1.1

Introduction

Exercise capacity, and specifically aerobic capacity, is commonly diminished among patients with central nervous system lesions, as a consequence of motor impairments, functional limitations, rest and immobility at the early phase of the disease. This is particularly true after cerebrovascular stroke because the functional limitation consequences are associated to the frequent cardiovascular diseases, frequently responsible of the stroke. Motor control, balance and muscle strength impairment are considered as the most limiting factors of walking recovery after cerebrovascular stroke . Although at the acute phase of stroke rehabilitation, muscle weakness and loss of coordination are considered as the primary impairments that affect gait recovery, at the subacute or chronic phase, impaired cardiorespiratory fitness is reported to affect gait performance by limiting walking endurance . Altered aerobic capacity of hemiplegic stroke patients, in comparison with healthy subjects of the same age, limits the ability to prolonged effort . So, exercise or training programs have been designed in order to improve functional abilities in chronic diseases, taking into account the decreased capacity for energy production. There is a growing body of literature on exercise training in stroke victims, but few studies have reported a relationship between walking performance and aerobic performance for individuals with hemiparetic stroke . The essential is reported at chronic phase of rehabilitation with a wide variety of objectives and procedures for training . It is reported walking training procedure with treadmill exercise with or without body weight support (BWS) or with a specific gait trainer ; aerobic training on cycloergometer , muscle strengthening exercises or electrical stimulation training and also exercises mimicking daily activities . Most studies reported an improvement of walking characteristics (balance, rhythm, speed…) with treadmill training with or without BWS , but for Moseley et al. there is not a real benefit for the speed of walking and for Macko et al. , there is a better effect of BWS at the acute phase.

Four studies specifically reported an increase in aerobic capacity after cycloergometer training program after stroke . The study of Bateman et al. does not include only stroke victims but also traumatic brain injury, subarachnoid haemorrhage . In these studies, the aerobic training associated an increase of workload and Potempa et al. and Tang et al. reported an increase of VO _2max . Concerning the walking performance, Katz-Leurer et al. showed an increase on stair climbing .

In a meta analysis, Pang et al. showed that cycloergometer was the most common method used for aerobic training (reported in four studies), and the exercise protocols described in the review studies are similar to the guidelines recommended by American College of Sports Medicine (ACSM) in terms of improving peak VO ₂ (55–90% of maximal heart rate or 40 to 85% of heart rate reserve; 3 to 5 days, 20 to 30 mins per session) . All reviewed studies used a protocol consisting of continuous exercise from 20 to 40 mins. All of theses studies reported statistically significant increase in peak VO ₂ in the experimental group after aerobic exercise training ranging from 9.0 to 34.8% and in peak workload from 7.3 to 176.9%. Only one study reported a training phase during early stage of rehabilitation . It is thus, the unique study witch compares an exercise group with a control group, in the early stage of stroke rehabilitation (less than 3 months). This recent study of Tang et al. reported an increase of gait speed and functional ambulation measured by 6-mins walking test (6MWT) . The variation of time spent during rehabilitation activities could be considered as a limit in this study, particularly because it is performed during the early rehabilitation stage period.

Finally, to our knowledge, improvement of VO _2max depended on training modality. For low conditioned subjects, interval-training induced a greater improvement in VO _2max than continuous training , but no study reviewed used this type of program in a population of stroke victims. Most of the effects of continuous training appeared when higher exercise intensity (75–85% of maximal HR) and/or longer training duration (3–6 months) were used . To program an interval-training activity could generate a lower training duration, and thus a lower time of hospitalization.

In summary, even though there is a general agreement regarding the overall benefits of exercise after stroke, wide differences in patient populations, in training (intensity, frequency, and in type of activity (walking, cycling, stimulation), made it difficult to ascertain the respective impact on cardiovascular capacity, muscle strength, and motor control, and finally on walking performance . However, only, Macko et al. report after treadmill training an effect on cardiovascular capacity and walking performances .

In this context, the primary objective of our study is to determine the feasibility of cycloergometer interval-training in stroke patients, and considering the relationship between walking performance (6MWT) and aerobic training ( VO _2max ) to determine the transfer effects of this training on walking performances in patients with hemiparetic gait following stroke at subacute and chronic phase.

