ABSTRACT
Background
Deficits in fine motor skills may impair device manipulation including touchscreens in people with Parkinson's disease (PD).
Objectives
To investigate the impact of PD and anti‐parkinsonian medication on the ability to use touchscreens.
Methods
Twelve PD patients (H&Y II‐III), OFF and ON medication, and 12 healthy controls (HC) performed tapping, single and multi‐direction sliding tasks on a touchscreen and a mobile phone task (MPT). Task performance was compared between patients (PD‐OFF, PD‐ON) and HC and between medication conditions.
Results
Significant differences were found in touchscreen timing parameters, while accuracy was comparable between groups. PD‐OFF needed more time than HC to perform single (P = 0.048) and multi‐direction (P = 0.004) sliding tasks and to grab the dot before sliding (i.e., transition times) (P = 0.040; P = 0.004). For tapping, dopaminergic medication significantly increased performance times (P = 0.046) to comparable levels as those of HC. However, for the more complex multi‐direction sliding, movement times remained slower in PD than HC irrespective of medication intake (P < 0.050 during ON and OFF). The transition times for the multi‐direction sliding task was also higher in PD‐ON than HC (P = 0.048). Touchscreen parameters significantly correlated with MPT performance, supporting the ecological validity of the touchscreen tool.
Conclusions
PD patients show motor problems when manipulating touchscreens, even when optimally medicated. This hinders using mobile technology in daily life and has implications for developing adequate E‐health applications for this group. Future work needs to establish whether touchscreen training is effective in PD.
Keywords: Parkinson's disease, touchscreen skills, dopaminergic medication, upper limb
Dopaminergic depletion in the basal ganglia, the main deficit underlying Parkinson's disease (PD), results in a variety of symptoms. 1 While loss of manual dexterity significantly impairs activities of daily living, 1 , 2 it has received less research attention in comparison to gait and balance problems. Poor manual dexterity may affect the use of touchscreens to operate mobile devices, which are an integral part of daily life. Furthermore, the interest in touchscreen applications to monitor disease progression or training programs is growing in PD. 3 Here, we aim to investigate the specific problems with touchscreen manipulations. Increasing the understanding of these deficits will inform the design of specific training interventions to improve touchscreen skills, so that people with PD are able to participate in using mobile technology.
Recent work revealed slower performance when using a smartphone application, including tapping and sliding movements, in PD patients with a higher score on the motor part of the Movement Disorders Society Unified Parkinson's Disease Rating scale (MDS‐UPDRS‐III). 4 Interestingly, 40% of patients reported to experience difficulties with the application due to ‘hand clumsiness’. 4 Also, deficits in manual dexterity impeded the use of touchscreen devices in PD. 5 Both studies included patients ON medication without a comparison with healthy controls (HC). Recent research demonstrated that slower performance and higher numbers of tapping errors on a smartphone discriminated PD patients ON medication from HC 6 as well as slower speed to type a telephone number on a smartphone. 7
As for the effects of medication on upper limb skills, some studies showed a faster performance and improved movement vigor, 8 though at a cost for movement accuracy. 9 , 10 Others found no beneficial effects. 2 For touchscreen skills specifically, Wissel et al. 11 revealed that improved tapping frequency and decreased tapping accuracy could distinguish between ON and OFF medication in PD.
Given these inconsistent effects and the fact that few studies investigated the impact of PD and medication on touchscreen skills, we developed a test battery involving tasks with a greater variety of motor demands than merely tapping, such as grabbing and sliding movements. Based on the literature, we hypothesized that touchscreen skills would be compromised in PD patients compared to age‐matched HC and that dopaminergic medication would result in invigoration of movement, i.e., an improvement of timing parameters, but not necessarily a more accurate performance.
Methods
Participants
Fourteen PD patients and 12 age‐matched HC were recruited from the database of the Department of Rehabilitation Sciences, KU Leuven, Belgium. Inclusion criteria for PD patients consisted of Hoehn and Yahr (H&Y) stage I‐III 12 , a PD diagnosis according to the United Kingdom PD Society Brain Bank criteria 13 and right handedness, measured by the Edinburgh Handedness Inventory. 14 Exclusion criteria for all participants were: Mini‐Mental State Examination (MMSE) < 24, 15 neurological disorders besides PD and upper limb deficits unrelated to PD that might interfere with task performance. This study was approved by the local Ethics committee UZ/KU Leuven according to the code of Ethics of the World Medical Association (Declaration of Helsinki, version 2013, S61793). Prior to participation in the study, an informed consent form was signed after explanation of the study protocol.
Experimental Procedure
This study consisted of one session, either in a quiet room at the Department of Rehabilitation Sciences of KU Leuven or at the participant's home. First, PD patients performed an extensive motor assessment OFF medication in the morning, between 12 and 15 hours after medication intake. Tests included the MDS‐UPDRS‐III 16 , the Purdue Pegboard test (PPT) 17 and a newly developed test battery of touchscreen skills (see below). A visual analogue scale (VAS) assessed fatigue experienced in hand/finger after each task of the test battery. Moreover, a mobile phone task (MPT) measured the time needed to type a predefined telephone number on a smartphone. 7 Subjects performed three trials of the MPT, each trial involving a different number. The average of the second and third trial was calculated.
