See All Evidence Sections
Respiratory Management (Rehab Phase)

Inspiratory Muscle Training

As expected, the loss of inspiratory muscle function is related to the level of injury as illustrated in Figure 3. Dyspnea, defined as a subjective report of breathlessness or shortness of breath, is common in people with SCI and is greatest in people with tetraplegia (Ayas et al. 1999). Approximately two-thirds of the prevalence of dyspnea in this group is attributed to the inspiratory muscle loss (Spungen et al. 1997). Improved inspiratory muscle strength and endurance could potentially improve cough and maximal exercise ventilation in addition to decreasing dyspnea. The inspiratory muscles can be trained similar to the limb muscles with inexpensive devices that increase the resistive or threshold inspiratory load on the inspiratory muscles (Reid et al. 2004). Table 9 outlines common measures that are indicative of respiratory muscle strength and endurance. In neuromuscular disorders like SCI, maximal lung volumes that measure IC also can reflect increased inspiratory muscle strength.

Term Abbreviation Definition
Maximal inspiratory pressure MIP or PImax

Estimate of inspiratory muscle force as reflected by the maximal pressure exerted by the inspiratory muscles measured at the mouth.

Maximal expiratory pressure MEP or PEmax

Estimate of expiratory muscle force as reflected by the maximal pressure exerted by the expiratory muscles measured at the mouth.

Maximum voluntary ventilation MVV

Maximum ventilation in 15 seconds, which reflects the “sprint” capacity of the respiratory muscles. The maximum ventilation can be measured over several minutes – between 4 and 15 minutes – which is more reflective of the endurance of the respiratory muscles.

Maximal sustainable mouth pressure SIP

Maximum mouth pressure sustained during a 10 minute period of SIP threshold loading, which is usually lower than the MIP. This is an estimate of the endurance of the inspiratory muscles.

Endurance time sustained on training load Tlim

The endurance time while breathing on a resistive or threshold trainer at a defined level of the MIP.

Maximal incremental threshold load TLmax

The maximal load (usually defined as an inspiratory mouth pressure) TLmax attained on an incremental threshold loading test whereby the load is progressively increased every 2-3 minutes.

Commercially available hand-held devices can be used for inspiratory muscle training. The three main types of devices are the resistive and threshold trainers (Figure 3) and isocapneic hyperpnea (see Reid et al. 2004 for details of these training techniques). These devices have a one-way valve that closes during inspiration so that the subject must breathe against a load. The resistive trainer imposes a load through a small diameter hole whereas the threshold trainer imposes a load via a spring loaded valve. Isocapneic hypernea imposes loading in a very different manner. The participant targets a prescribed ventilation that require higher inspiratory and expiratory flows. A bag attached to the device is adjusted to match the amount of rebreathing in order to maintain isocapnea i.e. a normal end-tidal CO2level. For all of the devices, the one-way valve opens during expiration such that no load is imposed during the expiratory phase of respiration. Evidence showing decreased dyspnea and improved strength and endurance after IMT is well documented in people with other health conditions such as chronic obstructive pulmonary disease (COPD) (Reid et al. 2004; Geddes et al. 2005).

Figure 3. Inspiratory Muscle Trainers
Threshold trainer: has an adjustable spring-loaded valve that imposes the inspiratory load. The inspiratory load can be increased by winding the spring more tightly. Advantage of this trainer is that the same load is imposed on the inspiratory muscles regardless of breathing pattern.
Threshold and P-Flex trainers available from Respironics HealthScan Inc., 41 Canfield Rd., Cedar Grove, NJ7, 0009-1201. 1-800-962-1266.
Resistive trainer: has holes of different diameters.  The inspiratory load can be increased by setting the dial to holes of lesser diameter. Disadvantage of this trainer is that the subject can reduce the inspiratory load by breathing more slowly.  If this device is used for training, a target must be used. Various targets have been designed that set a breathing rate (flow and/or inspiratory pressure) for the subject.
Isocapneic Hypernea Trainer: has a rebreathing bag that can be adjusted to ensure that the person’s CO2 level is maintained within a physiologic range. A target is provided for the person to increase the level of ventilation to a training intensity. This device enables training at low loads but much higher inspiratory and expiratory flow such that the inspiratory and expiratory muscles training at higher speeds of contraction. In contrast, the threshold and resistive trainers, place high loads while the speed of contraction is relatively low.
SpiroTiger™ trainer available from FaCT Canada Consulting Ltd. 1215 Cariboo Hwy N Quesnel, BC V2J 2Y3 Canada1-877-322-8348

