The Left Right Judgement Task (LRJT) involves determining if an image of the body part is of the left or right side. The LRJT has been utilized as part of rehabilitation treatment programs for persons with pain associated with musculoskeletal injuries and conditions. Although studies often attribute changes and improvement in LRJT performance to an altered body schema, imaging studies suggest that the LRJT implicates other cortical regions. We hypothesized that cognitive factors would be related to LRJT performance of hands and feet and that sensory, motor, and pain related factors would be related to LRJT in the affected hand of participants with wrist/hand pain. In an observational cross-sectional study, sixty-one participants with wrist/hand pain participated in a study assessing motor imagery ability, cognitive (Stroop test), sensory (Two-Point Orientation Discrimination, pressure pain thresholds), motor (grip strength, Purdue Pegboard Test), and pain related measures (West Haven Yale Multidimensional Pain Inventory) as well as disability (Disability of the Arm, Shoulder and Hand). Multiple linear regression found Stroop test time and motor imagery ability to be related to LRJT performance. Tactile acuity, motor performance, participation in general activities, and the taking of pain medications were predictors of LRJT accuracy in the affected hand. Participants who took pain medications performed poorly in both LRJT accuracy (p=0.001) and reaction time of the affected hand (p=0.009). These participants had poorer cognitive (p=0.013) and motor function (p=0.002), and higher pain severity scores (p=0.010). The results suggest that the LRJT is a complex mental task that involves cognitive, sensory, motor, and behavioural processes. Differences between persons with and without pain and improvement in LRJT performance may be attributed to any of these factors and should be considered in rehabilitation research and practice utilizing this task.
The Left Right Judgement Task (LRJT) involves determining, as accurately and as quickly as possible, if an image of a body part is of the left or right side. LRJT performance differences between persons with and without pain have been hypothesized to reflect changes in central nervous system processing, errors in judgement, and changes in bodily representations [
Studies involving the LRJT have been performed with persons experiencing pain associated with musculoskeletal injuries and conditions with variable findings. These include no changes in LRJT performance [
Imaging studies demonstrate that the LRJT is a complex mental task associated with activation of subcortical and cortical structures including frontal areas involved in working memory and attention, pre-motor areas, basal ganglia, cerebellum, and sensory integrative areas in the parietal cortex [
Studies with persons experiencing pain associated with musculoskeletal injuries and conditions demonstrate changes in peripheral [
The objective of the study was therefore to determine which of the cognitive, sensory, motor, and pain related factors were associated with LRJT performance in symptomatic participants with pain associated with musculoskeletal injuries and conditions of the wrist/hand. We hypothesized that motor imagery ability and cognitive aspects assessed with the Stroop test would be associated with LRJT performance of images of the hands and feet. We also hypothesized that sensory, motor performance, and pain related factors would be more specifically related to images congruent with the area of pain. A better understanding of the factors associated with LRJT performance provides valuable information into the variability of study results and the necessity to consider sensory, motor, cognitive, and even behavioural factors in research and practice involving the LRJT.
This was an observational cross-sectional study. The protocol and procedures conformed to the Declaration of Helsinki. The study was conducted at the Hand Clinic at the Centre Hospitalier de l’Université de Montréal, Notre Dame Hospital between June and December 2017. Ethical approval was granted from the institutional review board (CÉR-CHUM 16.372). Participants for the study were recruited when attending the hand clinic for consultation with plastic surgeons specialising in wrist and hand disorders. Participants were screened in the waiting area to explain the nature of the study, the requirements for their participation, and eligibility. Participants were required to be 18 years and older, experiencing pain associated with musculoskeletal injuries and conditions of the wrist/hand in their right dominant side that impacted their activities of daily living, were able to follow instructions and answer questionnaires in English or French, and suffer from no known neurological condition that impacted cognitive function and no musculoskeletal injuries and conditions of the lower extremities. Verbal and written informed consent was obtained prior to the commencement of the study. Demographic and descriptive information including gender, age, education, diagnosis, symptom duration, areas of pain, and taking of pain medications was documented. Handedness was verified utilizing the Edinburgh Handedness Inventory [
The LRJT involved determining if images of hands and feet were of the left or right side utilizing the Recognise™ (Neuro-Orthopedic Institute, Adelaide, South Australia) software [