Next, patients took their normal dose of medication. In the period between intake and optimal functioning of dopaminergic medication (±1 hour), a number of questionnaires were administered. These included the dexterity questionnaire (DEXTQ‐24) 18 and the Hospital Anxiety and Depression Scale (HADS). 19 Cognition was examined with the Montreal Cognitive Assessment (MoCA) 20 and the Trail Making Test (TMT). 21 The Mobile Device Proficiency Questionnaire (MDPQ‐16) 22 and smartphone specific questions (see Supplementary Appendix) assessed daily smartphone use. Further, a medication anamnesis was taken, allowing the calculation of the levodopa equivalent daily dosage (LEDD). 23 , 24 Finally, PD patients repeated the motor assessment in ON, when their medication was working optimally. The same experimental procedure was applied in HC, but without administering medication and the PD‐specific assessments and questionnaires.
Test Battery of Touchscreen Skills
A test battery of touchscreen skills, consisting of three tasks, was developed on a touch‐sensitive tablet (HP Elite x2 1012 G2 Hybrid Notebook) using a graphical programming environment with LabVIEW Software (version 18.0f2, National Instruments, Austin, TX, USA). The tapping task required participants to tap between two dots, 200 pixels apart, starting with the left dot (Fig. 1, Tapping). Three trials of 30 repetitions were completed. In the single sliding task, subjects had to slide a dot over a distance of 500 pixels to a predefined target from left to right (Fig. 1, Single Slide). This sliding movement was repeated 30 times during three trials. During the more complex multi‐direction sliding task the starting position of the dot randomly varied between four positions: 300 pixels to the (1) left, (2) right, (3) above or (4) below the target (Fig. 1, Multi Slide). This was repeated for three trials of 32 repetitions, i.e., eight slides in each direction. The size of the dots and the blue square were kept consistent throughout the session, i.e., 50 × 50 pixels and 100 × 100 pixels respectively. All tasks were performed as fast and accurately as possible with the right index finger. Considering the test battery was new, a repeatability analysis was performed (Supplementary Material S1).
FIG. 1.
Tablet tasks. Tapping between two dots (left panel). Sliding a dot towards a predefined target in a single direction (middle panel). Sliding a dot towards a predefined target in multiple directions (right panel).
Outcome Measures
Both timing and accuracy parameters were automatically recorded by the custom‐made application with a temporal resolution of 1 ms and a spatial resolution of 0.135 mm. For the sliding tasks, timing parameters included the total sliding time (ms), i.e., the time necessary to perform a separate sliding movement. The transition time (ms) involved the time in between these sliding movements. Onset was defined as the moment of releasing the dot and termination was determined as the moment of grabbing the next dot on the screen. Accuracy parameters for the sliding tasks, consisted of the error distance (pixels), measuring the deviation between the target center and the actual release point, as well as the number of correct responses (%), i.e., if the dot was released within the predefined target or not. For the tapping task, timing parameters consisted of the inter‐tap interval time (ms), defined as the time in between tapping movements. To measure tapping accuracy, the number of correct tapping movements (%), i.e., inside the green dot, was collected. For all tasks, the first repetition of each trial was excluded from the analysis.
Statistical Analyses
Statistical analysis was conducted with SPSS software (version 24 SPSS, Inc., Chicago, IL, USA) with a significance level of α < 0.05. Data distribution was assessed using Shapiro–Wilk tests and Q‐Q plots. Depending on the normality of the distribution, independent t‐tests or Mann–Whitney U tests compared PD‐OFF with HC and PD‐ON with HC. A Chi‐squared test compared gender distribution between groups. Paired t‐tests or Wilcoxon tests contrasted medication conditions. Also, a non‐parametric McNemar test examined H&Y stages between medication groups. We also calculated an upper limb score of the MDS‐UPDRS‐III between medication conditions, consisting of item 3.3 to 3.6 and 3.15 to 3.18 (if item 3.17 for left or right arm was ≥1). For each comparison, we corrected for the different parameters per task using a Bonferroni method for multiple testing. The corrected P‐values and effect sizes are reported. A non‐parametric effect size estimate r was calculated using the formula: r = z/ √ N (z = Z‐score and N = number of observations). Effect size estimates range from −1 to +1 with values further away from zero indicating larger effect sizes (i.e., ±0.1, ±0.3 and ±0.5 representing small, medium and large effect sizes, respectively). 25 , 26 Exploratory correlation analyses were performed between the timing parameters and clinical characteristics (age, Purdue Pegboard Test, DEXTQ‐24, MDS‐UPDRS‐III, MDS‐UPDRS‐III items 15a, 16a and 17a, MoCA, TMT, HADS) and between timing parameters and daily smartphone performance (MDPQ‐16, MPT). Spearman correlations were performed across groups and significant correlations (P < 0.05) were repeated for both groups separately (PD‐OFF and HC).
Results
Participants
Clinical characteristics are displayed in Table 1. Twelve PD patients and 12 HC completed the study. Two patients were excluded: One patient had already taken the morning dose of medication upon arrival and one was discontinued due to ill‐health. Other incomplete data related to left upper limb task execution, were found in two patients mostly due to fatigue.
TABLE 1.