1.2

Methods

1.3

Results

1.4

Discussion

There is a large variety of training procedure for stroke developed in different studies, including cycle ergometer, treadmill or functional activities such as brisk stepping . Exercising on a cycloergometer does not require as much postural control when compared to treadmill walking and thus may be a better alternative for those individuals with poor balance . Upon review of literature regarding training programs for hemiplegic stroke patients, the principal aim is to improve walking capacity, analyzing speed, symmetry, balance, endurance, as performance outcomes. Maximal exercise or aerobic capacity is seldom measured during these programs; therefore, it is difficult to draw valid comparisons. Only Potempa et al. and more recently, Tang et al. measured VO _2max parameters in the case of an aerobic cycloergometer training.

Concerning the effect of the aerobic cycloergometer training, we observe in our study, a significant improvement of the maximal oxygen consumption ( VO _{2 peak} ) and the maximal workload (Pmax) in a population of stable subacute or chronic hemiplegic patients ambulating independently, and who exhibit a baseline maximal capacity below that observed in healthy populations of the same age range. Cycloergometer training program improved VO _{2 peak} from 18.5 ± 3.7 to 21.3 ± 6.1 ml/kg per minute and Pmax from 80 ± 25.4 to 98.6 ± 24.4 W. These results are in agreement with those published by Potempa et al. , who found that exercise training significantly increased mean maximal VO ₂ ( P < 0.0001) (from 16.6 ml/kg per min at baseline to 18.8 ml/kg per minute after training) and workload ( P < 0.0001) (from 65.3 W to 94.2 W). Potempa et al.’s study was conducted in 19 subjects aged from 43 to 72 years, who had completed a stroke rehabilitation program for an average duration of 216 ± 40.3 days after stroke . Tang et al. reported also significant increase of VO ₂ ( P < 0.0001) (from 11.6 ± 0.7 ml/kg per min at baseline to 13.1 ± 0.9 ml/kg per minute after training) and workload ( P < 0.0001) (from 46.4 ± 4.3 W to 57.2 ± 6.4 W) . Tang et al. conducted the study with 18 patients with a mean age of 64.5 ± 3.6, mean post-stroke period of 19.2 ± 3.8 days, and FIM ^TM score mean of 84 ± 4.1 . Their results reported a lower increase than our results or them of Potempa et al., probably because the training program was proposed for all subjects at an acute or subacute period after stroke, and with a lower mean of VO ₂ or workload at baseline. Our results are also in agreement with those of Katz-Leurer et al., which reported a significant improvement of heart rate at rest ( P = 0.02), workload and work time ( P < 0.01) in 46 subjects aged 62 ± 11 years after a cycloergometer training of 8 weeks duration at the subacute stage after a cerebrovascular stroke . Bateman et al. also reported a benefit of exercise with different methodology and a mixed population . We can conclude positively for the feasibility of a cycloergometer interval-training program in stroke victim and for the increase in aerobic performance with a lower time duration than literature.

Concerning the walking performance as measured, we observed a significant increase of the 6MWT (15.87%) more demonstrative than the results reported by Katz-Leurer et al. with a trend for improvement in all functional parameters (walking distance, walking speed, stair climbing, and FIM™ score) in their treated group, and only significance for stair climbing ( P < 0.01) . Our results are lower than them of Tang et al. who reported an increase of 52.7 ± 19.7% for the exercise group but with a necessary moderation due to the increase of 22.7 ± 5.6% for the control group in relation with the spontaneous recuperation or the classical rehabilitation effects, and with the necessity to take into account that their program was conducted at the early phase of rehabilitation .

Very few studies analyzed the relationships between aerobic capacity and walking performance in stroke population, and the authors reported contradictory findings . Kelly et al. find a strong correlation ( r = 0.84) between the distance walked in 6 mins and the measure of peak cardiorespiratory fitness . The study was conducted in a sample of 17 subjects, between 4 and 6 weeks after stroke. The strength of the correlation may be due to the fact that none of Kelly et al.’s patients were taking betablockers. Courbon et al. report similar results with a weaker correlation ( r = 0.6), but some of the subjects are on betablockers . Eng et al. don’t find any correlation between VO _2max and the 6MWT, and Pang et al. reported only a low correlation ( r = 0.402) . Both authors concluded that measuring gait function via distance tests may not necessarily related to cardiovascular fitness, because other stroke-related impairments have a critical impact on walking function (strength, balance, coordination…). In our opinion, even if we acknowledge the importance of motor impairment on walking performance, their findings can be influenced by the fact that their subjects walked at a relatively high speed. In our study, we find also a significant correlation between the 6MWT and VO _{2 peak} before training, but not after. This lack of correlation between VO _{2 peak} and 6MWT improvements could be due to the fact that some stroke patients were taking betablockers, which limit VO _{2 peak} increasing.

