Journal Logo

CLINICAL SCIENCES: Clinical Investigations

Muscle Pain in Athletes with Locomotor Disability


Author Information
Medicine & Science in Sports & Exercise: February 2003 - Volume 35 - Issue 2 - p 199-206
doi: 10.1249/01.MSS.0000048635.83126.D4
  • Free


Sport for people with disabilities has achieved increasing popularity in recent years (7), with wheelchair sports as one of the most popular sporting activities (24). Although the health benefits are definitely recognized (22,24), sport injuries in athletes with locomotor disability (LDA) are increasing, paralleling the increase in sport participation (8). In spite of the number of articles dealing with LDA injuries (3,6–9,13,16,20,21,27), many topics are still controversial. Some authors (16,27) maintain that LDA are more vulnerable to sport injuries than able-bodied athletes, whereas others (21) have found a lower incidence rate of sports injury in active spinal cord injured patients than able-bodied athletes. Others (8) have concluded that the frequency of injury is the same, although LDA miss more practice and competition time because of injury.

Other controversial topics arise from the analysis of literature. Curtis and Dillon (3) found an association between the number of injuries and the number of hours per week spent in training. Taylor and Williams (27) found that the incidence of injury was not influenced by training activities such as the number of miles pushed with the wheelchair in training per year, amount of speed training, number of weight training sessions, or length of race. However, they found that restarting training before the pain had disappeared seemed to be linked to further injury. The United States of America’s “Athletes with Disabilities Injury Registry” (6) found that average sports time lost was 17.02 d for all injuries, whereas shoulder injuries accounted for the greatest time lost, at 31.71 d, followed by hand/finger injuries at 27.35 d. These data necessitate further study, as do the reasons and factors behind such a large amount of time lost from activity.

As a consequence of an acute trauma, there is immediate pain due to the direct receptor stimulation, followed by successive pain due to the release of sensitizing and/or algogenic substances, such as serotonin, bradykinin, kinins, and prostaglandins, in the affected area (1,11). Chronic disorders and other particular modalities of muscle stress can have different temporal relationships with the onset of pain (4,19). Pain is a fundamental and constant symptom of acute and chronic muscle fiber damage (5), serving as a defense mechanism to protect muscles against further injury (4). Characteristics of pain such as quality, duration, onset, intensity, as well as temporal relationship with physical activity might be used to differentiate between phases, types, and grade of injuries. Because the present study is a retrospective study based on a conversation between a medical doctor and a possible LDA patient, in our questionnaire we asked for the presence of pain, as opposed to injury. This is because pain represents a symptom that can easily be remembered for its high emotional impact. Furthermore, to our knowledge, no previous studies have, at the same time, differentiated between different types of muscle pain, location of pain, severity of pain, and underlying mechanisms and time loss to sport and work participation, as a result of sport-related muscle pain (SRMP).

To address some of the above quoted issues and because pain, per se, is an important factor to investigate, SRMP has been studied alone (excluding pain related to other tissues of the body or deriving from other causes) in this research. Muscle pain can greatly reduce the mobility of wheelchair-dependent subjects, thus decreasing their activities of daily living and reducing their participation in sport and recreation. This leads to an inactive lifestyle. Furthermore, some of our unpublished data (Italian Ministry of Health Research Project—RF96-324) demonstrated that SRMP is highly prevalent in able-bodied athletes and may be higher in LDA.

The purpose of the present research was: a) to estimate the prevalence of SRMP in a sample of the national LDA population and to describe the characteristics of the syndrome (pain duration, part of the body affected, etc.); b) to report the social impact of this kind of pain, such as medical care or loss of working days (sport or job activity); and c) to identify the determinants of SRMP.



