brown, Lamping, & Cano, 2014;Tsakos, Bernabe´, O'Brien, Sheiham, & de Oliveira, 2008). Among participants who had help completing questionnaires in our study (relative to those who completed questionnaires without help), scores of paralympic participants were 6% higher while those of control participants were around À10% lower. However, it is worth noting that there is an interesting imbalance in the numbers of self-administered and administered-with-assistance questionnaires between the paralympic and control samples, namely a larger fraction of questionnaires were administered with assistance in the control sample. This possibly suggests greater impairment in the control group, which may have influenced the self-perception of well-being.
Regarding gender, we observed that self-perceived well-being scores of males were 5-10% higher than that of females. However, because this finding was evident for both paralympic and control participants and was consistent with past findings (Gestsdottir et al., 2015;Harada et al., 2008;Li, Kao, & Wu, 2015), this observation is beyond the scope of our study and does not warrant extensive discussion here. Similarly, our finding that younger participants, aged 11-17 years, perceived higher psychological and emotional well-being than older participants, aged 18-23 years, applied to both populations. However, this finding was accentuated among control participants, mainly due to older participants' differences between the paralympic and control groups (64% difference in PGWBI scores, p < .0001). It seems that, with increasing age, there is a stronger awareness and sense of either goal achievement or goal discrepancy (Cheng, 2004;Nilsson, Leppert, Simonsson, & Starrin, 2010) and that this contributes to reinforce both positive and negatives among older participants.
It is worth noting that among control participants, congenital (vs. acquired) impairment was associated with higher well-being scores. This may be due to the greater difficulty accepting and adapting to a later occurring impairment condition. Yet, this difficulty seemed to be attenuated by the competitive sport experience of participants in the paralympic group in which scores from those with congenital and acquired impairments were similar. The difference between PGWBI scores of Paralympic and control acquired impairment subgroups was as large as 96% (p ¼ .00014).
Among the limitations of this study were the size and geographical constraints of the two participant populations, perhaps limiting the study's generalizability. Similarly, while competitive swimming is considered representative of competitive sport generally, future studies might investigate these variables among participants in other competitive sports, perhaps comparing team and individual sports. Another limitation of the study could be related to the sample size; indeed, high variance values of sample scores may indicate that a larger sample size is needed to enhance the statistical power of the instrument. Specifically for the case of participants affected by severe and progressive syndromes such as MS, further investigations over larger numbers could confirm the dramatic effectiveness of practicing competitive sport on mental well-being found in this study. In addition, this study relied entirely on participants' selfreported physical and psychological well-being; future studies might extend this research paradigm by seeking input regarding the participants' well-being by surveying others who know them well. Finally, among the most important challenges to our research is the possibility that our paralympic participants' positive attitudes developed independent from their competitive sport engagement or that they may have contributed to this group having self-selected competitive sport practice. As this was a cross-sectional research design, we can not be certain of a causal direction from sport practice to higher well-being scores. Longitudinal research is needed to affirm that assumption.

Conclusions
In this study of 100 paralympic and 100 control participants with physical and intellectual impairment conditions, the self-reported psychological and emotional well-being was higher by about 40% among the group who regularly practiced competitive sport. This finding was particularly strong for positive well-being, anxiety, depressed mood, and vitality, for which paralympic participants scored higher by 30-70%. We also found that the higher selfperceived well-being of the paralympic group as compared with the control group increased with older (i.e., older than 18 years) versus younger participants Also, somewhat remarkably, among paralympic participants, there were similar well-being reports for those with congenital and acquired impairments, while among control participants, perceived well-being was lower among those with acquired conditions. Thus, sport participation seemed to attenuate the negative impact of acquired impairments.
We suggest some possible social and psychological mechanisms triggered by the practice of competitive sport that may contribute to the improvement of the self-perceived well-being and motivated attitude (Eime et al., 2013;Fleisher & Zames, 2001;Fox & Corbin, 1989;Giacobbi et al., 2008;Kokaridas et al., 2009;Martin et al., 1995;Pack et al., 2017;Shephard, 1991;Wright & Cowen, 1986), namely motivation toward objectives, self-commitment as opposed to self-pity, acceptance of defeat, resilience, scarce liability to depression, propensity to mutual assistance, and development of long-lasting relationships.
We believe that all these factors have a direct impact on psychophysical wellbeing and general health of young people affected by impairments, making our results worthy of consideration in designing rehabilitations programs. These programs could be structured in training period cycles targeted at medium and long terms scheduled physical performance tests, possibly organized in groups, thus providing a motivational and social added value. Moreover, they could include high-intensity training sessions that are properly adapted to the type and severity of disorder. These results could have societal benefit through promoting inclusivity and the expression of the full potential of individuals with disabilities and through health-care cost savings. In perspective, we plan to carry out a similar study on paralympic athletes who practice different sports, to compare well-being scores between athletes of different sports (more or less dynamic) and between athletes of team versus individual sports.

Authors' Note
This study was the subject of an oral presentation at the 45th national congress of the Italian Society of Physical and Rehabilitation Medicine SIMFER (Genoa, Italy, October 22-25, 2017).

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical Approval
The project was approved by the local ethics committee CER-Liguria (regional ethical committee of region Liguria, Italy), reference number 351REG2017. The participation of subjects was on voluntary and anonymous basis. Participants, or participants' parents in case of participants under legal age, gave informed consent to engage in this research.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article. Note 1. Throughout this work, we use the term Paralympic swimmers to indicate those athletes who hold an Athlete's license issued by either the IPC or the FINP, valid for the relevant season.

ORCID iD
Luca Puce https://orcid.org/0000-0003-0825-2707 Laura Mori, MD, PhD, is an associate professor at the Department of Neurosciences, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health of University of Genoa and Physiatrist Physician affiliated with the National Health System at the IRCCS Ospedale Policlinico San Martino of Genoa. She has national and international scientific collaborations regarding different topics on neurorehabilitation. Particularly her interest are in developing and carrying out projects regarding Parkinson s disease, stroke, neuromuscular diseases, cerebellar ataxia, cognitive, swallowing, and motor recovery.
Marina Simonini is specialist in Physical Medicine and Rehabilitation. She is currently Director of the ASL 3 Genovese Rehabilitation Department. She is active in the research on motor control and movement analysis in acute and chronic neurological disabilities.