See All Evidence Sections
Respiratory Management (Rehab Phase)

Exercise training

As with able-bodied individuals, there is strong evidence in support for the use of exercise training for improving cardiovascular health among people with SCI (see Cardiovascular chapter). This is important because there is a high incidence of physical inactivity in individuals with SCI and as such, they are at increased risk of secondary conditions such as cardiovascular disease, diabetes, osteoporosis and obesity. There is clear evidence that the cardiovascular and skeletal muscle systems adapt positively to exercise training in both able-bodied and SCI people. However, the lungs and airways do not change appreciably in response to exercise training. It is likely that exercise is not sufficiently stressful to warrant an adaptive response. This may be even more so when considering the small muscle mass used in wheelchair propulsion or arm cranking exercise. On the other hand, respiratory muscles are both metabolically and structurally plastic and they respond to exercise training. This statement is based largely on direct evidence from animal models and indirect evidence from able-bodied humans.

Exercise training may influence the control of breathing and respiratory sensations (i.e., dyspnea). It is generally accepted that exercise training results in a lower minute ventilation at any given absolute oxygen consumption or power output. This is likely due to a reduction in one or more of the mechanisms (neural and/or humoral) purported to cause the hyperpnea (increased respiratory rate) associated with exercise. As such, the positive effects of exercise training in SCI may reside in an increase in respiratory muscle strength and endurance as well as a reduced ventilatory demand during exercise. A lower ventilation and/or sensation of dyspnea during exercise would lower the work of breathing and prevent early termination of exercise, respectively.

Author Year; Country
Research Design
Total Sample Size





Jung et al. 2014; South Korea Randomized Controlled Trial PEDro=5 N=20

Population: N=20 with SCI (12M, 8F)
Mean (SD) age: 46.6 (10.5) years
Mean (SD) DOI: 8.45 (3.56) years
Injury level C8-L5, AIS-B to D
Treatment: Aqua group (10, aquatic exercise)
Land group (10, control)
Both groups performed upper extremity exercises; 1h sessions 3 times/week for 8 weeks
Outcome Measures:
Forced vital capacity (FVC), forced expiratory flow rate (FER), forced expiratory volume in 1s (FEV1), FEV1-FVC ratio, (FEV1/FVC)

  1. Significant between-group difference in change values of FVC (Aqua=1.8±1.3L, Land=0.31±1.6L; mean±SD) and FEV1 (Aqua=1.1±1.2L, Land=0.21±0.3L).
  2. Significant within-group increase in FVC (2.5±0.7 to 4.3±1.4L), FER (80.5±15.5 to 90.5±17.0L/s), FEV1 (2.1±0.9 to 3.2±1.2L) and FEV1/FVC (89.3±3.8 to 93.0±3.6%) in aqua group.
  3. Significant within-group increase in FER (85.2±18.0 to 90.6±18.0L/s) in land group.
Effect Sizes: Forest plot of standardized mean differences (SMD ± 95%C.I.) as calculated from pre- and post-intervention data

Brurok et al. 2013; Norway Cross-over repeated measures N=15

Population: N=15 AIS-A SCI individuals
Mean (SD) age: 39.0 (12.9) years
Mean (SD) DOI: 13.2 (10.8) years
ACE: arm cycling
FESH: functional electrical stimulation (FES) hybrid cycling (leg cycling + ACE)
FESIH: FES iso hybrid cycling (lower extremity pulsed isometric muscle
contractions + ACE)
Outcome Measures: Mean peak ventilation (VE) and other physiological measures

  1. Significantly higher VE during FESIH (mean increase +8.21L/min) and during FESH (+11.0L/min) compared to ACE in individuals with SCI above T6.
  2. No significant difference in VE during FESIH and during FESH compared to ACE in individuals with SCI below T6.

Tiftik et al., 2015; Turkey Controlled Trial

Population: N=52 with SCI (40M, 12F)
Mean (SD) age: 33.4 (13.9) years
Mean (SD) DOI: 12.6 (13.0) months
18 AIS-A, 34 AIS-B/C/D
44 traumatic SCI, 8 non-traumatic SCI
17 cervical, 15 thoracic, 20 lumbosacral
Treatment: Group A (26): locomotor training (LT, using body weight supported treadmill training) + conventional rehab program; Group B (26): conventional rehab program only
Outcome Measures: VC, FVC, FEV1, FEV1/FVC, forced expiratory flow rate 25-75% (FEV25-75), PEFR, MVV

  1. Significant increase in FVC (3.5±0.8 to 3.6±0.9L; mean±SD), FEV1 (3.1±0.7 to 3.2±0.7L), FEV25-75 (3.8±1.0 to 4.0±1.1L) and VC (3.4±0.9 to 3.6±0.9L) in group A only
  2. Significant increase in FVC and VC in all group A subgroups after stratifying for injury completeness and severity.

