Retention of bimanual performance following hand arm bimanual intensive therapy in children with unilateral cerebral palsy: A six-month longitudinal study

Abstract

Hand-arm bimanual intensive therapy (HABIT) enhances upper extremity (UE) function and bimanual coordination in children with unilateral cerebral palsy (UCP). Previous studies assessed immediate improvements in UE function using clinical and self-reported measures, which may not accurately reflect real-world UE performance and their long-term retention effects. Therefore, this study aims to investigate the retention of real-world bimanual performance gains over time following HABIT in children with UCP. Thirty children with UCP, age 6-16 years underwent HABIT (6 hours/day for 5 days). Bimanual performance was assessed using GT9X Link accelerometers, worn on bilateral wrists for 3 days pre-, post-, 3-, and 6-month of HABIT. Accelerometer-derived variables-use ratio (UR), magnitude ratio (MR), bilateral magnitude (BM), median acceleration (MA), and acceleration variability (AV)-quantified bimanual performance during real-world activities. UE function was measured with standardized assessments. A mixed model analysis with repeated measures and paired t-tests analyzed the differences real-world bimanual performance and UE function respectively. There was a significant main effect of time in UR (F = 2.72, p = 0.05), BM (F = 4.36, p = 0.007), and MA (F = 3.68, p = 0.016). Post-hoc analysis (mean differences, 95% confidence interval [CI]) revealed improvements immediately post- compared to pre-HABIT in BM (14.99, 4.35-25.63) and MA (7.46, 2.55-12.36). However, subsequent assessments at 3- and 6-months displayed a regression in these gains, suggesting a lack of retention. A decline was observed at 3 months) and 6 months (BM; 16.94, 6.3-27.4, MA; 6.51, 1.61-11.41) in BM and MA compared to post-HABIT. UE capacity measures also showed improvements (p < 0.05) post-HABIT. Although HABIT initially may enhance performance of real-world bimanual tasks, its benefits diminish within six months, suggesting a need for repeating HABIT every 3-6 months to retain long-term improvements.

Fig 1. Representative examples of density plots.

Graphical representations of accelerometer data of both the upper extremities (UE) of (a) a typically developing child and (b) an age-matched child with right sided UCP pre-HABIT. The density plots are obtained by plotting the accelerations on a second-by-second basis recorded over a total wear time of 3 days pre-HABIT. The x-axis (magnitude ratio) indicates the contribution of each extremity to the task, and the y-axis indicates the overall intensity of movement. The right and left halves of the plots represent right and left UE use, respectively. In Fig 1b, the right half is more affected, and left is less affected UE of the child with UCP. The dots seen in the graphs represent the counts (number) of movements (accelerations) performed by each UE. The colors in the large color bar scale on the right of the plot indicate the frequency of movements, where brighter colors represent greater frequencies and vice versa. In the above representative examples, Fig 1a appears noticeably symmetrical as compared to Fig 1b. It indicates that the typically developing child used both UEs equally in terms of hours, whereas the child with UCP used the affected UE considerably less (right half of Fig 1b) relative to left, resulting in an asymmetric density plot, which is termed as use ratio, a measure of duration of affected relative to less affected UE. Magnitude ratio is the contribution of the affected relative to the less affected UE in terms of intensity of movements. The dots seen in both the halves of Fig 1a appear at similar heights suggesting symmetrical magnitude of both UE movements in a typically developing child. Whereas, in Fig 1b, the dots representing right (affected) UE are distinctly lower as compared to less affected UE, resulting in MR value moving towards negative side. Bilateral magnitude is the overall intensity of both UEs. The overall height of Fig 1a is taller relative to Fig 1b, indicating lower intensity of movements in the child with UCP.

Fig 2. Schematics of the study timeline.

Abbreviations: AHA = Assisting hand assessments, JHFT = Jebsen hand function test, NHPT = Nine-hole peg test.

Fig 3. CONSORT diagram.

A total of 81 participants were assessed for eligibility and 30 participants eventually participated and completed the HABIT protocol. Twenty-nine participants were included in the immediate post-HABIT analysis. Six participants were excluded during the follow-up longitudinal accelerometer assessments. Twenty-three participants were included in the final analysis.

Fig 4. Differences in mean scores of upper extremity performance in all five accelerometer-derived variables across four time points- pre-HABIT, post-HABIT, 3-month, and 6-month post-HABIT.

Values reported are means ± SEM as determined by distribution of data during each assessment time point. (a) Use ratio, (b) Magnitude ratio, (c) Bilateral magnitude, (d) Median acceleration, and (e) Acceleration variability. Pre-HABIT refers to baseline assessment and post-HABIT refers to assessments immediately within a week following HABIT. 3-month and 6-month time points refers to accelerometer data acquired 3- and 6-months following HABIT to assess long term retention of bimanual performance. The trends seen in the line graphs above indicate a decrease in all the accelerometer derived variables at 3-month time point (except use ratio) suggesting decrease in bimanual performance. Furthermore, all the accelerometer variables nearly returned to baseline or pre-HABIT levels at the 6-month follow up assessments. *denote significance with p value ≤ 0.01 on Bonferroni pairwise comparison between the study time points.

Fig 5. Comparison of differences in the mean scores of capacity measures pre- and post-HABIT training.

Values reported are means ± SEM as determined by distribution of data during each assessment time point. Pre-training refers to baseline assessment and post-training refers to assessment within one week following HABIT. There were significant changes seen in mean scores of (a) Assisting Hand Assessment, (b) Jebsen Hand function test (more affected UE), and (c) Nine Hole Peg test (more affected UE) from pre- to post-training. Children demonstrated greater change in the AHA indicating greater function of the affected arm and bimanual coordination, faster speeds in JHFT and NHPT during post-training assessment compared to baseline. *denotes p < 0.05, ** denotes p < 0.005.