Author Year; Country
Research Design
Total Sample Size

Methods Outcome

Postma et al. 2014; Netherlands
Randomized Controlled Trial

Population: N=40 individuals with SCI (35M, 5F) Mean (SD) age: 46.8 (14.3) years Median (IQR) DOI: 74 (57-109) days for RIMT group & 88 (59-121) days for control group 30 tetraplegia, 10 paraplegia 24 motor complete SCI
Treatment: Resistive IMT (RIMT) group (19): 8 weeks using inspiratory muscle training (IMT) trainer + usual care; Control group (21): Usual care
Outcome Measures: FVC, FEV1, PEFR MVV, MIP, MEP, visual analogue scale for subjective breathing, and Short-Form-36

  1. Significantly greater increase in MIP in RIMT group (56.4±29.5 to 82.7±29.7cmH2O; mean±SD) than control group (56.1±23.5 to 70.7±28.1cmH2O) 1 week after intervention period, but loss of significance at 8 weeks and 1 year follow-ups.
  2. MIP improved over longer period for those who continued RIMT post-intervention, compared to those who discontinued.
  3. No significant between-group difference in changes of any other pulmonary outcome measure.
Effect Sizes: Forest plot of standardized mean differences (SMD ± 95%C.I.) as calculated from pre- and post-intervention data

West et al. 2014; United Kingdom
Randomized Controlled Trial
N = 10

Population: N=10 athletes with cervical SCI (9M, 1F) Mean (SD) age: 29.2 (2.7) years Mean (SD) DOI: 9 (2.2) years 7 AIS-A, 3 AIS-B
Treatment: IMT group (5):  6-week inspiratory muscle training (IMT); Placebo group (5)
Outcome Measures: Diaphragm thickness (DT), MIP, MEP, FEV1, peak inspiratory flow rate (PIFR), PEFR, MVV and other cardiovascular and physiological measures
  1. Increase in DT (+22% IMT vs -3% placebo) and MIP (+11% vs -6%) is significant between-groups
  2. Significant increase in MVV for both groups; increase insignificant between-groups
  3. No evidence of activity-related dyspnea in either group pre- or post-intervention
  4. No correlation between percentage change in diaphragm thickness and maximum static inspiratory pressure.

Fischer et al. 2014; Italy

Population: N=12 hand bike athletes with SCI (10M, 2F) Mean (SD) age:  43 (5.4) years Median (SD) DOI: 16.4 (7.3) years All lesions between T2-T1
Treatment: Control (5): no intervention; Experimental (7): 20 sessions of respiratory muscle endurance training (RMET)
Outcome Measures: VC, FVC, TV, maximal TV, FEV1, FEV1 / FVC, PEFR, MVV, maximal VE (VEmax), maximal fb (fRmax), respiratory endurance time and other physiological measures
  1. No significant between group changes in all resting lung function measurements.
  2. Significant within-group increase in fRmax, VEmax & respiratory endurance time after RMET in experimental group only.

Aslan et al. 2016; United States
N = 11

Population: N=11 SCI individuals (8M, 3F) Mean (SD) age:  32(9) years Median (SD) DOI: 53(72) months 10 cervical, 1 thoracic AIS-A/B/C: 3/4/4
Treatment: 1 month of respiratory motor training (RMT)
Outcome Measures: FVC, FEV1, MIP, MEP, respiratory rate and other physiological measures
  1. Significantly increased FVC after RMT.
  2. No significant changes in other pulmonary measures.