    Significant increase in MVV in both groups (Group A: 82.3±22.8 to 89.1±24.8L/min; Group B: 76.4±18.2 to 84.4±23.9L/min)

Taylor et al. 2014; United States
Pre-post N=14

Population: N=14 SCI individuals (13M 1F)
Mean age (SD): 39.2(3.3)
Mean DOI (SD): 9.7(2.6) years
All AIS-A, level T3-T11
6 months of functional electrical stimulation row training
Outcome Measures:
Peak minute ventilation, peak aerobic capacity

  1. Significantly increased peak minute ventilation after training
  2. Significant relation between level of injury and peak minute ventilation before and after training

Terson de Paleville et al. 2013; United States
Pre-post N=8

Population: N=8 AIS-A tetraplegic SCI (7M, 1F) individuals
Mean (SD) age: 37 (18) years
Mean (SD) DOI: 25 (12) months
5 cervical, 3 thoracic
Treatment: Locomotor training (LT) with body weight support and treadmill
Outcome Measures: FVC, FEV1, MIP, MEP, respiratory muscle surface electromyography (sEMG) and respiratory motor control assessment

  1. Significantly increased FVC, MIP, MEP, FEV1 post-LT compared to pre-LT
  2. Significantly less baseline overall sEMG activity in SCI compared to NI*
  3. Significantly increased overall sEMG activity post-LT for all tasks**
  4. 7 participants had increased sEMG amplitudes for all tasks** post-LT
  5. No significantly changes in distribution of sEMG activity post-LT for all tasks**
  6. 1 subject developed activation in muscles post-LT which were not activated pre-LT
  7. Lower rate of muscle unit recruitment in SCI patients compered to NI*
  8. Significantly faster muscle unit recruitment post-LT compared to pre-LT
  9. *Non-injured controls (NI), 9M 5F

**Cough, inspiration/expiration tasks

Moreno et al. 2013; Brazil
Pre-post N=15

Population: N=15 male tetraplegic individuals with SCI divided into control (n=7) and rugby players (n=8) groups.
Control group: mean (SD) age: 33(9) yrs; DOI: 73(53) months.
Rugby player group: mean (SD) age: 26(6) yrs; DOI: 87(52) months.
Treatment: Experimental group participated in a regular 1-year wheelchair rugby training program that involved stretching, strength exercises, and cardiovascular resistance training (2-hour sessions 3- 4x per week).
Outcome measures: FVC, FEV1, MVV

  1. There was a significant increase in all variables after training: mean (SD) FVC increased from 2.7 (0.9) L to 3.0 (1.0) L; FEV1 increased from 2.5 (0.9) to 2.8 (1.0) L; MVV increased from 107 (28) to 114 (24) L/min. However, comparisons with the control group are not presented.

Lee et al. 2012; Korea Prospective cohort

Population: N=38 C-SCI patients divided into experimental (Mechanical Insufflation-Exsufflation [MI- E] Feedback Resistive training) (n=19) and control groups (n=19). MI-EFRT group: 17M 2F; mean (SD) age: 45.7 (3.4) yrs; DOI: 20.0(1.5) months.
Control group: 16M 3F; mean (SD) age: 50.1(3.6); DOI: 21.4(1.2) months
Treatment: Joint mobilization, stretching, and muscle strengthening for both groups 2x / day for 30 minutes, 5 x per week over 4-week period. A forced positive measure MI-E, along with expiratory muscle feedback respiration exercise was practiced by the experimental group, each for 15 mins.
Outcome measures: Lung capacity, FVC, FEV1, FEV1/FVC

  1. In the comparison of the values of respiratory function before and after the respiratory rehabilitation treatment, the experimental group showed a significant increase in VC(SD) from 42.3(4.9) to 47.0(4.7)%, FEV1 from 1.3(1.1) to 1.5(0.1)L, and UPCF from 153.4(29.0) to 188.1(30.2) L/min.
  2. Treatment had no significant effect on FEV1/FVC.
  3. In the comparison of changes in respiratory function after the respiratory rehabilitation treatment between the experimental and control group, there were significant differences between the changes in VC%(%), FEV1 (L), and UPCF (L/min).

Jacobs 2009; USA Prospective Controlled Trial
N = 18

Population: N=18 SCI participants with complete motor paraplegia (level of injury T6-T10); participants were assigned either resistjve training (RT) or endurance training (ET): RT group: 6M 3F; mean(SD) age: 33.8 (8.0) yrs
ET group: 6M 3F; age: 29.0(9.9) yrs
Treatment: ET: 30 min of arm cranking exercise 3 times per week for 12 weeks; RT: similar training but with training weights gradually increased every week
Outcome Measures: Peak oxygen uptake (VO2peak); peak values of minute ventilation (VEpeak)

  1. Significant increase in VO2peak in resistive training group (15.1%) and endurance training group (11.8%).
  2. No significant change in VEpeak in either group.