A cross-disability survey (approved by “Santa Lucia Foundation Ethics Committee,” July 11, 1997) was performed, through interviews carried out by medical doctors, to estimate SRMP period prevalence (over a period of 1 yr) in a sample of the national LDA population. The subjects consisted of both genders of between 12 and 60 yr of age and were active in competitive official sports events for the disabled. In this study, all the recruited LDA were members of the Italian Federation of Sports for Disabled (FISD). This organization includes 1866 LDA of a total of 6473 disabled athletes. FISD is the national association that organizes all competitive sport activities for disabled people in Italy. It is a member of the International Paralympic Committee (IPC) and is responsible for Italian LDA participation in the IPC sports events. For wheelchair tennis, fencing, and winter games, athletes were interviewed during the annual National Championships. For athletics and swimming, subjects were recruited from the two regional (northern and southern) qualifiers for the National Championships. These competitions for athletics and swimming allow for the greatest potential number of subjects with the widest range of athletic ability. For wheelchair basketball, four matches of the National Championship were selected from which to recruit LDA. The interviews were carried out between February 1998 and February 2000. Among 46 national sport events for LDA (6 National Championships, 4 Regional Championships, and 36 wheelchair basketball matches) over a 2-yr period, 11 events were selected from which to sample the subjects of this study. Each type of event was selected once (e.g., if interviews were carried out during the 1998 Winter Games National Championship, they were not repeated later). Of a total of 567 LDA participating in the above listed sports events, 227 LDA were randomly recruited. No subject refused to be interviewed. All subjects signed an informed consent form before being interviewed. Subjects included LDA with different types of disorders: spinal cord injured subjects, amputees, subjects with cerebral palsy, and “Les Autres.” Spinal cord injured LDA included subjects with tetraplegia or paraplegia due to any lesion of the spinal cord. Amputees included subjects with thigh or leg amputation of one or both lower limbs. Subjects with cerebral palsy included those with problems of movement and posture for ataxia and/or athetosis and/or spasticity due to a brain disorder. “Les Autres” included all athletes affected by disorders resulting in locomotor disabilities that did not fit into the previously mentioned categories. The sport group of LDA-labeled “others” included 15 skiers, 6 fencers, and 5 wheelchair tennis players.

Questionnaire and definition of SRMP.

A multiple-choice questionnaire was developed to obtain information about SRMP. It derived from an original instrument devised to measure SRMP prevalence, possible factors associated with SRMP (determinants), and its specific medical cost in able-bodied athletes (Italian Ministry of Health RF 96.324 Research Project, unpublished data). Content validity of the questionnaire used in the present study was obtained by a panel of judges (N = 9) with expertise in epidemiology, sports medicine, physiatrics, statistics, and muscle physiology, after a series of 11 meetings. The instrument was then pilot tested on 20 LDA to refine methods and definitions used in the study. Appropriate changes were then made within the questionnaires, in accordance with trouble occurring during data collecting and following suggestions provided by the interviewers and interviewees. The final questionnaire consisted of five sections (45 items). The sections were demographic information, sport activity level, presence of SRMP, characteristics of SRMP quality and onset, and the socioeconomic impact of SRMP. SRMP was defined as any muscle pain experienced during the past 12 months that occurred during sport activity (training or competition) and/or was reported as a consequence of physical exercise, causing discomfort for at least 1 d and not being related to systemic disease. Discomfort was recognized when the LDA referred a sensation that caused her or him to stop, limit, or modify their participation in physical activity.

Data collection.

Each data collector was adequately trained in questionnaire administration and in the assessment of the presence of SRMP. The collectors were all sports physicians, i.e., medical doctors either specialized in sports medicine or in the school of sports medicine (post graduate school). During each interview, after the first two sections of the questionnaire, the sports physician asked whether the LDA had experienced muscle pain in the previous year. If an affirmative answer was obtained, key questions about the localization, characteristics, and onset of pain were asked to assess the actual presence of SRMP. The localization of pain was identified by the region of the body involved (head, upper limb, upper and lower trunk, shoulder, upper and lower back, thigh, and lower part of the leg) and the corresponding muscle(s). Relevant information was taken about the subjective feeling of pain (sharp, cramping, insertional, stretching, like a contracture, or like a blow), the event promoting the muscle pain (pain occurred after a brisk movement, after a direct trauma, after an accidental fall, or on starting a new training period), and the onset of pain (during sport, after sport, at hours-days distance from sport, not identified). The pain sensation associated with the pain localization served to characterize the discomfort. Only if the sport physician believed the referred characteristics to be consistent with our definition of SRMP was the questionnaire continued. Other information related to muscle pain such as duration, intensity, symptoms, and appearance of the affected area were then recorded. Limitations to work and sport were also noted, i.e., the length of absence (days/weeks/months) from work and/or sport activities. Possible medical treatments and diagnostic aids were determined as well. LDA could answer more than one of the proposed features of the question regarding part of the body affected by pain, pain sensation, body area appearance, health care, and diagnostic aids.

