The patient followed an exercise programme including strength, endurance and flexibility training, as well as neuromuscular activities [14] . The duration of the programme was 2 months with 3 hourly sessions per week. Each session started with warming-up for 5 to 10 minutes, with low-intensity aerobic exercises, and stretching exercises. Then the main part was taking place, consisting of : 1) neuromuscular coordination exercises twice a week for 20 minutes; 2) strength exercises 2 days per week; 3) moderate-intensity aerobic exercise 3 times per week for about 30 minutes; and 4) flexibility exercises twice a week. Finally, cooling down was performed with low-intensity aerobic exercises and stretching exercises for five minutes. Both the intensity and the main principles of exercise were determined according to the guidelines of the American College of Sports Medicine (ACSM) [18] .

Patient’s gait mechanics were evaluated before and after the intervention using a three-dimensional gait analysis system, with six infrared cameras, two force plates, and an electronic timing system. Detailed description of the equipment and measurement process is presented elsewhere [19] .

Table 2 presents the means of the spatiotemporal gait parameters for both lower limbs, with and without paresis, according to the measurements before and after the intervention, as well as the % difference between the initial and final measurement. Increase of the spatial parameters (stride and step length) and decrease of the temporal parameters (stride and step duration, single and double support duration) were observed on both lower limbs. Gait velocity increased by 328%.

Figure 1 and Figure 2 present kinematic and kinetic parameters of the 5 gait cycles before and after the intervention. Before the intervention, severe impairment of the gait pattern was observed. After the completion of the exercise programme, the waveforms of all parameters approached the typical pattern. As regards kinematic parameters increased range of motion (ROM) was observed in the joints of the hips, knees and ankles in the anterior-posterior plane in both lower limbs. In addition, faster initiation of triple joint flexion in the lower limbs (hip-knee-ankle dorsiflexion) was observed during the swing phase (immediately after 60% of gait cycle). These changes were correlated with remarkable increase in muscle powers in the lower limbs from the three joints at the anterior-posterior plane throughout the gait cycle. Moreover, improved waveforms of the moments were observed, mainly during the support phase. Improved waveforms were also observed in the frontal and transverse planes, in all the parameters studied.

The parameter showing the most pronounced change was gait velocity. Gait velocity plays a crucial role in human functionality. For a pedestrian to move around in the city, to cross the road, for example, gait speed must be from 1.1 to 1.5 m per sec [20] [21] . The patient’s final gait velocity was approximately 0.5 m per sec, which is still much lower. Nevertheless, the initial value of her gait was only 0.12 m per sec, and her improvement ensured her autonomy for domestic activities. Indeed, the modified Rankin Scale after the completion of the intervention decreased to 3. Moreover, the above changes were achieved after only eight weeks of intervention. An extension of the intervention programme could lead to further increases in gait velocity.

Moreover, temporal gait parameters improved, consistent with the observations of Onley and Richards [22] on the spatiotemporal gait parameters in stroke patients, which are attributed to muscle weakness. Therefore, the improved spatiotemporal gait parameters suggest her muscle strength improvement.

Another improvement was the increase of the spatial gait parameters, attributable to improved balance. Indeed, Espy et al. [23] observed that the elderly adopt a gait pattern characterised by slow velocity and small step length to allow better balance, but this slow gait velocity actually facilitates imbalance, which is simply compensated by smaller step length. Moreover, Teixeira-Salmela et al. reported that the reduced gait velocity, reduced length of stride and increased duration of double support are mainly related to the fear of falling and not the actual risk of falling [24] .

The increase of joint ROM at the anterior-posterior plane could be attributed to the reduction of spasticity of the extensor muscles and/or to the improvement of flexor muscle strength. An alternative explanation is improved balance. The latter may have rendered the patient more confident, without the need to keep the lower limb close to the floor through an incomplete triple flexion during the swing phase, so as to react promptly in case of imbalance. Thus, released from the fear of falling, she could achieve more effective triple flexion [22] .

The patient achieved increased moments during the support phase, probably due to her increased strength [22] . Increased gait velocity explains the power waveform improvement in all 3 joints bilaterally [25] .

In conclusion, the exercise programme that was based on AHA/ASA guidelines had a positive effect on this patient’s gait pattern. There was an increase in spatial and decrease in temporal gait parameters, increase in joint ROM and lower limb muscle powers during the entire gait cycle and increase in the moments during the support phase. These findings are indicative of the effectiveness of the above exercise programme on the gait pattern of patients with chronic stroke. A study on the effects of exercise in more patients with different levels of disability and during different rehabilitation phases will allow drawing safe conclusion and generalise the findings.