Janssen & Pringle 2008; The Netherlands Pre-Post
N = 12

Population: N=12 men with SCI (6 with tetraplegia and 6 paraplegia), including 4 participants (mean (SD) age: 44(14) yrs, DOI: 13(8) yrs) who had previous training on ES-LCE
Treatment: Computer controlled electrical stimulation induced leg cycle ergometry (ES-LCE); total of 18 training sessions with each session lasting 25-30 minutes
Outcome Measures: Oxygen uptake (VO2); Carbon dioxide production (VCO2); pulmonary ventilation (VE)

  1. Significantly higher peak values for VO2 (+29%), VCO2 (+22%), and VE (+19%)

Valent et al 2008; The Netherlands Cohort
N = 137

Population: N=137 SCI participants; C5 or lower; aged 18-65 years. Hand cycling group: 35 participants with paraplegia, 20 with tetraplegia. Non-hand cycling group: 56 with paraplegia, 26 with tetraplegia.
Treatment: All participants followed the usual care rehabilitation program in their own rehabilitation centres, with or without regular hand cycling exercise. Study included three measurements: 1) when participants could sit in a wheelchair for three hours; 2) on discharge; 3) 1 year after discharge.
Outcome Measures: Peak oxygen uptake (VO2peak); FVC; PEFR

  1. Significant increase (26% in hand cycling group vs 8% non-hand cycling group) in VO2peak in paraplegic patients, whereas tetraplegic patients showed no change
  2. No change in pulmonary function (FVC or PEFR) found in either participants with paraplegia or tetraplegia.

de Carvalho et al. 2006; Brazil
Prospective Controlled Trial

Population: (1) Treatment group: 11 males with complete tetraplegia, ages 22-50, C4-C7, 25-180 months post-injury
(2) Control group: 10 males with complete tetraplegia, ages 23-42, C5-C8, 24-113 months post-injury
Treatment: Treadmill training with NMES: 20 min 30-50% BWS, 2x/wk. Conventional physiotherapy for 2. control group.
Outcome Measures: Metabolic and cardiorespiratory responses before and after training.

  1. Significant differences were found in all parameters after treadmill training with NMES, except for HR and diastolic BP. During gait, VO2 increased by 36%, VCO2 increased by 43%, VE increased by 30%, and systolic BP increased by 5%.
  2. For the control group, only VO2 and VCO2 increased significantly at rest (31 and 16%, respectively) and during knee-extension exercises (26 and 17%, respectively).

Fukuoka et al. 2006; Japan
Pre-post N=8

Population: N=8 (7M 1F); mean(SD) age: 46.5(8.3)
yrs; AIS B; T7-L1.
Treatment: Wheelchair training program: 30 min at
50% HRreserve, 3x/wk, 60 day.
Outcome Measures: VO2 peak, HR.

  1. Mean VO2 peak increased with training, became significant from 30th training day onwards (baseline = 17 ml/kg/min vs. T30 = 18 ml/kg/min).
  2. Steady state HR decreased significantly by 15th training day, reached a plateau from day 15 onwards (baseline HRss = 123±11 bpm vs. at day 15 = 109±6 bpm).

Sutbeyaz et al. 2005; Turkey
Pre-post N=20

Population: N=20 people with SCI (12 men, 8 1. women), 14 complete, 6 incomplete (T6-T12), mean(SD) age: 31.3(8.2) yrs; DOI: 3.8(5.8) yrs.
Treatment: Ventilatory and upper extremity muscle exercise: 1h, 3x/wk x 6 wks; Diaphragmatic, pursed lip breathing for 15min; Air shifting for 5min; voluntary isocapneic hypernea 10min; arm-crank exercise.
Outcome measures: Spirometry.

  1. After training, FVC, FEV1, and VC, were significantly higher than the baseline values.
  2. Exercise testing showed increased peak VE and peak workload and a reduction in the ratio of physiological dead space to tidal volume compared to baseline values.

Le Foll-de-Moro et al. 2005; France
Pre-post N=6

Population: N=6 participants (5M 1F), T6- & T11/12, mean (SD) age: 29 (14) yrs; mean DOI: 94 days.
Treatment: Wheelchair Interval-training Program – 30
min (6 x 5 min bouts: 4 min moderate intensity and 1 min of high intensity) 3x/wk for 6 wks; Progressed throughout training program to achieve 50% and 80% of heart rate.
Outcome measures: Spirometry.

