Effects of core stability exercises on balance among Chinese children and youth with intellectual disabilities

Study design

This study was a single-center, two-armed, evaluator-blinded randomised controlled trial (RCT). The study protocol was registered on the website of the Chinese Clinical Trial Registry (Registration no. ChiCTR2400088510). Ethical approval was obtained (ZSRT2024189) from the Ethics Committee of Zhejiang Normal University. The study was reported in accordance with the Consolidated Standards of Reporting Trials (CONSORT) statement for randomised trials (Schulz et al., 2010).

Participants and setting

Children and youth with IDs were recruited from a special school in a city in southeastern China. This school includes primary school (grades 1–6, aged 6–12 years) and middle school (grades 7–9, aged more than 13 years) and provides educational services for students with various types of disabilities.

Initially, each class teacher recommended potentially eligible students for the researchers according to the following criteria: (1) presence and severity of ID were diagnosed by a physician based on the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 2016); (2) aged 6–18 years old; (3) able to follow instructions and communicate verbally; and (4) had no history of core stability, strength or balance training during the past six months. The exclusion criteria included (1) an uncontrolled cardiovascular disorder; (2) neurological disorders (e.g., epilepsy); (3) vision or hearing impairments that hinder exercise performance; and (4) musculoskeletal limitations, such as spinal, hip, knee or foot deformities. Among the 53 individuals who consented to participate, 15 were not included because of the exclusion criteria (Fig. 1).

Randomisation and masking

After completing the baseline assessment, the participants were randomly assigned (1:1) to receive either core stability exercises (intervention group) or their regular school schedule (control group). Allocation was based on a computer-generated random number list prepared by a researcher with no involvement in the trial. Specifically, a sealed envelope contained a slip of paper with a random number indicating the group assigned to each subject after baseline assessments.

As in most trials, it was not possible to “blind” the subjects regarding the intervention options. In the present study, only the assessor was blinded; therefore, the study did not involve unblinding.

Intervention and control

Before the trial began, all assessment and intervention personnel received uniform training and developed common operating standards. Intervention and evaluation procedures were carried out in strict accordance with standardised methods.

The participants in the intervention group received Jeffreys’ core stability exercises (Jeffreys, 2002; Aly & Abonour, 2016; Ghaeeni, Bahari & Khazaei, 2015; Alsakhawi & Elshafey, 2019), which included lumbar–pelvic proprioception retraining, specific spinal stabilisation exercises, and different muscle contractions and abdominal maneuvers (Alsakhawi & Elshafey, 2019). In particular, this protocol consists of three progressive levels: level 1 is static contraction training conducted under stable conditions; level 2 is dynamic training, which is also performed under stable conditions; and level 3 includes dynamic and resistance training, which is performed under unstable conditions (such as with Swiss balls). The specific contents and methods of this training are shown in Table 1.

Participants in the intervention group received training three times a week, 35 min per session, for a total of 6 weeks. They were also encouraged not to seek any physical therapy other than core exercises during the intervention. Participants in the control group continued their regular school schedule (i.e., 35 min of PE lessons, three times per week) and were encouraged to maintain their current level of physical activity.

Outcome measurements

Two assessors who were masked with respect to group allocation conducted the assessments at baseline and at week 7. The height (cm) and weight (kg) of the participants in both groups were measured using a portable instrument (Inbody BSM370, Korea). Height was measured to the nearest 0.10 cm in bare feet, and weight was measured to the nearest 0.10 kg. Body mass index (BMI; kg/m²) was subsequently calculated.

Assessment of static balance

One-legged stance (OLS) test was used to evaluate static balance. The participants were instructed to stand on the dominant leg and place the heel of the other (free) leg on the knee of the supporting leg. Hands were placed on both sides of their hip. The assessment was conducted under four conditions: (1) standing with eyes opened on firm ground (EO, FI); (2) standing with eyes closed on firm ground (EC, FI); (3) standing with eyes opened on a foam pad (i.e., AIREX balance pad with 40  × 50 cm dimensions and six cm thickness; EO, FO); and standing with eyes closed on a foam pad (EC, FO). Under these four conditions, the participants were encouraged to stand for as long as possible to 60 s. The test is completed when the participant’s free leg is placed on the floor. If they opened their eyes during a test requiring closed eyes, they were required to take the test again. The assessors used a chronometer to record the participants’ standing time in seconds. The best of three attempts was recorded. The OLS test has been applied to children and youth with IDs (Rahmat & Hasan, 2013; Yılmaz et al., 2009) and has demonstrated good reliability and validity (Blomqvist et al., 2012).