Other possibilities to explain different results could be the type of training and the time between training and stroke. Macko et al. observed after a 6-month treadmill training in a group of 25 hemiplegic stroke patients more than 6 month post-CVA (mean age: 63 years), a 17% increase in VO _{2 peak} ( P = 0.001) . Fitness gains are progressive and evenly distributed across the 6-month period. These authors also found that the 6-min walk distance increased by 30%, with most of the improvement (77%) occurring within the first 3 months. We observe a lower improvement in 6-minute walk distance (15.87%), although our subjects have a similar average baseline walking performance to those of Macko et al. (231.36 m versus 232.01 m) . This can be due to the difference of training: treadmill versus cycloergometer. On the other hand, we find a comparable improvement in VO _{2 peak} (14.84% versus 17%) with a cycloergometer training program with a duration of only 2 months. However, the time elapsed between the stroke and inclusion into the training session is markedly shorter in our population (12.69 ± 7.53 months versus 35 ± 29 months). Looking back at our results, we observe a better improvement of maximal power for subjects with a short delay (< 1 year) than the other group of subject who has a longer delay (> 1 year). We suggest that fitness improvement can be in relation with the delay between stroke and training inclusion. On the contrary, the results of Tang et al. conducted at a mean post-stroke period of 19.2 ± 3.8 days show lower increase of VO ₂ or workload and higher increase of walking capacities .

According to Macko et al., cardiovascular adaptation after stroke may be contingent on improving training velocity rather than heart rate reserve . Our results suggest that walking improvement may be contingent essentially to Pmax improvement. This is an accord with the dynamic muscle strength activity evaluated by isokinetic muscle peak torque which increases on both sides at 60°/sec and 120°/sec, without isometric muscle strength increase.

In comparison with the study of Tang et al., the lack of a control group and the small sample size are some limit interpretations of our results . In particular, our patient sample is characterized by a wide variation in the severity of residual impairments, but we are unable to compare the outcome of training between levels of motor impairment, with the data of this study, but also with the data of other studies published. Many factors may contribute to the variation of the results of training after stroke: determinants of walking capacity (as motor recovery, balance, muscle strength, spasticity, motivation…) and of exercise capacity (as hypertension, cardiac arrhythmias, inactivity or learned disuse, sedentary versus active way of life, age, treatment with betablockers…) which can explain some difference to training responses . Nevertheless, such heterogeneous study populations are commonly encountered in the literature relative to stroke outcomes, and reproduce the variability encountered in routine clinical practice. The absence of a control group is also a limiting factor, although for the most part our results are consistent with those of other authors. But, because only an interval-training session is proposed in our study, we contribute particularly to determine specifically the cardiovascular training effects, without any other factors as frequently reported in other studies as muscle strength, walking or daily activities during the training procedure.

In a Cochrane Review, recently, Saunders et al. have demonstrated through literature that there is some evidence that cardiorespiratory fitness can be improved via exercise containing some cardiorespiratory training contents. And finally, there is good evidence that cardiorespiratory training during usual care, which involves walking as a mode of exercise, can reduce dependence on others during ambulation and improve walking performance in terms of speed.

1.5

Conclusion

An 8-week aerobic cycloergometer interval-training program is feasible and effective on cardiovascular, workload, and walking performances for hemiplegic patients. Although these results must be interpreted with caution, it is interesting to show that improvement of walking parameters can be obtained by another means that is treadmill training. The results obtained in our small sample need to be confirmed on larger populations, in a controlled study and prospective study with the comparison of different training modalities as functional or walking aerobic or muscle strength training exercises.

Conflicts of interest statement

No conflict of interest.