Statistical analysis.

To estimate the SRMP period prevalence (the prevalence during a year’s period) in the study population, we divided the number of persons reporting SRMP by the total number surveyed. The 95% confidence intervals (95% CI) of the prevalence were calculated using the formula reported by Fleiss (10). For the identification of SRMP determinants, univariate analysis of each variable of interest (age; gender; working activity (student, unemployed, employed); body mass (kg); stature (m); body mass index (BMI, body mass in kg divided by the square of the stature in m); type of sport (swimming, basketball, track and field, and others); training volume (hours of training per week); competition frequency (number of competitions during the previous 6 months); and disorder type (spinal cord injury, amputation, cerebral palsy, and “Les Autres”)) was performed. For nominal and ordinal variables (gender, working activity, type of sport, and disorder type), we produced a contingency table of outcome versus the k levels of the independent variable. In addition to the likelihood ratio chi-square test, we estimated the prevalence odds ratios (ORs) (with 95% CI) using the reference group as the level with the lowest prevalence of SRMP. For continuous variables (age, stature, body mass, BMI, training volume, and competition frequency), categories were defined by using quartiles of the control distribution (except for training volume where tertiles were used). We fitted a univariate logistic regression model to obtain the estimated coefficient (prevalence OR), and we aggregated levels with homogeneous ORs.

Multivariate analysis.

For the identification of SRMP determinants, the variables significantly associated with SRMP in the univariate analysis (working activity, disorder type, BMI, competition frequency, and training volume) together with basic characteristics of the study population, i.e., age, gender, and type of sport, were included in a multivariate logistic regression model to evaluate the independent role of the variables. Regardless of the obtained results, stature and body mass were not included in the multivariate model because they were synthetically expressed in the calculation of BMI. Due to the small sample size, no interaction tests were performed. The alpha level was set at 0.05. All statistical analyses were performed using the SPSS statistical package (SPSS, Inc., Chicago, IL) (25,26).


Characteristics of the Sample

The number and relative percentage of sampled subjects (227 LDA) expressed by age, sex, main sport practiced, and disorder type are shown in Table 1. The sample included mostly male athletes (74.4%), participating in swimming (44.1%) and track and field events (29.5%), and with a spinal cord injury (59.9%).

Characteristics of the study population (N = 227).

Characteristics of SRMP

The most relevant features of SRMP revealed in the present survey and other information related to the behavior of the 115 subjects who experienced SRMP are reported in Table 2. The most common part affected by SRMP was the shoulder muscles (56%). Subjects reported mostly moderate (49.1%) pain intensity. Pain duration lasted less than 7 d in 71.1% of the subjects and was caused by sprinting in 33.3% of cases. The onset of pain was mostly during the sport event (45.6%). Pain sensation was described as sharp in 46% of cases. In most cases, the body area appeared contracted (62.3%). In many cases (36.8%), pain did not cause an interruption in sport activities. On the other hand, work activities were almost never interrupted (89.5%). In most cases, subjects underwent exams by a physician (78.3%) and received medication (57.1%) and/or physiotherapy (44.6%). Diagnostic aids were not required in 58.2% of cases.

Characteristics of SRMP and other related data (expressed as a percentage manner of the total 115 subjects).

Prevalence of SRMP Univariate analysis: variables associated with SRMP.

The SRMP period prevalence rate was found to be equal to 50.7% (95% CI (44.0–57.4)). This result is shown in Table 3, where the SRMP prevalence rates for each variable of interest are also shown together with crude and adjusted prevalence ORs and 95% CI. In the univariate analysis, no statistical difference was observed in gender, age, or main sport practiced. Variables significantly associated with SRMP were stature, body mass, BMI, training volume, competition frequency, working activity, and disorder type.

Observed prevalence rate, with crude and adjusted prevalence odds ratio and 95% confidence intervals (95% CI) of sport-related muscle pain (SRMP) for the different variables of interest and SRMP prevalence and 95% CI for the total sample.

Multivariate analysis: determinants of SRMP.

