Effect of Body Weight–Supported Treadmill Training on Cardiovascular and Pulmonary Function in People With Spinal Cord Injury: A Systematic Review

A systematic literature search was conducted to examine the effect of BWSTT on measures of cardiovascular and pulmonary health in people with SCI. MEDLINE (PubMed), the Cumulative Index to Nursing and Allied Health Literature (CINAHL), and the Physiotherapy Evidence Database (PEDro) were searched. The search was limited to clinical trial studies in humans, with adult patients (≥18 years), and written in English. It was conducted for the time period of January 2004 until May 2018.

The search strategy included the terms for target population (spinal cord injury, paraplegia, quadriplegia or tetraplegia), outcome measures (heart rate, blood pressure, pulmonary function, respiratory function, and respiratory parameters), and intervention (locomotor training, walking training, gait training, or body weight support treadmill training). Different combinations of the terms were made using “AND” and “OR” in order to achieve a specific selection of literature (see Table 1).

Inclusion criteria were the following:

1. Studies with adult patients (≥18 years) with traumatic or nontraumatic SCI (cervical, thoracic, and lumbar), complete or incomplete lesions, and American Spinal Injury Association Impairment Scale (AIS) of A, B, C, or D.

2. Treadmill walking training, including manual-assisted BWSTT, robotic-assisted BWSTT, and treadmill walking training in water.

3. Studies focused on measurements of HR and BP and respiratory parameters after a course of treadmill walking training.

4. Clinical trial studies, including randomized controlled (RCTs), quasi-experimental, or pre-experimental trials.

5. Published in English.

Exclusion criteria were (a) overground walking training and (b) animal studies or studies on children.

After removing duplicated studies, two independent researchers (R.A and A.A) screened the articles by reading titles and abstracts. Then, the full text of articles relevant to study objectives was read in detail to determine eligibility (see Figure 1). Disagreements on study selection were resolved by discussions between two researchers (R.A. and A.A)

The study information was extracted by two reviewers (R.A. and A.A.) independently. The following data were extracted from included studies: author's name, year, study design, sample size, participants' characteristics (gender, age, injury level on the SCI, grade based on AIS, and time since injury), intervention program (length and frequency), and outcomes related to measurements of HR, BP, and respiratory parameters. There were no disagreements between the reviewers regarding data extraction.

Included studies were assessed using the modified Downs and Black scale to determine their methodological quality. The Downs and Black scale is composed of 27 questions to assess the quality of a study across five categories, including reporting (10 questions), external validity (3 questions), internal validity (bias and confounding; 13 questions), and power (one question) (see  Appendix).43 The total score of the scale ranges from 0 to 28. A higher score indicates a better quality of methodology. From the percentage of derived total score, studies were classified as high (>75%), moderate (50%–74%), or low quality (<50%). The level of evidence of each study was identified using the Spinal Cord Injury Rehabilitation Evidence (SCIRE) system, a five-level system that differentiates between studies of differing quality and incorporates the types of research designs commonly used in rehabilitation research (see Table 2).13 This scale has been used in published systematic review studies of exercise training in people with SCI.44,45 Two researchers evaluated and scored studies independently. In case of disagreement between two evaluators, a consensus was made through discussion.

An overview of the results of the literature search and screening process is provided in Figure 1. The electronic database search retrieved 264 articles. Removal of duplicates within and between the individual databases left 79 articles for further examination. Of the 79 retrieved articles, 53 articles were excluded after screening for titles and abstracts. The remaining 26 articles were screened by reading the full text to determine eligibility. Specific reasons for article exclusion are presented in Figure 1. Nine articles were evaluated in detail and are included in this review study.

A summary of the study designs, participants' characteristics, interventions, and outcome measures for each of the nine studies reviewed is provided in Table 3. The number of participants in the included studies ranged from 6 to 52 participants; a total of 121 participants took part in those studies. They included individuals with incomplete SCI, complete SCI, or both. The duration of interventions ranged from 2 to 5 times per week for 4 weeks to 6 months. Seven studies examined the effects of walking training on HR measurements.34–40 Four studies examined the effects of walking training on BP measurements.34–36,40 Three studies examined the effects of walking training on respiratory parameters.36,41,42 

The results of the quality review are presented in Table 4. Overall, the quality index of all included studies was low. All studies scored less than 21 out of 28 on the Downs and Black scale. All studies fulfilled most of the criteria for reporting. The objectives of the study, the main outcomes, the characteristics of the patients, the interventions of interest, and the main findings were clearly described in these studies.34–42 However, all studies34–37,39–42 except one38 were rated poorly on the measurements of the internal validity (bias and confounding) because of lack of control group and randomization. Two studies were scored poorly on the measurement of the external validity because only men were included, and this might have an influence on the generalizability of findings.38,40 All studies did not report estimation of sample size; thus, they were rated poorly on the measurements of the power.34–42 

The level of evidence varied from level 2 to level 4 (see Table 3). Two studies included a control group.38,42 One of the studies was a randomized crossover study design with small sample size and was classified as level 2.38 The other study was a prospective study with a control group (ie, quasi-experimental study design) and was also categorized as level 2.42 The remaining seven studies used a single group, pre- and post-test study design and were placed at level 4.34–37,39–41 The sample sizes of all studies34–41 were small (range, 6–12) except one study42 that included 52 participants.

