Print - Back

Week 8 Report: Hand Tremors
Generated Mar 3, 2018 by rosemary.plapp1

Introduction: Many individuals with Parkinson's disease struggle to maintain their independence as the disease progresses. According to O'Connell and Guidon, the disease makes it difficult for individuals to multitask and perform self-care tasks (O'Connell & Guidon, 2016). Hand tremors, a common result of the disease process, and make feeding especially difficult. It is hypothesized that weighted eating utensils can assist individuals with hand tremors in feeding themselves. This study investigates the impact of weighted utensils and wrist cuffs on tremor amplitude. 

Methods: 16 subjects with Parkinson's disease had their hand tremor amplitude measured in millimeters (mm) while holding a built-up spoon (108 grams), holding a weighted spoon (248 grams), and holding the built up spoon while wearing a weighted wrist cuff (470 grams). As the subjects subjects were measured multiple times, a repeated measures ANOVA was used to determine the relationship between weight and hand tremors. The hand tremor amplitude constitutes interval/ratio data. A QQ plot was utilized to ensure that residual variability was normally distributed. A scatter plot was constructed to ensure that the residual variability exhibited homogeneity of variance. Observations were independent, and an interaction plot was utilized to determine if there was an interaction between the block and treatment. As the assumptions for repeated measures ANOVA were satisfied, the test was performed. Tukey 95% confidence intervals were then obtained from the results. 

Analysis:

The QQ plot below indicates that residual variability is normally distributed.

<result1>

The scatter plot of residuals displays a fairly consistent vertical spread from left to right, with the exception of 2 outliers. It can still be assumed that the residual variability exhibits homogeneity of variance. 

<result2>

 There is no evidence of interaction between the block and treatment.

<result3>

Based on the p-value of 0.2741, there is a lack of evidence of differences in the mean hand tremor amplitude between the use of the built-up spoon, weighted spoon, and combination of built-up spoon and weighted wrist cuff. All confidence intervals obtained included "zero" as a value, indicating that these results are likely not clinically important. The largest difference in hand tremors was with weighted spoons, which improved tremors by up to 0.20466643 mm. This improvement is considered minimal

<result4>

Conclusion: Individuals with Parkinson's disease are able to maintain their independence longer with the incorporation of devices that reduce their symptoms. Unfortunately, this study found that there is a lack of evidence of improvement in tremors when individuals with the disease use items of increased weight compared to a built-up spoon when eating. Future studies may incorporate stratification methods, dividing subjects by disease stage. Additionally, it would be intersting to evaluate subjects' tremors when holding normal silverware in addition to with the weighted items.

References:

O'Connell, E., & Guidon, M. (2016). Fear of falling and dual-task performance in people with Parkinson's disease. European Journal of Physiotherapy, 18(3), 167-172. doi:10.3109/21679169.2016.1156156

 

Result 1: QQ Plot of Residuals   [Info]
Right click to copy



Result 2: Scatter Plot of Residuals   [Info]
Right click to copy



Result 3: Repeated Measures ANOVA Interaction Plot   [Info]
Right click to copy



Result 4: Repeated Measures ANOVA tables   [Info]
Repeated measures ANOVA results:
Responses: Tremor Amplitude
Treatments: Feeding Method
Blocks: Subject

Response statistics by treatment
Feeding MethodnMeanStd. Dev.Std. Error
built-up160.4956250.37945520.0948638
built-up and cuff160.5350.456172480.11404312
weighted160.57750.4840730.12101825


ANOVA table
SourceDFSSMSF-StatP-value
Feeding Method20.0536541670.0268270831.3516890.2741
Error300.59541250.019847083

Tukey 95% Simultaneous Confidence Intervals:
built-up subtracted from
DifferenceLowerUpperP-value
built-up and cuff0.039375-0.0834164330.162166430.7116
weighted0.081875-0.0409164330.204666430.2434
built-up and cuff subtracted from
DifferenceLowerUpperP-value
weighted0.0425-0.0802914330.165291430.6733

Fitted values stored in new column: Fit
Residuals stored in new column: Residuals