In the multivariate analysis, only three factors were found to be statistically associated with SMRP: a BMI between 24.6 and 30.9 (adjusted OR 3.4; 95% CI (1.5–7.4)), a training volume above 7 h·wk−1 (adjusted OR 3.8; 95% CI (1.4–10.0)), and, as regards disorder type, the categories of spinal cord injury (adjusted OR 11.2; 95% CI (3.8–33.3)) and amputation (adjusted OR 15.4; 95% CI (3.9–60.9)). The prevalence OR of SMRP as a function of training volume is shown in Figure 1. The line fitting the point has been developed with the purpose of highlighting that the prevalence OR of SRMP increases with a higher rate than the increase of time spent in training.

Sport-related muscle pain odds ratio versus training volume expressed as number of hours of training per week.


In the present study, aimed at investigating SRMP in athletes with a locomotor disability, we observed in our population a prevalence during a year’s period of 50.7% (95% CI (44.0–57.4%)). Variables considered of interest (age, gender, working activity, body mass, stature, body mass index, type of sport, training volume, competition frequency, and disorder type) were studied to find those significantly associated with SRMP. Based on a preliminary univariate statistical analysis and a successive multivariate logistic regression model training volume, body mass index and disorder type (spinal cord injured and amputee categories) were identified as determinants of SRMP.

A noteworthy result is the fact that the training volume is a SRMP determinant in athletes with a disability. In a previous study of ours on SRMP prevalence in able-bodied athletes (unpublished data of a study funded by the Italian Ministry of Health Research Project—RF96-324), we had obtained an association between SRMP and training volume similar to that of the present study. In the introduction of the present study, we reported the rationale about the link between pain and injury. Because to our knowledge there are very few previous studies where pain, as opposed to injury, has been investigated in disabled athletes (e.g., ref. 3 of the present article), we have to refer to literature about injuries. The relationship between training volume and injury is controversial in the literature. In a retrospective study carried out on 128 wheelchair athletes in 1985, Curtis and Dillon (3) noted that “a higher number of hours per week spent training was related to a higher number of injuries reported.” In this case, all kinds of injuries reported by the athletes during their sports history were taken into account and soft tissue injuries represented the highest percentage (33%) of all reported injuries. In contrast, Taylor and Williams (27), studying wheelchair racing disabled athletes, did not find significant association between occurrence of injury and a training variable, such as the distance pushed per week. Our findings are consistent with Curtis and Dillon’s results. The association between SRMP and training volume could also be explained by the fact that, differently from Taylor and Williams but similarly to Curtis and Dillon, the sample used in the present study included LDA participating in many sports activities. Careful monitoring of the training volume should be undertaken by coaches and medical personnel as part of a comprehensive injury-prevention program.

Literature about the same topic but related to physically active able-bodied athletes reveals results similar to our study. Recently Hootman et al. (14) showed a relationship between the duration of physical activity (hours per week) and an increased risk of developing activity-related injury in able-bodied subjects. Moreover, the risk of injury was positively correlated with higher measured cardiorespiratory fitness. Although these studies were carried out on able-bodied subjects and they are more focused on injury rather that SRMP, some conclusions can be drawn. Jones et al. (17) demonstrated in soldiers that a greater amount of exercise was associated with a higher risk of injury. These authors defined the total amount of exercise as the product of intensity × duration × frequency. In the present study, we had only duration and volume (hours per week) as parameters of the training dose. We did not have the possibility of estimating the total amount of work (such as the total amount of oxygen consumption) performed by our athletes during training, and therefore we could not determine the training intensity. Even within a very specific sport group, energy expenditure depends on the kind of disability, as the key factor is the total amount of recruitable muscle mass. On the other hand, the rough data of hours of training per week should still be considered a significant parameter of exposure to the sport environment, i.e., of the likelihood of experiencing pain.