To the best of our knowledge, this review is the first to systematically synthesize the evidence regarding the effects of BWSTT program on measurements of HR, BP, and respiratory parameters in people with SCI. An exhaustive search found nine studies that met inclusion/exclusion criteria. Given the heterogeneous nature of the SCI population and low research quality, there is weak to moderate evidence to support the effects of BWSTT on improving HR measures and respiratory parameters; however, the evidence regarding BP measures is still lacking or needed.

Changes in HR were measured pre- and post-BWSTT in seven single group studies.34–40 One study40 included only individuals with motor complete cervical SCI (ie, quadriplegia) and found no changes in resting HR after BWSTT (two sessions per week for 3 months). On the contrary, another study34 that included both motor incomplete and complete cervical SCI reported a significant decrease in resting HR after 6 months of BWSTT with three sessions per week. Two studies with mixed SCI lesion levels (ie, quadriplegia and paraplegia)36,37 also reported a significant decrease in resting HR after 6 weeks or 24 sessions over 10 to 16 weeks of BWSTT. Two other studies with mixed SCI lesion levels35,38 reported a nonsignificant decrease in resting HR after 4 months or 4 weeks of BWSTT. Exercise HR was measured in two studies37,39 with incomplete and mixed SCI lesion levels and was significantly decreased after 24 sessions of BWSTT.

The intensity and duration of BWSTT may influence HR adaptation. One study38 found no significant decrease in resting HR, but the duration of this study was very short (4 weeks of BWSTT) as compared to the other studies34,36,37 (8 weeks to 6 months of BWSTT). Authors of this study did not provide information about the intensity of training. Also, it might be challenging to find a significant change in HR after 4 weeks of exercise. It has been shown that at least 6 to 8 weeks of walking training would be ideal for producing HR adaptation.36,37 Furthermore, moderate to high intensity of arm aerobic exercise for 8 weeks has been shown to improve cardiopulmonary fitness in individuals with SCI.46 Future studies in walking training should focus on the impact of the training intensity on HR outcomes.

Completeness of injury (incomplete or complete SCI) might also have an influence on HR response to walking training. In two studies35,40 that included only individuals with motor complete injuries, BWSTT did not produce a significant decrease in resting HR. In Ditor's study,35 only participants who had HR response (ie, increased HR) during walking training showed a decrease in resting HR. It seems that HR response and voluntary contraction of leg muscles during BWSTT are essential factors for inducing HR adaptation.35 A repetitive and intensive locomotor training has been shown to improve the activity of leg muscles even in persons with motor complete SCI.47,48 It is still unknown how significantly increased activity of leg muscles after a course of walking training can contribute to HR adaptation.

Compared to arm cycling exercise, walking training can trigger the activity of larger muscles in the body (ie, leg and trunk muscles) through activation of central pattern generators located within the spinal cord.49–51 In addition, being in upright posture during walking training can provide greater stress and challenge to the cardiovascular system. It has been shown that in individuals with SCI, HR and oxygen uptake were significantly higher during BWSTT in comparison to exercise in a sitting position.32,33 Consequently, walking exercise might induce a greater positive adaptation in HR than conventional exercise approaches such as cycling or FES leg cycling exercise.52 However, the efficacy of walking training compared to other exercise approaches has not been studied.

In relation to cardiac autonomic function, one study34 showed an improvement in HRV, as reflected by a significant reduction in the ratio of low-frequency power to high-frequency power, in individuals with incomplete cervical SCI after 6 months of BWSTT. Two studies35,38 reported a trend toward an improvement in HRV after 4 weeks or 4 months of BWSTT. The duration of one of these studies that found nonsignificant improvement in HRV after training was very short (ie, 4 weeks).38 It might be difficult to find a significant improvement in HRV within a short term of training. However, this study has noted a significant improvement in HR complexity after training, which is an indirect measurement of cardiac autonomic function.38 The other study included only individuals with motor complete SCI, and only participants who showed HR response (ie, increased HR during training) to BWSTT improved their HRV after 4 months of training.35 

The findings of these studies34,36,37,39 suggest that BWSTT might result in decreased resting and exercise HR and improved cardiac autonomic function in individuals with SCI. These studies that support the use of BWSTT in improving HR measurements had a single group, pre- and post-test study design. Therefore, level 4 evidence is given for the use of BWSTT to improve HR measurements following SCI.