  1. At maximal exercise, peak VE (75%), peak fb (-13.4%), peak VT (+28.9%), and the ventilatory reserve (12.9%) improved after training. The oxygen cost of VE decreased significantly (-20%) after training.
  2. For the wheelchair test, at the same workload after training, VE and fb decreased and VT increased consistent with improved ventilatory efficiency and greater reliance on aerobic capacity after training.
  3. Spirometric values and lung volumes showed small trends towards improvement after training.

Silva et al. 1998; Brazil Pre-post N=24

Population: N=24 participants (12 people with 1. paraplegia, 12 able-bodied individuals), median age SCI: 31 yrs (range 22-54), control: 30 (range 22-52), T1-T12, all ASIA A, >3 yrs after injury.
Treatment: Arm cranking aerobic training: 30 mins, 3x/wk x 6 wks.
Outcome measures: Spirometry.

  1. After aerobic training, SCI participants showed significant increases in FVC and the ventilatory muscle endurance, so that max voluntary ventilation at 70% time values post-training were not different from the initial values of able bodied individuals.
  2. Severely limited ventilatory muscle endurance in people with paraplegia can be improved by arm cranking.

Hooker & Wells 1989; USA
Pre-post N=8

Population: N=8 SCI (4M 4F); Low intensity group: C5-T7 (age range 26-36yrs); Moderate Intensity group: C5-T9 (age range 23-36yrs)
Treatment: Aerobic training: WC ergometry 20 min 3x/wk for 8 wk
Low Intensity exercised at a power output = 50-60% of maximal heart rate.
Moderate Intensity exercised at a power output = 70-80% maximal heart rate.
Outcome measures: maximal oxygen uptake, peak power.

  1. After training, no changes to maximal oxygen uptake or peak power.
  2. No detectable changes during submaximal or maximal exercise were detected.
  3. Training intensity was insufficient, participants did not comply with the program, or study was underpowered due to small sample size and heterogeneity of subject responses.


Evidence for exercise training for the respiratory management of the SCI person includes 2 experimental, non-randomized controlled trials and 9 lower-level studies. Studies describing the acute responses to exercise in people with SCI were not included nor were studies that investigated competitive athletes with SCI. Included studies were difficult to interpret because of relatively small sample sizes, differences in exercise modality (wheelchair, arm crank exercise, body weight supported treadmill training) as well as inconsistency in the frequency, intensity and duration of exercise training. Four studies included a control group (Silva et al. 1998; de Carvalho et al. 2006; Lee et al. 2012; Moreno et al. 2013), and the control groups in the studies by de Carvalho et al. (2006), Lee et al. (2012) and Moreno et al. (2013) included subjects comparable to those in the treatment group. This is in contrast to the control group used in Silva and colleagues study that consisted of able-bodied subjects only. While healthy controls may be used for the normative values, they cannot be considered a true control group for SCI subjects.

There is insufficient evidence to strongly support exercise training as a means to improve pulmonary function or ventilatory responses to exercise in SCI people. Some evidence (Le Foll-de-Moro et al. 2005) indicated that following exercise training, peak VE, VT and ventilatory reserve improve. However, the training intensity needs to be relatively high (70-80% of maximum heart rate at a minimum of 3x/week for 6 weeks) whereas lower intensities did not show similar efficacy (Hooker & Wells 1989). Other studies, including one with level 2 evidence, show no change in pulmonary function or ventilation during exercise (Valent et al. 2008; Jacobs 2009). Although 6 months of body-weight supported treadmill training in conjunction with neuromuscular electrical stimulation was shown to be effective for improving peak measures of respiration, the intensity at which subjects worked to achieve these outcomes is unclear, as each performed according to their individual capacity (de Carvalho et al. 2006). Additional well-designed randomized controlled trials are necessary to elucidate if exercise training is an effective means by which to improve pulmonary function at rest and during exercise. Nonetheless, from the limited SCI data and the well-known able-bodied response to upper limb training, it appears that changes to exercise ventilation and ventilatory efficiency can be positively influenced.


There is level 2 evidence (based on 2 prospective controlled trials: de Carvalho et al. 2006; Tiftik et al. 2015) and level 4 evidence (based on 5 pre-post studies: Silva et al. 1998; Sutbeyaz et al. 2005; Le Foll-de-Moro et al. 2005; Fukuoka et al. 2006; Terson de Paleville et al. 2013) to support exercise training as an intervention that might improve resting and exercising respiratory function in people with SCI.

There is level 4 evidence (based on 1 pre-post study: Janssen and Pringle 2008) that computer controlled electrical stimulation induced leg cycle ergometry (ES-LCE) increases the peak values of oxygen uptake, carbon dioxide production, and pulmonary ventilation.

Related Downloads
Related Toolkits