Clinical characteristics and demographics for all participants
| PD patients (N = 12) | P‐value | P‐value | |||
|---|---|---|---|---|---|
| PD‐OFF | PD‐ON | HC (N = 12) | (PD‐OFF vs HC) | (OFF vs ON) | |
| Age (years) | 64.7 (6.7) | 68.7 (6.9) | 0.160 | ||
| Gender (M/F) | 9/3 | 8/4 | 0.653 | ||
| EHI (%) | 95 (90; 100) | 100 (100; 100) | 0.198 | ||
| MMSE (0–30) | 29.5 (26.8; 30) | 29 (28.8; 30) | 0.799 | ||
| MoCA (0–30) | 27.3 (1.9) | 25.6 (2.8) | 0.084 | ||
| TMT (B‐A) (s) | 45.5 (29.1) | 37.9 (19.0) | 0.455 | ||
| HADS‐Anxiety (0–21) | 7.6 (4.4) | 4.7 (3.0) | 0.073 | ||
| HADS‐Depression (0–21) | 6.8 (3.7) | 2.8 (1.7) | 0.002 a | ||
| DEXTQ‐24 | 33 (27.8; 35.5) | 24 (24; 24) | <0.001 a | ||
| PPT‐R (#pegs/30s) | 9 (7; 11) | 11 (8.8; 11) | 13 (12.8; 13) | <0.001 a | 0.064 |
| PPT‐L (#pegs/30s) | 8.6 (2.2) | 9 (2.1) | 10.8 (1.6) | 0.009 a | 0.496 |
| PPT‐RL (#pegs/30s) | 12.9 (5.0) | 12.82 (4.1) | 19.1 (3.0) | 0.003 a | 0.852 |
| PPT‐Combi (#parts/ min) | 14 (12.5; 17.5) | 19 (13; 22) | 21.5 (21; 25.3) | 0.001 a | 0.089 |
| NFOG‐Q (0–28) | 4 (0; 10.8) | ‐ | |||
| H&Y (II/III) | 8/3 | 9/2 | ‐ | 1.000 | |
| MDS‐UPDRS‐III (0–132) | 36 (32; 43.5) | 26 (24.5; 30.5) | ‐ | 0.003 a | |
| MDS‐UPDRS‐III‐UL (0–60) | 19 (17; 23) | 12 (11; 16) | ‐ | 0.003 a | |
| MDS‐UPDRS‐III item 15a (0–4) | 1 (0.8; 1) | 1 (0; 0) | ‐ | 0.317 | |
| MDS‐UPDRS‐III item 16a (0–4) | 1 (1; 1) | 1 (1; 1) | ‐ | 0.046 a | |
| MDS‐UPDRS‐III item 17a (0–4) | 0 (0; 1) | 0 (0; 0) | ‐ | 0.102 | |
| LEDD (mg/24 h) | 867.1 (290.1) | ‐ | |||
| Disease Duration (years) | 10.6 (3.9) | ‐ | |||
Normally distributed variables are displayed as the mean (standard deviation). Non‐normally distributed variables are presented as the median (1st quartile; 3rd quartile).
Group significant different at P < 0.050.
PD, Parkinson's disease; HC, healthy controls; OFF, OFF medication; ON, ON medication; EHI, Edinburgh Handedness Inventory; item 15a, postural tremor, right hand; item 16a, kinetic tremor, right hand; item 17a, rest tremor amplitude, right upper extremity; MMSE, Mini Mental State Examination; MoCA, Montreal Cognitive Assessment; TMT, Trail making test; HADS, Hospital Anxiety and Depression Scale; DEXTQ, Dexterity questionnaire; PPT, Purdue Pegboard Test; R, right; L, left; RL, Bimanual; Combi, combination; H&Y, Hoehn and Yahr; MDS‐UPDRS‐III, Movement Disorders Society – Unified Parkinson's Disease Rating Scale part 3; UL, upper limb; LEDD, L‐dopa equivalent daily dosage; #, number of.
PD patients and HC did not differ significantly (Table 1), except for a higher HADS‐Depression score in PD patients (P = 0.002). In general, PD‐OFF patients had worse upper limb skills than HC, reflected by the DEXTQ‐24, Purdue Pegboard Test and MPT (all P < 0.050) (for MPT performance see Fig. S2). Dopaminergic medication improved disease severity, indicated by the lower MDS‐UPDRS‐III score and upper limb scores ON compared to OFF medication (both P = 0.003). However, there were no significant medication effects on the Purdue Pegboard Test or the MPT (P > 0.050). Importantly, tremor in the right upper limb was generally low and did not improve with dopaminergic medication (P > 0.100), apart from the kinetic tremor (P = 0.046).
Tablet Task Performance
In the Supplementary Material S1, we report on the repeatability analysis showing some learning from trial 1 to trial 3 for timing parameters, though without effects on the analysis of the pooled results.
Effect of PD
Table 2 and Figure 2 reveal that PD‐OFF performed both sliding tasks significantly slower compared to HC (P < 0.050, r < − 0.500, see Fig. 2A–B ). Moreover, PD‐OFF needed more time to capture the dot in between the slides (i.e., a longer transition time) in both sliding tasks (P < 0.050, r < − 0.500, see Fig. 2C–D ). Accuracy of sliding performance did not differ. Looking at the tapping task, neither the inter‐tap interval time, nor tapping accuracy differed significantly between PD‐OFF and HC. PD‐OFF had higher VAS scores (i.e., more fatigue) for all tablet tasks compared to HC (P < 0.050, see Table 2 ).
TABLE 2.