Tamplin et al. 2013; Australia

Population: N=24 individuals with chronic tetraplegia (C4-C8, AIS A & B) were randomized to the experimental group (n=13) or control group (n=11). Intervention group: mean (SD) age: 44 (15) yrs; DOI: 13(7) yrs. Control group: mean (SD) age: 47(13) yrs; DOI: 8(6) yrs.
Treatment: The experimental group received group singing training 3 times weekly for 12 weeks. The control group received group music appreciation and relaxation for 12 weeks. Assessments were conducted pre, mid-, immediately post-, and 6-months postintervention.
Outcome measures: Standard respiratory function testing, surface EMG from accessory respiratory muscles; sound pressure levels during vocal tasks, assessments of voice quality, voice handicap index, profile of mood states, and assessment of quality of life

  1. The singing group increased projected speech intensity and maximum phonation length significantly more than the control group.
  2. Both groups demonstrated an improvement in mood, which was maintained in the music appreciation and relaxation group after 6 months.
  3. No change in respiratory muscle strength was shown.
Effect Sizes: Forest plot of standardized mean differences (SMD ± 95%C.I.) as calculated from pre- and post-intervention data

Van Houtte et al. 2008; Belgium

Population: C4-T11 AIS A,B, or C; 2-6 months since injury
Treatment:  sham or normocapnic hyperpnea training  for 15-30 min x 8 wks; average of 27 sham and  28 training sessions
Outcome measures: MIP, VC, MVV, respiratory muscle endurance, respiratory infections.
  1. Significant increase in MIP, VC, MVV, and respiratory muscle endurance. and lung volumes after IMT
  2. Number of respiratory infections was less in the training than the sham group (1 versus 14).

Mueller et al. 2012 & 2013; Switzerland

Population: N=24 individuals with traumatic complete tetraplegia (C5-C8, AIS A) were randomly assigned to 1 of 3 groups. Placebo group: 6M 2F; mean (SD) age: 41.6(17.0) yrs; DOI: 6.6(1.4) months. Isocapnic hyperpnea (IH) group: 6M 2F; mean (SD) age: 33.5(11.7) yrs; DOI: 6.6(0.9) months. Inspiratory resistive training (IRT) group: 6M 2F; mean (SD) age: 35.2(12.7) yrs; DOI: 6.0(0.0) months.
Treatment: All participants completed 32 supervised training sessions over 8 weeks.
Outcome measures: inspiratory and expiratory muscle strength

  1. Compared to placebo training, IRT showed high effect sizes for inspiratory muscle strength (d=1.19), VAS values of “cleaning the nose” (d=0.99), and the physical component of subjective quality of life (d=0.84).
  2. IH compared with placebo showed a high effect size for breathlessness during exercise (d=0.81).
  3. Friedman analysis showed a significant effect for IRT versus placebo and versus IH on inspiratory muscle strength.
Effect Sizes: Forest plot of standardized mean differences (SMD ± 95%C.I.) as calculated from pre- and post-intervention (IH and IRT respectively) data

Loveridge et al. 1989; Canada
N = 12

Population: 12 participants with complete motor loss below C6-C7 (n=6 control, n=6 training) >1yr post injury, mean(SD) age IMT:31(4.1) yrs, Controls: 35(12) yrs.
Treatment: Resistive IMT without target at 85% SIP for 15 minutes twice daily, 5 days per wk x 8 wks.
Outcome measures: Spirometry.
  1. Increase in MIP and Maximal sustainable mouth pressure (SIP) in both the control group (30%±19% and 31%±18% respectively), and IMT group (42% + 24% and 78%±49% respectively) but no difference in post-training improvements between groups.
  2. The increased MIP and SIP resulted in a slower and deeper breathing pattern and a significantly shorter inspiratory time:total time of respiratory cycle in both trainers and control participants.

Liaw et al. 2000; Taiwan

Population: N=30 participants with SCI (C4-C7, 30-134 post-injury); 20 participants completed (13 control,17 IMT group), 8M:2F in each group, mean(SD) age RIMT:30.9(11.6) yrs; control: 36.5(11.5) yrs
Treatment: Target resistive IMT or control; 15-20min 2x/day x 6wks; other rehab activities continued.
Outcome measures: Spirometry, MIP.
  1. Pre-post % change of VC and TLC in IMT group was greater compared to change in control values.
  2. MIP improved in both groups which might be due to natural progression of improvement from SCI, learning to do the maneuver, and/or insufficient length of training.

Derrickson et al. 1992; USA

Population: 40 participants met admission criteria; 11 participants completed (9M 2F), neurologically complete, C4-C7; 2-74 days post-injury, studied at >24hrs after spontaneous breathing, mean (SD) age: 28.5(5.6) yrs.
Treatment: Resistive IMT without target (n=6) and (n=5) breathing with abdominal weights, 5 days per wk x 7 wks.
Outcome Measures: Spirometry.
  1. Significant improvements within both groups in FVC, MVV, PEFR and MIP between week 1 and week 7.
  2. No significant differences between treatment groups for any of the improvements in pulmonary variables; however, mean changes between week 1 and 7 tended to be larger for the IMT group.