Assessment of dynamic balance

A lower quarter Y balance test (YBT-LQ) kit (Functional Movement Systems, Chatham, VA, USA) was used to measure the participants’ dynamic balance performance. The test is a valid and reliable measure of dynamic balance in children and youth with IDs (Ahmadi, Hasan & Hosin, 2012; Balayi, Sedaghati & Ahmadabadi, 2022; Jouira, Rebai & Sahli, 2023). The test kit consisted of a centralised stance platform to which three pipes were attached to represent the anterior (AT), posteromedial (PM) and posterolateral (PL) reach directions. Each pipe was marked in 1.0 cm increments for measurement purposes and was equipped with a moveable reach indicator.

Each participant was asked during the YBT-LQ to stand barefoot (to eliminate potential effects of varying footwear). One foot was on the center foot plate, and the most distal aspect of the toes was behind the starting line (Gribble, Hertel & Plisky, 2012). While maintaining a single-leg stance, each participant was asked to push the reach indicator with the reaching foot as far as possible and return to the original start position while maintaining balance. The test direction order was standardised as follows: right AT, right PM, right PL, left AT, left PM and left PL (Jouira, Rebai & Sahli, 2023). Each participant performed three practice trials (unaccounted tests) for each direction so that participants became comfortable with performing the task. After two minutes of rest, each participant performed three test trials in each direction (accounted tests). To avoid the potential effects of fatigue and learning, a 60 s rest period was provided between trials, during which participants were allowed to sit down. A trial was classified as invalid when the participants (1) lost their balance (i.e., stepped with the reach leg on the ground), (2) lifted the stance leg from the stance platform, (3) stepped on top of the reach indicator for support or (4) kicked the reach indicator (Muehlbauer & Waldermann, 2022; Plisky et al., 2009). When an invalid trial occurred, the corresponding data were discarded, and trials were repeated until a total of three valid trials were completed. The maximal reach distance (cm) per reach direction was used for further analysis.

Statistical analyses

The sample size was determined in advance using G*power software (version 3.1.9.2). Based on a confidence interval of 95%, a power of 80% and an attrition rate of 10%, 19 participants were needed per group, yielding a total sample size of 38.

Data were analysed by researchers using SPSS 27.0 (IBM Corp., Armonk, NY, USA). The normal distribution of the data was analysed by the Shapiro–Wilk test and histogram graphs. Descriptive analyses are presented as the mean and standard deviation for continuous variables and numbers and percentages for categorical variables. Prior to conducting independent t tests, Levene’s test for equality of variances was performed. The baseline demographic characteristics of categorical and continuous variables were compared between groups using chi-square tests and independent t tests, respectively. Differences within each group before and after the intervention were analysed using paired t tests. The differences between the groups before and after the intervention were analysed via independent t tests. Statistical significance was set at P < 0.05 for all tests.

Comparison of static and dynamic balance between groups before and after the intervention

The demographic characteristics of the participants in both groups are presented in Table 2. The intervention and control groups were homogenous in terms of all the demographic characteristics (P > 0.05).

The results in Table 3 show no statistically significant difference between the intervention and control groups in terms of static and dynamic balance before the intervention (P > 0.05).

After the intervention, independent t tests (Table 4) revealed statistically significant differences between the two groups in one stance condition (i.e., OLS time; EO, FI) of static balance (t = −3.105, P = 0.004) and in all directions for both legs of the dynamic balance (left leg: AT direction: t = −4.510, P < 0.001; PM direction: t = −4.241, P < 0.001; PL direction: t = −4.403, P < 0.001; right leg: AT direction: t = −4.920, P = 0.004; PM direction: t = −4.821, P < 0.001; PL direction: t = −4.963, P < 0.001).

Effects on static balance performance

Pre- and post-comparisons within the groups revealed significant differences in the intervention group in only one stance condition (i.e., OLS time; EO, FI; t = −5.269, P < 0.001) but not in the control group in all four stance conditions (Fig. 2).

Effects on dynamic balance performance

As shown in Fig. 3, the results of the paired t tests revealed that in the intervention group, the AT, PM, and PL directions for both legs of the dynamic balance were significantly different before and after the intervention (left: AT direction: t = −3.197, P = 0.005; PM direction: t = −4.608, P < 0.001; PL direction: t = −4.706, P < 0.001; right: AT direction: t = −3.324, P = 0.004; PM direction: t = −5.614, P < 0.001; and PL direction: t = −5.905, P < 0.001) but not in the control group.