Comparison of performance on touchscreen skills between PD patients and healthy controls
| OFF vs HC | OFF vs ON | ON vs HC | |||||||
|---|---|---|---|---|---|---|---|---|---|
| HC (N = 12) | PD‐OFF (N = 12) | PD‐ON (N = 12) | P‐value | r | P‐value | r | P‐value | r | |
| Slide Single | |||||||||
| Total time (ms) | 400.7 (257.6; 771.7) | 805.6 (752.2; 985.2) | 645.0 (601.4; 935.6) | 0.048 a | −0.507 | 0.164 | −0.589 | 0.272 | −0.377 |
| Transition time (ms) | 503.4 (463.9; 684.1) | 736.7 (641.6; 928.1) | 645.1 (546.7; 984.1) | 0.040 a | −0.518 | 1.388 | −0.272 | 0.133 | −0.436 |
| Error (pixels) | 31.9 (19.1; 46.8) | 18.8 (13.7; 25.1) | 18.0 (16.5; 25.6) | 0.404 | −0.342 | 4.000 | 0.000 | 0.312 | −0.365 |
| Accuracy (%) | 91.1 (81.7; 94.4) | 89.4 (80.8; 96.7) | 90.6 (81.1; 96.1) | 3.372 | −0.041 | 3.156 | −0.077 | 3.020 | −0.065 |
| Slide Multi | |||||||||
| Total time (ms) | 362.6 (276.2; 497.2) | 620.5 (583.5; 836.5) | 551.7 (513.1; 701.7) | 0.004 a | −0.625 | 0.076 | −0.679 | 0.032 a | −0.530 |
| Transition time (ms) | 754.1 (698.0; 812.3) | 982.2 (868.5; 1100.9) | 847.4 (803.7; 1112.1) | <0.001 a | −0.707 | 0.468 | −0.453 | 0.048 a | −0.507 |
| Error (pixels) | 25.0 (20.6; 36.4) | 19.7 (17.7; 23.7) | 20.5 (19.0; 22.7) | 0.312 | −0.037 | 1.232 | −0.294 | 0.512 | −0.318 |
| Accuracy (%) | 83.9 (80.5; 89.8) | 85.4 (75.0; 89.3) | 79.2 (71.1; 89.3) | 3.728 | −0.024 | 0.240 | −0.544 | 1.388 | 0.195 |
| Tap | |||||||||
| Time (ms) | 263.5 (196.4; 310.2) | 332.1 (239.0; 459.4) | 277.85 (218.0; 369.4) | 0.120 | −0.389 | 0.046 a | −0.657 | 0.638 | −0.212 |
| Accuracy (%) | 92.8 (89.7; 95) | 88.9 (79.7; 97.7) | 78.3 (67.8; 85) | 1.686 | −0.041 | 0.130 | −0.533 | 0.028 a | −0.490 |
| VAS score | |||||||||
| Tap | 0.8 (0.3; 1.0) | 1.5 (0.8; 1.9) | 1.4 (1; 1.7) | 0.024 a | ‐ | 0.759 | ‐ | 0.033 a | ‐ |
| Slide Single | 0.7 (0.3; 1.3) | 1.95 (1.6; 2.75) | 1.9 (1.2; 2.6) | 0.014 a | ‐ | 0.432 | ‐ | 0.052 | ‐ |
| Slide Multi | 0.9 (0.3; 1.5) | 2.4 (1.9; 4) | 2.3 (1.5; 3.3) | 0.014 a | ‐ | 0.255 | ‐ |
0.033 a |
‐ |
Data are presented as the median (1st quartile; 3rd quartile). Mann–Whitney U tests compared performance of PD patients with HC. Wilcoxon signed ranks tests compared PD patients OFF and ON medication.
Group significant different at P < 0.050.
PD, Parkinson's disease; OFF, OFF medication; HC, healthy controls; ON, ON medication; VAS, visual analogue scale; r, effect size estimate.
FIG. 2.
Performance on tablet tasks. (A) Total single direction sliding time (ms). (B) Total multi‐direction sliding time (ms). (C) Transition time (ms) on the single sliding task. (D) Transition time (ms) on the multi‐direction sliding task.
Effect of Medication
The comparison between medication conditions did not reveal significant effects for either sliding task. Figure 2 and Table 2 show that both the performance and transition times improved following medication intake with a large effect size, although not significantly. Similarly, the number of correctly performed sliding movements on the multi‐direction sliding task worsened with a large effect size, though not significantly.
When comparing PD‐ON with HC, no significant differences were found in the performance or transition time for the single sliding task (Fig. 2 A−C). In the multi‐direction sliding task, PD‐ON had a significantly longer performance and transition time compared to HC (P = 0.032, r = − 0.530; P = 0.048, r = − 0.507; resp., Fig. 2 B−D). The error distance and the number of correctly performed sliding movements did not differ significantly on either sliding task.
Looking at the tapping task, medication led to a significant reduction in the inter‐tap interval time (P = 0.046, r = − 0.657). Despite the large effect size, medication conditions did not differ significantly in tapping accuracy. When compared to HC, PD‐ON revealed a similar timing performance, although they were less accurate (P = 0.028, r = − 0.490). Details are provided in Table 2. VAS values for all tablet tasks were similar between medication conditions (P > 0.200), yet PD‐ON patients reported higher VAS scores than HC (P < 0.100), see Table 2.
Correlation Analysis
A detailed overview of correlations between timing parameters of tablet tasks and general characteristics across groups (PD‐OFF and HC) are displayed in Table S1. Looking at the association with daily life smartphone skills, longer performance and transition times of all tablet tasks were correlated with longer performance times on the MPT (R > 0.400, P < 0.050, Fig. S1A–B), though not with the self‐reported MDPQ‐16 scores. A better manual dexterity, indicated by a lower score on the DEXTQ‐24 (R > 0.390, P < 0.060,Fig. S1C) and higher score on the Purdue Pegboard Test (R < − 0.350, P < 0.100), was associated with faster performance on all tablet tasks.