Litchke et al. 2012; USA
N = 24 (22 SCI)

Population: N=24 males (22 with tetraplegia, 1 with spastic cerebral palsy, and 1 with congenital upper and lower limb deformities) randomly assigned to 1 of 3 groups: 1) inspiratory and expiratory resistive training (n=8); 2) inspiratory and expiratory threshold training (n=8); 3) controls (n=8). Age range: 17-35 yrs; DOI range: 6 months to 17 years.
Treatment: Resistive group trained with the Expand-a-Lung;1 set of 10 breathing cycles 3x per day for 9 weeks. Threshold group trained with the PowerLung Performer model; 3 sets of 10 breathing cycles 3 times per day every day for 9 weeks.
Outcome measures: SF-36v
  1. 16 participants completed the study (Threshold=4, Resistive=5, CON=7).
  2. Resistive RMT showed reductions in bodily pain and improvements in vitality domains of the SF36 versus CON values. The mechanism of decreased pain as a consequence of RMT is difficult to determine. However, due to the significance of pain on health related quality of life, this outcome is worthy of further consideration.

Ehrlich et al. 1999; Canada
Case Series
N = 1

Population: 26 yr old male,C3-C4
Treatment: Threshold IMT and Positive expiratory pressure value (Peripep) for one year.
Outcome measures: MIP, infection number.
  1. Number of respiratory infections decreased from 3 to 2.
  2. Number of respiratory infections requiring acute care hospitalization decreased from 2 to 0.
  3. MIP increased from 10 to 42 cmH2O.
  4. Daily suctioning 10x daily decreased to intermittent suctioning not required daily.

Uijl et al. 1999; The Netherlands
Prospective Controlled Trial
N = 10

Population: 10 participants recruited; 9 participants completed (8M 1F), all with tetraplegia C3-C7, 2-27yrs post-injury; AIS A (n=3), B (n=3), C and D (n=3); Age: mean 34.4 yrs (range 20-49 yrs)
Treatment: No resistive sham training (6 weeks) then Target flow IMT (6 weeks). 15 min twice daily for each phase of 6 wks.
Outcome measures: Spirometry, MIP, Maximal incremental threshold load (TLmax).
  1. TLmax, a measure of inspiratory muscle endurance increased after both sham training and IMT.
  2. No significant improvement in MIP for either group or differences in post-training change between groups.
  3. Significant increase in peak power, VT and oxygen consumption during maximal exercise test at 6-12 wks of IMT.

Rutchik et al. 1998; USA
N = 9

Population: N=9 people with SCI; C4-C7;  >1 yr since injury;  Age: 24-65 yrs with mean 36 yrs
Treatment: Resistive IMT without target 15 min twice daily x 8 wks.
Outcome measures: MIP, spirometry.
  1. Significant increase in MIP and lung volumes after IMT.
  2. At 6 months, 4 months after training stopped, trends towards baseline and repeat measures in 7 of 8 participants showed no difference between baseline and 6 months outcomes.
  3. Compliance ranged between 48 and 100% of IMT sessions.

Hornstein & Ledsome 1986; Canada
Case Series
N = 20

Population: N=20 participants (18M 2F) in acute post-traumatic phase; 10 tested at 4 months, 10 others were discharged, non-compliant or had medical complications.
Treatment: Resistive IMT without target 15min 2x/day x 6wks.
Outcome measures: MIP
  1. Four months after IMT began, 10 participants showed improvement in MIP from mean (SD) 45(1) mmHg to 59(6.8) mm Hg but no statistics were performed on data.
  2. Two case reports showed improvement in MIP and decreased dyspnea.

Gross et al. 1980; Canada
N = 6

Population: N=6 people with SCI (4M 2F); age range:18-41 yrs; >1 yr post injury
Treatment: Resistive IMT without target 30 min per day, 6 days per wk x 16 wks.
Outcome measures: MIP
  1. During training, progressive and significant increases in MIP and the critical mouth pressure that resulted in EMG signs of diaphragm fatigue.