Four studies34–36,40 investigated the effects of BWSTT on changes in BP measurements in individuals with SCI. Two studies35,36 included individuals with cervical, thoracic, or lumbar SCI, and the other two studies34,40 targeted only individuals with cervical SCI. In the studies with mixed SCI levels, no significant changes in resting BP were observed in individuals with motor incomplete SCI after 6 weeks of BWSTT36 or in motor complete SCI after 4 months of BWSTT.35 In the studies of cervical SCI, one study observed no significant change in resting BP in motor incomplete SCI after 6 months of BWSTT.34 Another study with motor complete cervical SCI reported a significant increase in resting BP after 3 months of BWSTT combined with neuromuscular electrical stimulation.40 The authors of this study suggested that the increase in resting BP might be due to improvement in sympathetic activity after the BWSTT.40 

It is well known that regular exercise at moderate/high intensity can lead to a decrease in resting BP.53,54 Past studies indicated that regular exercise might decrease resting BP in people with paraplegia (SCI at T1 or below) for whom the prevalence rate of hypertension is high.6,55 However, people with cervical SCI (quadriplegia) tend to have abnormally lower resting BP compared to able-bodied individuals and those with thoracic or lumbar SCI (paraplegia) due to impaired cardiovascular autonomic function after SCI.16–18 Additional reduction in BP after a period of exercise in people with quadriplegia might provoke symptoms of hypotension.40 A physical exercise program of FES leg cycling for several months increased resting BP in people with quadriplegia.56 The use of BWSTT in people with quadriplegia may or may not lead to an increase in resting BP, as shown in two reviewed studies.34,40 Studies that included participants with quadriplegia and paraplegia35,36 were therefore inconclusive in terms of changes in BP due to the fact that resting BP may change in opposite directions in these populations.

As discussed previously, the four studies34–36,40 that reported outcomes in resting BP after BWSTT were inconclusive due to their heterogeneity in terms of study participants and the response of the participants. In addition, all four studies had small sample sizes, ranging from 6 to 12 participants. This limited number of studies, small sample sizes, different levels of SCI, and variable responses to walking exercise allow no conclusions to be drawn regarding the effects of BWSTT on resting BP. Future clinical trials need to include large sample sizes and focus on either paraplegia or quadriplegia.

Two studies34,35 also measured BP variability pre- and post-BWSTT training. One study34 reported a significant decrease in low-frequency systolic BP (SBP) in individuals with incomplete cervical SCI after 6 months of BWSTT. The other study from the same investigation team found no significant changes in measurements of BP variability in individuals with motor complete cervical or thoracic SCI after 4 months of BWSTT.35 No conclusion can be made at present.

Only three out of nine studies that were included in this review measured pulmonary function as an outcome.36,41,42 In all three studies, there were improvements in some of the respiratory parameters after a course of walking training. Soyupek et al36 found significant increases in forced vital capacity (FVC) and inspiratory capacity (IC) in individuals with motor incomplete SCI after 6 weeks of BWSTT. Another prospective study with a control group showed that vital capacity (VC), VC%, FVC, forced expiratory volume in one second (FEV1), and forced expiratory flow rate 25% to 75% were significantly increased in participants who received 4 weeks of BWSTT (the experimental group) but not in those who received standard rehabilitation program (the control group).42 Both groups showed a significant increase in maximum voluntary ventilation.42 However, this study did not use a random procedure in assigning the participants into one of two groups.

In addition to measuring pulmonary function, a study explored the underlying mechanisms of improvement in pulmonary function after walking training.41 They reported significant increases in FVC, FEV1, and maximum expiratory pressure after 62±10 sessions of BWSTT. Also, the amplitude and motor unit recruitment of all respiratory muscles during respiratory tasks significantly increased after training as compared to baseline measures. The findings of the study suggest that BWSTT could induce neuroplasticity in spinal neural circuitry that is responsible for the activation of respiratory muscles.

In comparison to exercises in sitting position, upright posture during walking training can trigger the activity of trunk muscles including abdominal muscles, which play an important role in respiration.49–51 It is also a highly effective stressor of the cardiopulmonary system since the lungs need to work harder to deliver more oxygen to larger working muscles.57 Furthermore, it has been shown that walking training increases the connectivity of neural spinal circuity between motor cortex and leg muscles.58 The increased muscle activity of inspiratory and expiratory muscles after BWSTT may be explained by neuroplastic changes within the neural spinal circuity that controls respiration.41 All of those factors might induce greater adaptive changes in pulmonary function. Based on this information mentioned, the use of BWSTT to improve pulmonary function after SCI is supported by one study42 with level 2 evidence.

To maintain cardiovascular and pulmonary health after SCI, it is crucial for people with SCI to engage in regular physical activity. There is some evidence that the use of BWSTT as an exercise in individuals with SCI has positive effects on cardiovascular and pulmonary health by improving resting and exercise HR and respiratory parameters. No clear evidence is currently available about which type of physical rehabilitation produces the best results. In addition, it is not clear whether BWSTT leads to a better outcome in cardiovascular and pulmonary function compared to conventional physical rehabilitation methods, such as arm cycling or FES leg cycling exercise. Furthermore, current studies have not provided information about the effects of the intensity level of walking training on these outcome measurements. Because of limited studies, further investigations are necessary. Future randomized controlled trial studies are needed to investigate the effects of BWSTT on cardiovascular and pulmonary health compared to other physical rehabilitation interventions. Further studies also should investigate the influence of walking training intensity (ie, walking speed or amount of body weight support) on cardiovascular or pulmonary health.

The authors declare no conflicts of interest.