In PD‐OFF, a higher MDS‐UPDRS‐III score (i.e., worse disease severity) correlated with a longer inter‐tap interval time, total sliding time on the single sliding task as well as transition time on the multi‐direction sliding task (R > 0.600, P < 0.050). In contrast, patients with more severe right kinetic tremor (higher score on MDS‐UPDRS‐III item 16a) had faster total multi‐direction sliding times (R = − 0.583, P = 0.047). Lastly, higher HADS‐scores (i.e., worse mental wellbeing) were significantly correlated with slower performance (R > 0.450, P < 0.050). No correlations were found with the other characteristics.
Daily Smartphone Skills
The questionnaires on daily smartphone use (Table S2) revealed that HC owning a smartphone (92%) did not report problems with smartphone use, whereas the majority of the smartphone owning patients (100%) did experience problems (92%). The most commonly reported problem was the small size of the icons and the keyboard (45%), followed by difficulties with tapping (27%). Also, difficulties with swiping (18%), double tapping (9%), button use (9%) and enlarging an image by swiping over the screen (9%), were indicated. Interestingly, more PD patients mentioned to play games daily on the smartphone compared to HC (P = 0.047).
Table S3 provides sub‐scores of the MDPQ‐16, assessing the ability to perform different tasks on mobile devices. Lower scores indicate more difficulties, though the origin (motor or cognitive) is not specified. 22 The total score did not differ significantly between groups (P = 0.221). However, PD patients experienced more difficulties with the performance of basic skills (i.e., Mobile Device Basics), consisting of navigating through menus and using the keyboard, compared to HC (P = 0.019). Also, searching and finding information on the internet (i.e., Internet) and setting up passwords as well as deleting the search history (i.e., Privacy) appeared to be more difficult for PD patients than for HC (P < 0.050). As for performance on the MPT, PD patients were significantly slower in OFF than HC (P = 0.012, See Fig. S2).
Discussion
This pilot study aimed to examine the effects of PD and dopaminergic medication on touchscreen skills. We found a slower performance on most tablet tasks in PD‐OFF compared with HC, while accuracy did not differ between groups. After administration of dopaminergic medication, performance times of the tapping and single sliding tasks improved to comparable levels as HC. However, the complex multi‐direction sliding movements remained abnormal in PD.
PD Affects Timing Parameters
We investigated if and why PD patients report difficulties with the motor aspects of touchscreen manipulation and found that PD‐OFF performed sliding movements more slowly compared to HC in both single and multi‐direction sliding tasks. These slower timing parameters OFF medication partially support a lack of movement vigor or the presence of bradykinesia in PD. 27 Here, we showed for the first time that this symptom also affects the motor components of using a touchscreen device, particularly in the most difficult sliding task. Also, slower transition times were found, i.e., patients needed more time to grab the dot after terminating a sliding movement. Using a mobile device is not limited to actual movement performance, but also requires transitioning towards the next movement, pointing towards the complexity of touchscreen use. Although not recorded in the current study, the amount of pressure exerted on the touchscreen may also influence transition performance. 28 As such, daily use of touchscreen devices requires various complex skills, the exact coordination of which needs further research.
As for accuracy measures, PD‐OFF and HC did not differ in the number of correct sliding movements. These different results for timing and accuracy parameters might be explained by a difference in priority. PD‐OFF might have moved more slowly towards the fixed targets offered by the tablet tasks than HC, prioritizing accuracy over movement time. 29
Regarding the tapping task, we found no significant differences in inter‐tap interval time between PD‐OFF and HC, as opposed to the results of Wissel et al. 11 Patients in this latter study had worse disease severity compared to our study sample, suggesting that disease severity may be related to tapping performance. Our findings of a significant correlation between MDS‐UPDRS‐III score and inter‐tap interval time further support this. The differing findings also need to be interpreted against the correction for multiple testing applied in our study. Contrary to the timing, tapping accuracy was similar between groups in both studies. 11 Overall, the results of the tapping task further support the importance of comprehensive test batteries containing multiple tasks to identify specific problems with touchscreen use in PD.
Impact of Clinical Characteristics
Generally, tapping, sliding and transition times correlated positively with the more functional MPT and DEXTQ‐24 scores, indicating the relevance of the tasks for capturing the capacity of touchscreen manipulation. Together with the negative correlations with the Purdue Pegboard Test, it confirms that dexterous deficits affect touchscreen usage in PD, even more so when patients are OFF medication. 5 Correlations with MDS‐UPDRS‐III scores were found, indicating that patients with worse disease severity have a worse tablet task performance. Moreover, a worse right upper limb kinetic tremor was associated with a faster total multi‐direction sliding time. This is an intriguing result as we expected a correlation in the opposite direction, but tremor did not otherwise impact the findings. Apart from the stronger correlations with motor capacity, we also demonstrated an association between timing parameters and measures for mental wellbeing mainly for PD‐OFF, suggesting that the presence of depression may have affected the motivation for touchscreen performance. The high cognitive scores in the current study might explain the lack of significant correlations with executive function (TMT performance) and cognitive function in general. Therefore, future studies should consider participants with a broad spectrum of cognitive impairments to clarify the impact of cognitive function on touchscreen manipulation. 30
Partial Effects of Dopaminergic Medication
Medication improved performance times of the tapping and single sliding tasks to the level of HC, corroborating the known dopaminergic effects on bradykinetic symptoms and on tapping performance. 11 According to a recent review, dopaminergic medication increases the activity in the cortico‐subcortical network related to the invigoration of movements. 8 In contrast, transition times did not differ between medication conditions in both sliding tasks. It is likely that transition times reflected the ability to chunk motor components as it consisted of stopping the sliding movement, transitioning towards the dot, grabbing the dot and then starting the sliding movement again. One can thus expect that transition times capture not only motor function but also cognitive flexibility, suggesting that both motor and cognitive aspects were involved in the relatively simple touchscreen manipulations tested in this study. The discrepancy in medication effects on these timing parameters also underscores the need for the design of novel training interventions that target the varied aspects of touchscreen skills.