Two RCTs provide level 1a evidence (Mueller et al. 2012, 2013; Van Houtte et al. 2008) that inspiratory resistive and normocapneic hypernea training, respectively, significantly increases inspiratory muscle strength. A case report provides level 5 evidence that threshold IMT improved inspiratory muscle strength and decreases the number of respiratory infections (Ehrlich et al. 1999). One RCT (Tamplin et al. 2013) provides level 1 evidence that group singing exercises 3 times weekly for 12 weeks significantly improves phonation, and projected speech intensity.

Earlier reports were not comparable and could not be combined in a meta-analysis (Brooks et al. 2005) because of research design, heterogeneity of subject characteristics or differences in training techniques. Several previous studies that used an RCT design incorporated with suboptimal IMT protocol. In particular, several used an inspiratory resistive device with no target to control for decreasing resistance with slower flows so the training methods may have induced an alteration in breathing pattern towards slower inspiratory flows rather than a training response against higher inspiratory pressures. The few studies that utilized a RCT design also showed improvement in both control (or sham) and training groups. Comparable improvement in measures of inspiratory muscle and lung function in the control and IMT groups may reflect learning of testing maneuvers, benefit from other rehabilitation or lifestyle activities, and/or natural progression of improvement after SCI.

The single subject report by Ehrlich et al. (1999), utilized the threshold trainer, which imposes a constant inspiratory load regardless of breathing pattern. Given that threshold IMT has consistently shown improvements in inspiratory muscle strength and endurance in people with chronic obstructive pulmonary disease, this technique warrants a larger RCT to determine its benefit for people after SCI.

The RCT by Van Houtte et al. (2008) and case report by Ehrlich et al. (1999) provides level 5 evidence that respiratory muscle training has the potential to dramatically reduce respiratory infections.

Future research to determine a potential treatment effect of IMT after SCI, should utilize: 1) larger samples; 2) a research design that controls for the influence of learning or recovery from SCI on IMT outcome measures of inspiratory muscle strength and endurance, and dyspnea; 3) optimal training techniques of threshold loading, targeted resistive devices, or normocapnic hyperpnea; 4) outcomes of inspiratory muscle strength and endurance; dyspnea; quality of life; daily function; 5) a comparison of the effectiveness of IMT relative to or as an adjunct to other rehabilitation interventions. Of equal importance, overly aggressive prescription of IMT has the potential to fatigue and injure the inspiratory muscles, which can increase the person’s predisposition to respiratory compromise. The article by Reid et al. (2004) provides a table that outlines parameters to monitor during IMT in order to avoid untoward responses such as muscle fatigue and hypercapnia. Parameters include: intensity of load, mode of load, duration, frequency and length of training to ensure adequate training protocol; blood pressure, heart rate, respiratory rate, other signs and symptoms of respiratory distress or inability to tolerate exercise load as signs of exercise intolerance; discoordinate chest wall movement, excessive dyspnea during training, long lasting complaints of fatigue after training sessions to avoid inspiratory muscle fatigue; signs of delayed-onset muscle soreness, reduced strength and endurance to avoid muscle injury; and end-tidal CO2, SpO2 and signs of headache, confusion to avoid hypercapnea (Reid et al. 2004). Van Houtte et al. (2008) provided 48 hours rest after their participants were unable to tolerate an overly intense workload.

For inspiratory muscle training to improve ventilation, decrease dyspnea, and to improve daily function after SCI, parameters to optimize IMT are only available for people with other respiratory conditions. For people with chronic obstructive pulmonary disease, the optimal IMT protocol should utilize threshold or targeted resistive trainers, at an intensity of 30-70% of MIP, for a duration up to 30 minutes per session, performed continuously or in intervals, 4-6 days/week and be continued indefinitely (Geddes et al. 2006). Progression of intensity (MIP) should not exceed 5% per week.


There is level 1b evidence based on 2 RCTs (Van Houtte et al. 2008; Postma et al. 2014), level 2 evidence based on 5 RCTs (Mueller et al. 2012, 2013; Loveridge et al. 1989; Liaw et al. 2000; Derrickson et al. 1992; West et al. 2014), level 3 evidence from 2 studies (Aslan et al. 2016 and Fischer et al. 2014), and level 4 evidence from several pre-post and case studies to support IMT as an intervention that will improve inspiratory muscle strength and might decrease dyspnea and respiratory infections in some people with SCI.

Related Downloads
Related Toolkits