Aerobic treadmill training effectively enhances cardiovascular fitness and gait function for older persons with chronic stroke

Blikk på forskning i Fysioterapeuten 3/2013


Blikk på forskning utarbeides i samarbeid med Journal of Physiotherapy (Australia), som trykker forskningspresentasjonene under betegnelsen Critically appraised Papers, CAPs.


Summary of: Globas C et al (2012) Chronic stroke survivors benefit from high-intensity aerobic treadmill exercise: a randomized controlled trial.Neurorehabil Neural Repair26: 85–95. [Prepared by Marco YC Pang, CAP Editor.]

Question: Does high-intensity aerobic treadmill exercise improve cardiovascular fitness and gait function in people with chronic stroke?

Design: Randomised, controlled trial.

Setting: An outpatient rehabilitation centre in Germany.

Participants: Individuals with chronic stroke > 60 years of age with residual gait impairment, and ability to walk on the treadmill at ≥ 0.3 km/h for three minutes were eligible. Serious cardiovascular conditions (eg, angina pectoris, heart failure, valvular dysfunction, peripheral arterial occlusive disease), dementia, aphasia, and major depression were exclusion criteria. Randomisation of 38 participants allocated 20 to the intervention group and 18 to the usual care group.

Interventions: The intervention group underwent treadmill training (three times/week) for three months. The program was intended to achieve 30–50 minutes of treadmill training at 60–80% of the maximum heart rate reserve as determined by a maximum effort exercise test. The training was supervised by a physician and/or physiotherapist. The usual care group received conventional care physiotherapy for one hour 1–3 times a week without any aerobic training.

Outcome measures: The primary outcomes were peak oxygen consumption rate and the six-minute walk test. Secondary outcome measures were self-selected and maximum walking speeds as measured in the 10-m walk test, Berg balance score, five-Chair-Rise test, Rivermead Mobility Index, and Medical Outcomes Study Short-Form 12 (SF-12). The outcomes were measured at baseline, immediately after completion of training, and at 12 months.

Results: 36 participants completed the study. After the three-month training period, the change in peak oxygen consumption rate was significantly more in the treatment group, by 6.3 mL/kg/min (95% CI 5.7 to 6.9). The change in distance achieved in the six-minute walk test was also significantly more in the treatment group by 53 metres (95% CI 32 to 75). Among the secondary outcomes, maximum walking speed (by 0.14 m/s, 95% CI 0.08 to 0.20), Berg balance score (by 2.6 points, 95% CI 0.5 to 4.7), and SF-12 Mental score (by 4.0 points, 95% CI 3.4 to 4.6) improved significantly more in the treadmill training group than the usual care group after the treatment period. The groups did not differ significantly on the remaining secondary outcomes. It was reported that compared to baseline peak oxygen consumption rate and six-minute walk test distance were significantly improved at 12 months.

Conclusion: A high-intensity treadmill training program improves cardiovascular fitness and gait in older adults with chronic stroke.
[95% CIs calculated by the CAP Editor.] 


Evidence is accumulating of the profound benefits conferred by aerobic training on cardiovascular function, mobility, brain health, and overall quality of life after stroke. However, when subjected to the rigors of systematic review, available data have failed to demonstrate superiority of such training over traditional therapies in optimising recovery post-stroke (Moseley et al 2005). The trial by Globas and colleagues contributes in important ways to elucidating the role fitness training plays in improving cardiovascular function and mobility after stroke. Level 2 evidence (ie, randomised controlled trial with < 100 subjects) is provided regarding the safety and effectiveness of a moderately intense training protocol for older individuals in the chronic post-stroke period (subjects were 5–10 years older than those in most previous trials). Considering the average age of stroke rehabilitation participants is > 70 years, use of a representative cohort speaks to the relevance of the study. Mean gain in exercise capacity of the training group (5.5 mL/kg/min or 1.6 metabolic equivalents, METS) is clinically meaningful – 1 MET improvement is associated with significantly fewer adverse events in people with coronary artery disease (Hambrecht et al 2004) and 12% increase in survival of men with cardiac disease (Myers et al 2002). Clinically meaningful change was also achieved in the six minute walk (ie, 49 m), but not comfortable walking speed (0.14 m/s) (Perera et al 2006) and Berg Balance Scale (5.8 points) (Stevenson 2001). The significant training-induced improvement in the SF-12 mental subscore is of interest, particularly given the recent links drawn between brain health and cardiovascular conditioning after stroke (Quaney et al 2009). That benefits were largely sustained at 12-month follow-up is encouraging. Use of a crossover design helped deal with the lack of dose equivalency in the intervention protocols (39 versus ~24 sessions in training and usual care groups, respectively) but unequal exposure precludes drawing conclusions about the ‘relative’ effectiveness of treadmill training. The troubling statement ‘current conventional care for chronic stroke survivors in Germany does not lead to improvements over 3 months’ is counter to findings reported elsewhere (Duncan et al 2003) and warrants further attention. We are reaching the stage where large multi-centred trials of aerobic training after stroke are necessary to answer definitively the central question of what attributes define ‘responders’ to this intervention. 

Marilyn MacKay-Lyons

School of Physiotherapy, Dalhousie University, Canada


Duncan P et al (2003) Stroke 34: 2173–2180.

Hambrecht R et al (2004) Circulation 109: 1371–1378.

Moseley AM et al (2005) Cochrane Database of Systematic

Reviews Issue 4. Art. No.: CD002840.

Myers J et al (2002) New Engl J Med 346: 793–801.

Perera S et al (2006) J Am Geriatr Soc 54: 743–749.

Quaney B et al (2009) Neurorehab Neural Repair 23: 879–885.

Stevenson TJ (2001) Aust J Physiother 47: 29–38.