Looking at accuracy, no significant effects of anti‐parkinsonian medication were found for both sliding tasks. Tapping accuracy, on the contrary, was worse in PD‐ON compared to HC. This is in line with previous work showing that medication had a deleterious effect on movement accuracy, while improving movement time. 11
Overall, these findings imply that dopaminergic medication has a positive effect on the simpler aspects of touchscreen motions as opposed to the more complex sliding tasks. These results underscore that simple, repetitive tasks, underestimate the problems with touchscreen manipulation in PD. Although simple assessments are most frequently used, future research should include more comprehensive test batteries revealing the complex reality of touchscreen use.
Medication administration resulted in similar performance levels as HC, as differences in the single sliding task and in inter‐tap interval time between patients and HC were no longer significant. This is an important finding as it can be assumed that most patients use their mobile devices while ON medication in daily life. However, medication did not ameliorate all aspects. These partial effects of dopaminergic medication could also explain the many self‐reported difficulties with using mobile devices by patients, probably further exacerbated by the distraction experienced in daily life.
Clinical Implications and Future Research
Previous research showed that patients who experienced more difficulties with technology, might dropout of studies using smartphone interventions. 31 This suggests the need for specific training programs to address these problems, making sure that transfer is addressed to daily life. Such interventions could be delivered in the home setting, shown to be successful for improving dexterity 32 as well as micrographia in PD. 33 Furthermore, we recently demonstrated benefits of short‐term learning of unlocking a touchscreen trace. 7 Importantly, all these training programs proved feasible without much supervision yet high adherence, 34 suggesting a cost‐effective approach for tackling touchscreen deficiencies in the future.
The slower performance on the tablet tasks together with the self‐reported difficulties experienced by PD patients, suggest the need for thoughtful development of smartphone applications for patients. Considering the more pronounced problems with the complex aspects of tablet tasks, more simple handling of E‐health applications should be provided, e.g. by avoiding multidirectional movements or double tapping. Nunes et al. 5 provided preliminary guidelines for such developments, needing further validation.
Study Limitations
Several limitations should be considered when interpreting our findings. First, the small sample size may have increased the risk of type II errors, which may have underestimated the medication effects on movement accuracy. We used the Bonferroni method to correct for multiple testing, ensuring an overall conservative approach to our statistical analysis. The small sample size also prevents generalization to the broader PD population. Additionally, we focused on the motor aspects of touchscreen use supporting the need for future research implementing additional cognitive load. The ON tests were performed after the OFF tests possibly resulting in order effects, although we did not find differences in VAS scores for fatigue. In addition, and despite some familiarization after trial 1, we showed acceptable reliability of repeated trials.
Conclusion
Overall, we found that PD patients had poorer touchscreen skills compared to age‐matched healthy controls, especially when performing multi‐direction sliding movements and when capturing a target. Some of the milder difficulties were alleviated with dopaminergic treatment, but the more complex tasks remained below the levels of healthy controls. These findings underscore the message that efficient utilization of mobile devices should not be assumed in PD. Therefore, future research is needed to investigate the developments of E‐health applications and novel neurorehabilitation programs, which are tailor‐made to ensure that people with PD can partake optimally in society.
Author Roles
(1) Research project: A. Conception, B. Organization, C. Execution; (2) Statistical analyses: A. Design; B. Execution, C. Review and Critique; (3) Manuscript preparation: A. Writing of the first draft, B. Review and Critique.
J.D.V.: 1C, 2A, 2B, 3A
S.B.: 2C, 3B
L.J.: 1A, 3B
E.H.: 1A, 3B
A.N.: 1A, 2C, 3B
E.N.: 1A, 1B, 1C, 2A, 2C, 3A, 3C
Disclosures
Ethical Compliance Statement: This study was approved by the local Ethics committee UZ/KU Leuven according to the code of Ethics of the World Medical Association (Declaration of Helsinki, version 2013, S61793). Prior to participation in the study, an informed consent form was signed after explanation of the study protocol. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this work is consistent with those guidelines.
Funding Sources and Conflict of Interest: This project was supported by the Research Foundation Flanders–FWO [grant numbers: 12F4719N, 1520619 N, G0A5619N] and King Baudouin Foundation [J1811020‐E003]. The authors declare that there are no conflicts of interest relevant to this work.
Financial Disclosures for the Previous 12 Months: EN is a postdoctoral fellow funded by the Research Foundation Flanders–FWO [grant number 12F4719N]. SB is a doctoral fellow funded by the Research Foundation Flanders–FWO [grant number 1167419N].
Supporting information
Supplementary Material S1 Repeatability analysis
Supplementary Table S1. Correlation analysis across groups and in both groups separately.
Supplementary Table S2. Smartphone specific questions. Descriptive use of mobile devices
Supplementary Table S3. Mobile Device Proficiency Questionnaire (MDPQ‐16)
Supplementary Figure S1 Correlations between performance on tablet tasks and clinical characteristics across groups (PD‐OFF and HC). (A) Total sliding time (ms) on the Single sliding task and MPT performance (s). (B) Transition time (ms) on the Multi‐direction sliding task with MPT performance (s). (C) Transition time on the Multi‐direction sliding task with scores on the dexterity questionnaire (DEXTQ‐24). Filled circles = PD‐OFF patients; unfilled circles = healthy controls.