The association between SRMP and training poses a cost/benefit problem. Exercise is commonly considered an effective factor in promoting physical health: the scientific community is adamant in affirming this (18). A relevant aspect of the studies performed on able-bodied subjects is the dose-response effect of an exercise regimen (18). As the dose of activity increases, the increments of health benefits are progressively reduced. Conversely, the risk of damage (cardiac or orthopaedic) is closely linked to exercise volume. Kallinen and Markku (18) showed that the increasing benefit curve (e.g., protection against heart problems) reaches a plateau at higher exercise levels and the curve assumes a half-dome shape. The same authors found that the curve of drawbacks (e.g., orthopaedic injuries) increases gradually with increasing dose of exercise, i.e., assumes the opposite shape of the benefit curve. Thus, the differential between benefits and drawbacks assumes a maximum at a certain value of exercise. This value corresponds to the optimum exercise dose. Our results show that the categories of training volume from 2 to 4 and from 5 to 7 h·wk−1 were not associated with SRMP, whereas a training volume greater than 7 h·wk−1 constituted a significant determinant associated with SRMP. If the health benefits are actually progressively limited after a certain level of training dose, we can postulate an optimum level of training (i.e., compromise between health and SRMP prevalence). Obviously, further research is required to offer sound advice on this topic. At present, our results only suggest conservative conduct in training LDA, and further prospective studies are recommended.

For BMI, the category between 24.6 and 30.9 was also found to be a determinant of SRMP with respect to the category between 14.7 and 24.5. This is an important finding from a prevention point of view, considering that BMI can be largely affected by the amount of body fat and the latter can be modified by appropriate diet and physical exercise. Integrated with skin-fold measurements or other methods to assess body fat, BMI could be considered to devise prospective studies to assess risk factors of SRMP.

Other important intrinsic determinants of SRMP were found among the categories of type of disorder. Spinal cord injured and amputee LDA were found to be more prone to SRMP than other categories of LDA. Regarding spinal cord injured LDA, this fact can be interpreted as an effect of the relatively small residual musculature in these subjects faced with the very heavy workload applied on the upper limbs. It has been noted that the demands of sport activity plus those of daily wheelchair pushing do not allow a sufficient amount of rest for healing and recovery of musculature (9). This susceptibility must be considered when training programs are designed for athletes with a disability. The amputee athlete was found to be the most liable to suffer from SRMP. We have found no solid explanation for this fact except that these athletes are probably the most active and therefore the most prone to overload their muscles. Another mechanism that we can suggest is that SRMP could be a result of lower coordination between different muscles due to a less than perfect fit of the prosthesis. Moreover, the modified biomechanics and less efficient movements that occur in these subjects might increase the possibility of experiencing SRMP. It is interesting to note that, similarly to our results, a previous retrospective study on the injury experience of wheelchair athletes found the lowest percentage of injured athletes among those with cerebral palsy (7).

The great majority of the subjects with SRMP reported a pain that was not due to direct trauma. It seems that 16 of our subjects had delayed onset muscle soreness (DOMS). This we suppose on the basis of SRMP onset (between 24 and 48 h after the sport event), the activity causing pain, and the pain sensation (23). We suppose that in these subjects an eccentric contraction was probably involved (12,19) in causing pain. In very few cases SRMP was related to a direct trauma (2.7%), whereas in a relevant number of cases (19.3%) SRMP appeared several hours after sports events. In the latter case, therefore, a completely different mechanism, to be studied in further research, could be involved. A small amount of LDA (5.3%) were not able to remember the precise onset of pain. Although these subjects still attributed SRMP to be related to previous sport activity, we cannot exclude that other reasons contributed to cause the muscle pain. A possible explanation might be that the underlying factors related to the disability might have accounted for the onset of pain. For example, we can suppose that wheelchair pushing in everyday activity contributed to causing pain. Furthermore, we cannot exclude that sport activity had produced micro-trauma, which led to subtle chronic overuse.

Curtis and Black (2) assessed the prevalence and intensity of shoulder and upper-extremity pain experienced by female wheelchair athletes during functional activities. They reported that approximately 14% of female wheelchair users had shoulder pain before their injury, whereas 72% reported shoulder pain afterward. Of this 72%, 52% were frequently in pain and 11% reported a reduction in activities of daily living due to pain. The highest intensity of shoulder pain was reported when carrying out household chores, propulsion up and down ramps, overhead lifting, and sleeping. The interest in shoulder pain is justified by the results of the present study, which revealed the highest occurrence of SRMP in the shoulder (56% of all cases). Even though this is not a surprising result considering the type of disorders of the subjects of the present study, this finding, already shown by authors who studied the “injury experience” of wheelchair athletes (7), is important because of the possible impact on their activities of daily living. Fortunately, the global results of this study show that the severity of the pain is not so high given that, different from other studies (6), SRMP had a minimal impact on working and sport activity. Indeed 89.5% of the athletes did not interrupt their working activity because of the pain and 36.8% did not interrupt their sport activity after the pain.

