Implicit Hand Representations in Typical Ageing and in Parkinson's Disease
Implicit Hand Representations in Typical Ageing and in Parkinson's Disease
16 September 2022
Having an internal representation of one’s own body is important for many interactions with the environment, and in making decisions about what actions we are capable of performing. However, even in healthy adults, these representations are known to be distorted. In the hand specifically, individuals are likely to underestimate the length of all fingers, but overestimate the distance between each adjacent pair of knuckles. Both healthy ageing and Parkinson’s Disease (PD) include apsects which are known to further distort body representations, including, but not limited to, diminished tactile sensitivity and impaired action capabilities. This study was designed to investigate the accuracy of hand representations in typical ageing and in PD. Fourteen participants with mild to moderate PD, 17 healthy age-matched controls and 20 younger controls made estimates about the location of hand landmarks when the hand was hidden from view. Estimations were compared with actual hand size. Older controls and individuals with PD both demonstrated more accurate representations of thumb length, and of distance between the index and middle knuckles than younger controls, with older controls also showing differences in their perception of distance between thumb and index knuckles. However, no differences were found between the PD group and older controls, suggesting that the formation of body representations is an ability which is preserved in PD. Possible explanations for, and implications of these results are discussed.
LUSTRE, aquisition form, wordpress
Participants To determine the number of participants necessary, a priori power analysis was conducted in G*Power (Faul et al., 2009), using α= 0.05, β= .08 and effect size = 0.32. This effect size was calculated from Longo (2014), which employed a similar methodology. The analysis determined that 10 participants in each condition were required to yield sufficient power. Previous studies using this methodology have included sample sizes ranging from 12-22 participants (Longo & Haggard, 2010, 2012; Peviani & Bottani, 2020). The intended sample size, therefore, was 20 participants per condition. Due to the time constraints of the study, this number was not reached for all conditions, but all conditions included more than the 10 participants needed as suggested by the priori analysis. 20 younger controls were tested (15 female). Their ages ranged from 19 to 30 (M = 22.40 yrs, SD = 2.21 yrs). 17 were right-handed, and 3 were left-handed, with handedness ranging from -89.5 to 100 (M = 64.52, SD = 61.84) on the Edinburgh Handedness Inventory (EHI; Oldfield, 1971). 17 healthy older controls were tested (11 female). Their ages ranged from 52 to 79 (M = 66.12 yrs, SD = 9.16 yrs). 14 were right-handed, and 3 were left-handed, with handedness scores ranging from -100 to 100 (M = 65.29, SD = 77.31). 14 individuals with PD were tested (4 female). Their ages ranged from 54 to 78 (M = 65.93 yrs, SD = 8.43 yrs). All PD participants were right-handed, with handedness scores ranging from 33.5 to 100 (M = 88.31, SD = 21.20). There was no significant difference between the ages of the participants in the typically ageing and the PD condition, t(29) = 0.06, p = .95. For the PD participants, the most recent onset of PD was 3 years ago, with the longest diagnosis of 20 years (M = 7.75 yrs, SD = 4.81 yrs). All presented with a Hoehn and Yahr Stage of 3 or below. This indicated that all participants were physically independent. All participants had been prescribed antiparkinsonian medication, and they were all tested under their normal medication regime. Younger controls were recruited through use of social media and personal connections of the researcher. PD participants were recruited through a Parkinson’s Research Interest Database which was developed by the researcher’s supervisor (Dr Megan Readman), and by contacting a local branch of Parkinson’s UK. Older controls were primarily friends and family of PD participants. Materials 24 hours before testing, participants were asked to submit demographic information in a questionnaire created using the design software Qualtrics (Qualtrics, Provo, UT). Participants’ hand movements were recorded by an Xbox Kinect camera, mounted on the ceiling directly above the hand. The camera had a resolution of 640x480 pixels, and a frame rate of 30 captures per second. The recording was made using the Kinect Studio application. Within the frame of the recording, a 30cm ruler was placed, to allow for conversion of pixels to centimetres during analysis. During the experiment, the board used to hide the participants’ hands from view was a piece of black cardboard, approximately 85x60cm. The board was 2mm thick and completely opaque. The board was positioned approximately 10cm above the hand, and was supported in this position by 5 cylindrical weights (one under each corner of the board, and one placed centrally). At each side of the board was a small mark of duct tape. This was to indicate where the participants should point between each trial. A mark was placed on each side of the board, as the handedness of the participant determined which hand they used during testing, and therefore determined which side of the board was easier to point to. Participants were asked to point using a red straw, approximately 10cm long. All participants completed the EHI (Oldfield, 1971). This includes a list of tasks (for example, writing or striking a match), for which the participant must indicate which hand they prefer to use. The response options include a strong or slight preference for the right or left hand, or no preference. A score of 100 indicates pure right-handedness, while a score of 100 indicates pure left-handedness. Participants in all conditions were screened for cognitive impairments using the Addenbrookes Cognitive Examination (ACE-III; Hodges & Larner, 2017). This assessment included 19 tasks which examine cognitive function on 5 separate domains; attention (e.g. ‘count down from 100 in 7’s’), memory (e.g. ‘remember this name and address’), fluency (e.g. ‘name as many animals as you can in one minute’), language (e.g. ‘write two full sentences’) and visuospatial reasoning (e.g. ‘draw a clock which reads 10 past 5). Typically, a score of less than 87 out of 100 would be considered abnormal, however, as some aspects of the ACE-III require participants to perform motor tasks, it is accepted that the best cut-off score to identify cognitive impairment in Parkinson’s is 80 points (Kaszás et al., 2012). Using this assessment as an exclusion criterion, only 1 PD participant’s data was removed from further analysis. There was no significant difference in the ACE-III scores of the remaining participants between the three conditions, F(2, 48) = 2.10, p = .13. Participants in the PD condition were also assessed using the Movement