Supplementary Figure S2. Comparison performance on mobile phone task (MPT) between groups. Mann–Whitney U test compared patients with healthy controls. Differences are indicated by square brackets. Wilcoxon Signed Ranks test compared medication conditions. * Group differences at P < 0.050; # Group differences at P < 0.100.
Supplementary Appendix. Smartphone specific questions (translated to English)
Acknowledgments
We are grateful for all the patients and controls to participate voluntarily in this study. We also want to thank Prof. Wim Vandenberghe from University Hospitals Leuven for his role in patient recruitment.
Relevant disclosures and conflicts of interest are listed at the end of this article.
References
- 1. Jankovic J. Parkinson's disease: clinical features and diagnosis. J Neurol Neurosurg Psychiatry 2008;79(4):368–376. 10.1136/jnnp.2007.131045. [DOI] [PubMed] [Google Scholar]
- 2. Gebhardt A, Vanbellingen T, Baronti F, Kersten B, Bohlhalter S. Poor dopaminergic response of impaired dexterity in Parkinson's disease: Bradykinesia or limb kinetic apraxia? Mov Disord 2008;23(12):1701–1706. 10.1002/mds.22199. [DOI] [PubMed] [Google Scholar]
- 3. Rovini E, Maremmani C, Cavallo F. Automated systems based on wearable sensors for the management of Parkinson's disease at home: a systematic review. Telemed e‐Health 2019;25(3):167–183. 10.1089/tmj.2018.0035. [DOI] [PubMed] [Google Scholar]
- 4. Lee W, Evans A, Williams DR. Validation of a smartphone application measuring motor function in Parkinson's disease. J Parkinsons Dis 2016;6(2):371–382. 10.3233/JPD-150708. [DOI] [PubMed] [Google Scholar]
- 5. Nunes F, Silva PA, Cevada J, Correia Barros A, Teixeira L. User interface design guidelines for smartphone applications for people with Parkinson's disease. Univers Access Inf Soc 2016;15(4):659–679. 10.1007/s10209-015-0440-1. [DOI] [Google Scholar]
- 6. Lee CY, Kang SJ, Hong SK, Ma HI, Lee U, Kim YJ. A validation study of a smartphone‐based finger tapping application for quantitative assessment of bradykinesia in Parkinson's disease. PLoS One 2016;11(7):1–11. 10.1371/journal.pone.0158852. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Nackaerts E, Ginis P, Heremans E, Swinnen SP, Vandenberghe W, Nieuwboer A. Retention of touchscreen skills is compromised in Parkinson's disease. Behav Brain Res 2020;378(August 2019):112265. 10.1016/j.bbr.2019.112265. [DOI] [PubMed] [Google Scholar]
- 8. Meder D, Herz DM, Rowe JB, Lehéricy S, Siebner HR. The role of dopamine in the brain ‐ lessons learned from Parkinson's disease. Neuroimage 2019;190(May 2018):79–93. 10.1016/j.neuroimage.2018.11.021. [DOI] [PubMed] [Google Scholar]
- 9. Fernandez L, Huys R, Issartel J, Azulay JP, Eusebio A. Movement speed‐accuracy trade‐off in Parkinson's disease. Front Neurol 2018;9(897):1–13. 10.3389/fneur.2018.00897. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Levy‐Tzedek S, Krebs HI, Arle JE, Shils JL, Poizner H. Rhythmic movement in Parkinson's disease: effects of visual feedback and medication state. Exp Brain Res 2011;211(2):277–286. 10.1007/s00221-011-2685-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Wissel BD, Mitsi G, Dwivedi AK, et al. Tablet‐based application for objective measurement of motor fluctuations in Parkinson disease. Digit Biomarkers 2018;1(2):126–135. 10.1159/000485468. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Hoehn M, Yahr M. Parkinsonism: onset, progression and mortality. Neurology 1967;17(5):427–442. 10.1212/wnl.17.5.427. [DOI] [PubMed] [Google Scholar]
- 13. Hughes AJ, Daniel SE, Kilford L, Lees AJ. Accuracy of clinical diagnosis of idiopathic Parkinson's disease: a clinico‐pathological study of 100 cases. J Neurol Neurosurg Psychiatry 1992;55(3):181–184. 10.1136/jnnp.55.3.181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 1971;9:97–113. 10.1016/0028-3932(71)90067-4. [DOI] [PubMed] [Google Scholar]
- 15. Folstein MF, Folstein SE, McHugh PR. “Mini‐mental state” a practical method for grading the cognitive state of patients for the clinician. J Psychiat Res 1975;12:189–198. [DOI] [PubMed] [Google Scholar]
- 16. Goetz CG, Tilley BC, Shaftman SR, et al. Movement Disorder Society‐sponsored revision of the unified Parkinson's disease rating scale (MDS‐UPDRS): scale presentation and clinimetric testing results. Mov Disord 2008;23(15):2129–2170. 10.1002/mds.22340. [DOI] [PubMed] [Google Scholar]
- 17. Tiffin J, Asher EJ. The Purdue pegboard: norms and studies of reliability and validity. J Appl Psychol 1948;32(3):234–247. 10.1037/h0061266. [DOI] [PubMed] [Google Scholar]
- 18. Vanbellingen T, Nyffeler T, Nef T, Kwakkel G, Bohlhalter S, van Wegen EEH. Reliability and validity of a new dexterity questionnaire (DextQ‐24) in Parkinson's disease. Park Relat Disord. 2016;33:78–83. 10.1016/j.parkreldis.2016.09.015. [DOI] [PubMed] [Google Scholar]
