Convenience sampling was used to select participants with DS from a special education school. The control group (CG) was recruited from a university in Barcelona, Spain. Twelve young adults with DS and 12 young adults without DS agreed to participate. Both groups were aged between 17 and 22 years. The inclusion criteria for the DS group were a level of ID between 30% and 59%, which implies an Intellectual Quotient percentile (IQ) of 33–70%. The ID classification was obtained from patient medical files and represents a combination of level of intelligence and adaptive behaviour. The National Government classifies the percentage of disability (physical, intellectual and/or sensory) in 5 degrees, as follows: non-existent (0%), low (1–29%), moderate (30–59%), severe (60–74%), and very severe (>75%). Classification is based on items such as the ability to adapt to different daily life environments (professional, cultural, family, or social) and the (IQ), among others.38 According to the Government, our participants had moderate ID.

The exclusion criteria for the DS group were: mobility problems, standing difficulty, vestibular or neuromuscular disease, and any additional psychiatric diagnoses requiring drug therapy. Identical criteria were applied to the CG group, except those related to intellectual impairment and disability.

Initially, 12 volunteers with DS and 12 without DS enrolled in the study, but one volunteer with DS dropped out due to surgery, and one volunteer from the CG moved out of the area.

All participants and their parents/guardians provided informed consent to participate in the study. A medical examination was performed prior to the initial study assessments in order to rule out possible contraindications to PA.

The study protocol followed the criteria of the Declaration of Helsinki39 and was approved by the Ethics Committee.

Evaluations were performed at the beginning (pre) and at the end (post) of the 18-week PA programme.

Participants in both groups followed an 18-week PA programme (two 90min sessions per week) based on classical, modern, and creative dance. Volunteers in the CG performed these activities in a college course during the same timeframe that school students with DS devoted to physical education and art. Sessions were led by two physiotherapists with special experience in dance and body expression and were conducted in classrooms especially designed for dance and PA. Both classrooms had the same materials: elastic bands, exercise balls, rings, mats, mirrors, and a ballet barre.

The same programme was performed by both groups. It consisted of warm-up activities (5–10min), abdominal strengthening, lumbar and ballet barre exercises (15min), proprioception and balance exercises with opened eyes and closed eyes (20min), choreography (20min), improvisation exercises and image recognition in a mirror (15min), and relaxation (5–10min). The DS group received greater explanation and demonstration of the exercises during the first two weeks to ensure understanding. Adherence to the programme by both groups was over 90%.

Table 1 presents the characteristics of the participants. Eleven individuals in the DS group (4 men and 7 women) and in the CG (6 men and 5 women). According our assessment of the classification of the degree of disability in the DS group, one person has a low degree and ten people have a moderate degree.

The only significant differences were in the average values for height (p<.01) and BMI (p=.02), due to the morphological differences associated with DS. Height values were higher in CG. BMI was higher in DS group.

Significant differences between the CG and DS group were observed in the soleus (area and mean amplitude p<.01) of the left leg and lateral gastrocnemius (area and mean amplitude p<.01) of the right leg (Table 3).

In the CG, significant differences between the opened eyes and closed eyes conditions were observed in the soleus (mean amplitude p=.04) of the left leg and the medial gastrocnemius (area p=.02) of the right leg, with an increase in EMG activity in the closed eyes condition. In participants with DS, significant differences between the opened eyes and closed eyes conditions were observed in the tibialis anterior (area and average p=.03) of the left leg, with a decrease in EMG activity observed with the closed eyes (Table 5).

In the CG, significant differences between pre- and post-training situations were observed in the tibialis anterior (area and mean amplitude p=.02) and soleus (area and mean amplitude p=.03) of the left leg with an increase in EMG activity in the tibialis anterior and a decrease in the soleus. In participants with DS, no significant differences between pre- and post-training situations were observed (Table 6).

When comparing groups in the opened eyes condition, we noted that young people with DS exhibited more muscle activity, although the differences were significant only in the tibialis anterior, soleus and lateral gastrocnemius. In the closed eyes condition, these differences were also observed. In our previous studies, we observed a large displacement of the COP in the group with DS41 which makes us think that although there is a muscle overexertion stabilization effect, it is potentially ineffective at reducing COP displacement. Other authors have analyzed muscle activity by EMG in people with DS, although many have studied the evolution of gait from childhood to adolescence or adulthood.16,21,42 When analysing standing position, authors analysing have observed increased activation and co-activation of agonist-antagonist muscle groups in people with DS,4,19,42,43 which coincides with our results. Postural adjustments and pre-adjustments during the execution of exercises can be altered. In children (1–6 years), Shumway-Cook18 observed latencies in muscle responses at the ankle during standing balance perturbations. Their results showed that muscle responses were delayed, with longer latencies than the CG. A distinct evolution in the organization of these muscle responses was also observed according to age group. More recently, changes in people with DS have been observed in the transition from adolescence to adulthood; adults tend to have longer latencies in muscle response but are better able to adjust and adapt to various situations.20 With regard to participant age, that study group was comparable to ours. Differences in people with DS have also been observed in muscle activation and the stabilizing role of the muscles.17

We know that the role of the joint-stabilizing muscles is important for the implementation of standing balance exercises.44 The role of anticipatory postural adjustments in unstable conditions is extremely important, as they should enable a response to varying types of instability in order to provide improved stability and performance of tasks.45 Probably, the high degree of muscle activity observed in DS group is related to compensatory safety strategies.