Disorder Society- Unified Parkinson’s Disease Rating Scale (MDS-UPDRS; Goetz et al., 2008), to determine the severity of PD symptoms at the time of testing. The UPDRS assesses both the motor and the non-motor symptoms of PD. The non-motor assessment involves questions about the individual’s experience of symptoms during the past week, for example how well they are sleeping, and if they are experiencing tremors regularly. A motor assessment is also conducted, with the participants performing tasks such as opening and closing their hand as quickly as possible, and walking from one side of the room to the other. The researcher was also required to make judgements about the severity of typical PD symptoms such as tremors and rigidity present throughout the examination. All questions and tasks are scored on a scale of 0 to 4, with 0 indicating no impairment, and 4 indicating severe impairment. This assessment has previously been validated and determined to be a reliable indicator of the severity of PD symptoms at the time of testing (Gallagher et al., 2012; Martinez-Martin et al., 2013). Testing occurred in the action and perception lab in the Whewell building at Lancaster University. This study received ethical approval from the Ethics Department of Lancaster University. Procedure Participants were emailed an information sheet 24 hours in advance to inform them of the requirements of the study. This email also directed them to a Qualtrics survey, where they were asked to submit their demographic information (age and sex). Here, they also completed the EHI, and were asked to confirm that they had normal or corrected-to-normal vision. On the day of testing, participants were first screened for cognitive impairment using the ACE-III. At this point PD participants also completed the full MDS-UPDRS. After the recording had started, participants were asked to place their dominant hand (as determined by the EHI) on the table in front of them. They were asked to move their chair so that their hand was aligned with the middle of their body. The participants were instructed to not move their hand throughout the experiment, before an occluding board was placed so that the participants could no longer see their hand. They were asked verbally to confirm that this was the case. Participants were given a straw to use as a baton with which to point. They were then directed to use the straw to point on the board, directly above where they believed specific locations of the hand to be. 10 different locations were used: the tips of each finger, and the knuckle where each finger meets the palm of the hand. Small duct tape marks were placed on the knuckles of each finger. This was done both to ensure that the participants were clear about which knuckles were intended, and also so that location of the knuckle would be clearer on the recording. The location for each trial was read aloud by the experimenter. Between each trial, participants were asked to move the straw to point at a duct tape mark on the side of the board. This was to ensure that all estimations were made where participants believed their hand to be, instead of them using alternative methods such as measuring where they believe one location to be based on the previous location. One block of testing consisted of 10 trials (one trial for each hand landmark). For the younger control condition, participants were directed to each landmark 10 times, meaning that data were obtained over 10 blocks. However, testing of the first PD participant determined that asking participants in this condition to complete all 10 blocks was not a viable option. Individuals with PD suffer from motor fatigue ability (Fabbrini et al., 2013) and multiple repetitive tasks led to an increased severity of PD symptoms such as tremors. For these reasons, all subsequent participants only completed 5 blocks of 10 trials each. This ensured we still had 5 estimations for each landmark, without causing distress to participants. Two different random orders were created for the presentation of the locations, and these were randomly assigned to participants. After testing, the occluding board was removed so that the recording could be used to ensure that the hand had not significantly moved throughout the testing period, before the recording was ended. Data Analysis To determine both the actual and estimated locations of the hands, the recordings were replayed using the Kinect Studio software. For each trial, the footage was paused when the participant had the straw pointed at the estimated location. The cursor was then moved to this point, and the x and y coordinates of the cursor was manually inputted into a spreadsheet. The same method was used to determine the actual position of each hand while the occluding board was not in place. The beginning and end of each recording was examined to confirm that the hand had not moved between the start and the end of the experiment. It was often the case that although the hand had not moved in any significant way, there was a couple of pixels difference in the position of a few landmarks. For this reason, the x and y coordinates of the hand position was recorded both before the board was placed, and after it was removed, and the average of these locations was used. For analysis, we were interested in the overestimation of the length of each finger and of the distance between each pair of adjacent knuckles. To calculate the length of each finger, the difference between the x coordinates of the tip and knuckle of the finger was calculated, and the same was done for the y coordinates. Pythagoras’s theorem was then employed to determine the distance, leading to the following formula: The same formula was adapted to determine the distance between each pair of knuckles. These distances were calculated for each block of 10 trials, and then the average was taken for each participant, before being compared to the actual measurements to calculate the percentage overestimation of each distance. For the detection of outliers, all estimations were plotted using RStudio. Code was adapted from Helbing (2020) to plot an ellipse for each hand location per participant, which encompassed at least 80% of all data points. Estimations outside these ellipses were treated as outliers and removed from further analysis. Setting the inclusion of data points to 80% meant that even for older participants, who only performed 5 trials per location, it was still possible for outliers to be seen outside of the ellipse. RStudio did not have the capacity to plot 10 separate ellipses at once, therefore 2 separate plots had to be made per participant. Before analysis, hand maps were also created using RStudio. Although these plots were not used for analysis, they helped to visualise the data. All hand maps can be found in the Appendices.
Statistical Analysis Type
Cati Oates, “Implicit Hand Representations in Typical Ageing and in Parkinson's Disease,” LUSTRE, accessed June 6, 2023, https://www.johnntowse.com/LUSTRE/items/show/177.