- 19. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983;67:361–370. 10.1111/j.1600-0447.1983.tb09716.x. [DOI] [PubMed] [Google Scholar]
- 20. Nasreddine ZS, Philips NA, Bédirian V, et al. The Montreal cognitive assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc 2005;53:695–699. [DOI] [PubMed] [Google Scholar]
- 21. Llinàs‐Reglà J, Vilalta‐Franch J, López‐Pousa S, Calvó‐Perxas L, Torrents Rodas D, Garre‐Olmo J. The trail making test: association with other neuropsychological measures and normative values for adults aged 55 years and older from a Spanish‐speaking population‐based sample. Assessment 2017;24(2):183–196. 10.1177/1073191115602552. [DOI] [PubMed] [Google Scholar]
- 22. Roque NA, Boot WR. A new tool for assessing Mobile device proficiency in older adults: the mobile device proficiency questionnaire. J Appl Gerontol 2018;37(2):131–156. 10.1177/0733464816642582. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Tomlinson CL, Stowe R, Patel S, Rick C, Gray R, Clarke CE. Systematic review of levodopa dose equivalency reporting in Parkinson's disease. Mov Disord 2010;25(15):2649–2653. 10.1002/mds.23429. [DOI] [PubMed] [Google Scholar]
- 24. Schade S, Mollenhauer B, Trenkwalder C. Levodopa equivalent dose conversion factors: an updated proposal including opicapone and safinamide. Mov Disord Clin Pract. 2020;7(3):1–3. 10.1002/mdc3.12921. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. New Jersey, USA: Lawrence Erlbaum Associates; 1988. [Google Scholar]
- 26. Fritz CO, Morris PE, Richler JJ. Effect size estimates: current use, calculations, and interpretation. J Exp Psychol Gen 2012;141(1):2–18. 10.1037/a0024338. [DOI] [PubMed] [Google Scholar]
- 27. Mazzoni P, Hristova A, Krakauer JW. Why don't we move faster? Parkinson's disease, movement vigor, and implicit motivation. J Neurosci 2007;27(27):7105–7116. 10.1523/JNEUROSCI.0264-07.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Drotár P, Mekyska J, Rektorová I, Masarová L, Smékal Z, Faundez‐Zanuy M. Evaluation of handwriting kinematics and pressure for differential diagnosis of Parkinson's disease. Artif Intell Med 2016;67:39–46. 10.1016/j.artmed.2016.01.004. [DOI] [PubMed] [Google Scholar]
- 29. Rand MK, Stelmach GE, Bloedel JR. Movement accuracy constraints in Parkinson's disease patients. Neuropsychologia 2000;38(2):203–212. 10.1016/S0028-3932(99)00059-7. [DOI] [PubMed] [Google Scholar]
- 30. Kudlicka A, Clare L, Hindle JV. Executive functions in Parkinson's disease: systematic review and meta‐analysis. Mov Disord 2011;26(13):2305–2315. 10.1002/mds.23868. [DOI] [PubMed] [Google Scholar]
- 31. Ginis P, Nieuwboer A, Dorfman M, et al. Feasibility and effects of home‐based smartphone‐delivered automated feedback training for gait in people with Parkinson's disease: a pilot randomized controlled trial. Park Relat Disord 2016;22:28–34. 10.1016/j.parkreldis.2015.11.004. [DOI] [PubMed] [Google Scholar]
- 32. Vanbellingen T, Nyffeler T, Nigg J, et al. Home based training for dexterity in Parkinson's disease: a randomized controlled trial. Park Relat Disord. 2017;41:92–98. 10.1016/j.parkreldis.2017.05.021. [DOI] [PubMed] [Google Scholar]
- 33. Nackaerts E, Heremans E, Vervoort G, et al. Relearning of writing skills in Parkinson's disease after intensive amplitude training. Mov Disord 2016;31(8):1209–1216. 10.1002/mds.26565. [DOI] [PubMed] [Google Scholar]
- 34. Vanbellingen T, van Beek J, Nyffeler T, Urwyler P, Nef T, Bohlhalter S. Tablet app‐based dexterity‐training in patients with Parkinson's disease: pilot feasibility study. Ann Phys Rehabil Med 2020. 10.1016/j.rehab.2020.06.004. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Material S1 Repeatability analysis
Supplementary Table S1. Correlation analysis across groups and in both groups separately.
Supplementary Table S2. Smartphone specific questions. Descriptive use of mobile devices
Supplementary Table S3. Mobile Device Proficiency Questionnaire (MDPQ‐16)
Supplementary Figure S1 Correlations between performance on tablet tasks and clinical characteristics across groups (PD‐OFF and HC). (A) Total sliding time (ms) on the Single sliding task and MPT performance (s). (B) Transition time (ms) on the Multi‐direction sliding task with MPT performance (s). (C) Transition time on the Multi‐direction sliding task with scores on the dexterity questionnaire (DEXTQ‐24). Filled circles = PD‐OFF patients; unfilled circles = healthy controls.
Supplementary Figure S2. Comparison performance on mobile phone task (MPT) between groups. Mann–Whitney U test compared patients with healthy controls. Differences are indicated by square brackets. Wilcoxon Signed Ranks test compared medication conditions. * Group differences at P < 0.050; # Group differences at P < 0.100.
Supplementary Appendix. Smartphone specific questions (translated to English)