The present study has two main sources of limitations, one related to the questionnaire and one to the sample (the latter does not allow us to infer the SRMP prevalence of the global LDA population). The content validity of the questionnaire was based on a series of consensus meeting among experts (face validity). There was validity deriving neither from a comparison with a “gold standard” of muscle pain nor a reproducibility study. The information reliability obtained from the questionnaires should have been guaranteed by training carried out by the medical doctors who interviewed the LDA. This allowed us to ensure homogeneous results and assessment of the SRMP. About the sample, we did not follow a predefined sample design. The LDA were interviewed randomly during the Italian Championships. The interviewed athletes were not identified a priori. The selection of large sports events guaranteed that people were recruited from all parts of Italy and had different levels of performance. However, only LDA who participate in organized competitive sports activity could be part of the sample. Practically the sample included only the LDA registered with the Italian Federation of Sports for Disabled. This organization included, in the 2 yr of the study, 3083 athletes with intellectual disabilities, 937 athletes with visual impairments, and 2453 athletes with physical disabilities. Among the latter, 1866 athletes constituted the total number of LDA. This constitutes only a small number of subjects compared with the 664,000 subjects corresponding to the total Italian disabled population in the range of age between 6 and 64 yr and who do not live in hospitals or other kinds of organized residential places (15). Among these subjects, 44% of them are estimated to practice physical activities (either recreational or rehabilitative or competitive physical activity). Therefore, only a small representation of the whole complete population practicing sport is given in the study. The study could present another source of limitation. Indeed, it must be noted that an underestimation or underreporting of the real prevalence of SRMP could have occurred. Some athletes who were suffering from SRMP at the time of the meeting may have been absent from that competition precisely because of experiencing muscle pain or injury and therefore missed the interview. In this case, the real prevalence of SRMP in LDA could be even higher than that of our reported values.

The strength and the novelty of the present study was the examination of the 1-yr prevalence and the main determinants of one single kind of pain (the SRMP and not a specific or general injuries) in a wide group of LDA including athletes with different disorders and participating in different sports. Other studies in which data were collected during sports events including a large number of disabled athletes (7,9) investigated all kinds of sport-related injuries, expressing their statistics in a different time interval. Consequently, comparisons are difficult to be made. The present study is also aimed at targeting one condition (SRMP) that may have a significant impact on sports and/or working activities. Research on disabled athletes is needed to improve the knowledge on the possible impact of sport and exercise upon the quality of life (22).

In conclusion, we have shown that sport-related muscle pain has a 1-yr prevalence of 50.7% (95% CI (44.0–57.4)) in our sample of athletes with locomotor disability active in competitive sports events. We suggested training volume and body mass index as determinants associated with sport-related muscle pain. The results of this study could be used to design prospective control studies in the future in order to identify possible risks of developing sport-related muscle pain. Coaches might take advantages from this information and could opportunely set-up specific training programs to verify the importance of training volume on the occurrence of sport-related muscle pain. These observations highlight the importance of a strict collaboration among physicians, coaches, and physical trainers for the well-being and the performance of the athlete.

The authors thank Dr. John F. Ditunno for his useful advice and revision of the manuscript and Fondazione Santa Lucia for its administrative and financial support.

This study was supported by a grant from the Italian Ministry of Health (Ministero della Sanità, RF 96.324).