Differences between the groups, with the opened eyes, were smaller following training, and were observed in the plantar flexors (soleus and gastrocnemius) but not in the tibialis anterior. Essentially, the behaviour of the tibialis anterior became more similar between the two groups. The differences in activation of the tibialis anterior likely disappeared because of a decrease in muscle activity in the DS group following training, while the tibialis anterior remained more active in the CG. It would have been useful to assess COP position in relation to the changes in activation of the flexor and extensor muscles. This would have allowed us to analyze whether the change in the behaviour of the flexors and extensors in both groups was due to a change in the positioning of the COP after training. In the closed eyes condition, differences between groups also disappeared for the tibialis anterior following training because the CG displayed increased activity while activation in the DS group decreased. These changes may be due to repositioning of the COP after training, which entails changes in the stabilizing activity of the flexor and extensor muscles. In our previous study, we observed that the DS group demonstrated improved stability of the COP with the opened eyes while the CG demonstrated improvements with the closed eyes.14 It is possible that the DS group gained stability with the opened eyes by learning to better integrate visual signals, while the CG learned to gain stability in the more difficult condition for them, with the closed eyes, through the improved use of touch and proprioceptive information. Perhaps these improvements stabilize the COP as well as result in position change.

When comparing each group according to visual condition before the programme implementation, the CG demonstrated increased activation of the soleus with the closed eyes. Activity in the extensors tended to decrease, but with no statistically significant differences. In contrast, the DS group showed no difference in muscle activation between the two visual conditions. We think that this difference in behaviour could be due to better sensory integration by the CG. A lack of integration of visual signals in the DS group may explain the absence of differences in terms of muscle response from one visual condition to another. Several theories explain the pathophysiologic cause of postural control deficits in people with DS, including deficits in proprioception, the somatosensory system,4 and cerebellum.18 Our observations support a possible deficit in the use of sensory information, in this case, visual information. This would result in a lack of differences in muscle reaction based on visual information.

Following training, the CG maintained increases in soleus activation and displayed some increases in gastrocnemius activation as well. The DS group displayed a trend towards reduced activity in the tibialis anterior. These changes caught our attention. In the CG with the closed eyes, the muscles were more active, but only in the plantar flexors (soleus, gastrocnemius); although this is more evident post-training. However, in the DS group, differences were observed in the tibialis anterior, with lower activation with the closed eyes post-training, when no differences were observed before. Again, we can correlate this finding to our previous studies regarding the behaviour of COP. There, we could see that the CG demonstrated improved control of their COP with the closed eyes after training (just as the eyes were also closed when the differences in flexor activity increased). We also observed that people with DS demonstrated improved control of their COP with their opened eyes (just as we observed in this study with the action of the tibialis muscle, which was higher with the opened eyes). Therefore, we think that the CG learned to control their COP in the closed eyes condition by using their muscles while the DS group did this with opened eyes.

Again, it would have been useful to assess the position of the COP in relation to the support base to see if predominant activation of the flexor or extensor muscles corresponds to different positions of the COP. Similarly, it might be useful to perform an assessment by muscle group instead of individual muscles in order to study muscle group activity via flexor or extensor moments.

The CG exhibited greater pre- to post-training differences than the DS group. In the CG, less activation of the soleus was observed in the opened eyes condition after the implementation of the programme. However, this group displayed higher activation of the tibialis anterior with the closed eyes after training.

We must also mention the variability in the results of the subjects in the DS group, which is consistent with the findings of other authors.46 This variability could explain the discordant fact that after training, minor differences between the DS group and CG were detected, while, in contrast, the intragroup pre- and post-training differences were not significant in the DS group.

One limitation of the study was the sample size. The sample was intentional, as a compact group from a single special education school. It did not allow for a control group within the DS group. We were also not able to separate the groups by gender. We were also unable to perform long-term follow-up since most students in the study transitioned from the education centre to a workplace. On the other hand, it would have been interesting to be able to study both lower extremities and to compare the behaviour of their musculature.

On the strengths side, the sample was very well controlled and confined to the same level of disability. Since the students were from a single centre, good dance programme monitoring was ensured. The consistency between the programmes applied to the DS group and CG was stringently maintained. Our study provides data for the evaluation of biomechanics and neurophysiology in DS subjects.