1. Blais, C. Jr., A. Adam, D. Massicotte, and F. Peronnet. Increase in blood bradykinin concentration after eccentric weight-training exercise in men. J. Appl. Physiol. 87: 1197–1201, 1999.
2. Curtis, K. A., and K. Black. Shoulder pain in female wheelchair basketball players. J. Orthop. Sports Phys. Ther. 29: 225–231, 1999.
3. Curtis, K. A., and D. A. Dillon. Survey of wheelchair athletic injuries: common patterns and prevention. Paraplegia 23: 170–175, 1985.
4. Ebbeling, C. B., and P. M. Clarkson. Exercise-induced muscle damage and adaptation. Sports Med. 7: 207–234, 1989.
5. Edwards, R. T. Hypothesis of peripheral and central mechanisms underlying occupational muscle pain and injury. Eur. J. Appl. Physiol. 57: 275–281, 1988.
6. Ferrara, M. S., and W. E. Buckley. Athletes with disability injury registry. Adapt. Phys. Act. Q. 13: 50–60, 1996.
7. Ferrara, M. S., W. E. Buckley, B. C. Mccann, T. J. Limbird, J. W. Powell, and R. Robl. The injury experience of the competitive athlete with a disability: prevention implications. Med. Sci. Sports Exerc. 24: 184–188, 1992.
8. Ferrara, M. S., and R. W. Davis. Injuries to elite wheelchair athletes. Paraplegia 28: 335–341, 1990.
9. Ferrara, M. S., G. R. Palutsis, S. Snouse, and R. W. Davis. A longitudinal study of injuries to athletes with disabilities. Int. J. Sports Med. 21: 221–224, 2000.
10. Fleiss J. L. Statistical Methods for Rates and Proportions, 2nd Ed. New York: Wiley, 1981, pp. 14–15.
11. Fock, S., and S. Mense. Excitatory effects of 5-hydroxytryptamine, histamine and potassium ions on muscular group IV afferent units: a comparison with bradykinin. Brain Res. 105: 459–469, 1976.
12. Garrett, W. E. Muscle strain injuries: clinical and basic aspects. Med. Sci. Sports Exerc. 22: 436–443, 1990.
13. Hoebering, J. H., H. B. L. Debets-Eggen, and P. M. L. Debets. Sports medical experiences from international Flower Marathon for disabled wheelers. Am. J. Sport Med. 18: 418–421, 1990.
14. Hootman, J. M., C. A. Macera, B. E. Ainsworth, M. Martin, C. L. Addy, and S. N. Blair. Association among physical activity level, cardiorespiratory fitness, and risk of musculoskeletal injury. Am. J. Epidemiol. 154: 251–258, 2001.
15. I stat (N ational I nstitute for S tatistics ). Health services for the disabled. In:ISTAT Annual Report: Reorganisation and Diversification in the Offer of General Services, Rome, 2000, pp. 329–332 [L’assistenza sanitaria per i disabili. In:Rapporto Annuale dell’ISTAT: Riorganizzazione e diversificazione nell’offerta di servizi collettivi].
16. Jackson, R. W., and A. Fredrickson. Sport for the physically disabled 1976 Olympiad (Toronto). Am. J. Sport Med. 7: 293–296, 1979.
17. Jones, B. H., D. N. Cowan, and J. J. Knapik. Exercise, training, and injuries. Sports Med. 18: 202–214, 1994.
18. Kallinen, M., and A. Markku. Ageing, physical activity, and sports injuries. Sports Med. 20: 41–52, 1995.
19. Kuipers, H. Exercise-induced muscle damage. Int. J. Sports Med. 15: 132–135, 1994.
20. Madorsky, J. G. B., and K. A. Curtis. Wheelchair sports medicine. Am. J. Sports Med. 12: 128–132, 1984.
21. Nilsen, R., P. Nygaard, and P. G. Bjorholt. Complications that may occur in those with spinal cord injuries, who participate in sport. Paraplegia 23: 152–158, 1985.
22. Noreau, L., and R. J. Shephard. Spinal cord injury, exercise and quality of life. Sports Med. 20: 226–250, 1995.
23. Sbriccoli, P., F. Felici, A. Rosponi, et al. Exercise induced muscle damage and recovery assessed by means of linear and non-linear sEMG analysis and ultrasonography. J. Electromyogr. Kinesiol. 11: 73–83, 2001.
24. Shephard, R. J. Sports medicine and the wheelchair athlete. Sports Med. 5: 226–247, 1988.
25. SPSS® B ase. Applications Guide, Version 9.0. Chicago: SPSS Inc., 1999, pp. 64–77, 103–117.
26. SPSS®. Regression ModelsTM, Version 9.0. Chicago: SPSS Inc., 1999, pp. 35–59.
27. Taylor, D., and T. Williams. Sports injuries in athletes with disabilities: wheelchair racing. Paraplegia 33: 296–299, 1995.


©2003The American College of Sports Medicine