The West Haven Yale Multidimensional Pain Inventory (MPI) [
Pressure Pain Threshold (PPT) was determined by using a digital pressure algometry (Wagner Instruments, Greenwich, CT, USA, model# Wagner FPX25). PPT was measured bilaterally on the palmer aspect of the first carpometacarpal joint and the hypothenar eminence lateral to the pisiform. The average of three trials was recorded [
Tactile acuity was assessed with the Two-Point Orientation Discrimination (TPOD) task utilizing a hand-held caliper (Fowler, Model # 54-101-150-2, Newton, MA, USA) [
Proprioception was measured by evaluating Joint Position Sense (JPS). JPS was performed in the same manner as described by Kalisch et al. (2012) where subjects were blindfolded and instructed to compare sizes of two polystyrene balls of different diameters placed in their hands. Three different diameter polystyrene reference balls (7.0, 8.0, and 9.6 cm diameter) were placed in the participant’s left hand by the examiner. A second polystyrene ball, of seven possible different diameters (6.6, 7.0, 7.3, 8.0, 9.0, 9.6, and 10 cm diameter), was placed in the right (affected) hand. Participants were instructed to squeeze the polystyrene balls and then relax the tension to control for thixotropy effects influencing JPS [
Motor performance was assessed by dynamometric evaluation of strength performed utilizing a hand-held Jamar dynamometer (Sammons Preston Rolyan, Bolingbrook, IL, USA) following recommended protocols [
Fine and gross motor function was assessed with the Purdue Pegboard Test (PPG) (Lafayette Instruments, Lafayette IN, USA, Model #32020A), a standard manual dexterity test commonly utilized in research and in clinical settings that involves placing pins in slots with their right hand, left hand, and both hands in 30-second time epochs. A total score consists of the aggregate of these three measures. Finally, participants perform the building of small assemblies involving pins, washers, and collars in a one-minute epoch. The PPG has been assessed for reliability and validity [
Disability of the Arm, Shoulder and Hand questionnaire (DASH) was utilized to assess both symptoms and functional status in patients with upper extremity musculoskeletal injuries and conditions. It is a self-rated assessment with documented construct validity and reliability [
Studies of the LRJT allude to attention/concentration and motor imagery ability as possible confounding factors explaining experimental results in LRJT studies [
Sample size was predetermined based upon an
Statistical analysis was performed utilizing GraphPad Prism 7 (GraphPad Software Inc, La Jolla, CA, USA) and SPSS 24 (IBM Corporation, Armonk, New York, USA) statistical software. Normality of data was assessed by visual inspection of the data and D’Agostino Pearson normality test.
Differences between LRJT performance measures between hands and between feet were performed utilizing paired T-tests. Pearson correlation coefficients were performed between LRJT performance (Accuracy and RT) and the independent variables. Adjustments for multiple comparisons were made when necessary using the False Discovery Rate Benjamini-Hochberg procedure with an
Multiple Linear Regression models were performed for each of the dependent variables (LRJT accuracy and LRJT RT for the hands and feet) with the sensory, motor, and cognitive measures. Variables that have previously been found to be related to LRJT performance in some studies such as age, pain severity, symptom duration, motor imagery ability, and concentration/selective attention (Stroop test) and inserting different permutations of the independent variables were entered into the multiple linear regression models, the choice influenced by correlation coefficients values and relevance. Choice of best model and which variables to maintain was based upon minimizing of the mean squared error, including independent variables where the coefficients had p values below p=0.10 and had the highest R and R2 adjusted values. Models were checked for multicollinearity and homoscedasticity.
As pain medication was a strong and significant predictor in the multiple linear regression model for LRJT performance accuracy, the participants were divided into two groups, those taking pain medication (PainMeds) and those who had not taken pain medication (NoPainMeds). Nonparametric tests were performed on demographic, pain, and disability measures between groups. Paired comparisons were performed for LRJT Accuracy and RT between these groups. As some LRJT performance data violated the assumptions of homogeneity and equality of variance, Mann–Whitney U nonparametric tests were performed.
Sixty-one subjects participated in the experiment (31♂, 30♀). Participants experienced pain associated with musculoskeletal injuries and conditions of the wrist and hand including postoperative fractures/amputation, tendinitis, first carpometacarpal osteoarthritis, Dupruytren’s, trigger finger, and wrist sprains. Descriptive information is found in Table
Participant descriptive information.
Mean | Standard Deviation | |
---|---|---|
Age (years) | 55.82 | 13.57 |
|
||
Symptom Duration (months) | 43.68 | 45.79 |
|
||
West Haven Yale Multidimensional Pain Inventory (max scores – 6) | ||
|
||
Pain Severity | 3.09 | 1.17 |
|
||
Pain Interference | 3.11 | 1.38 |
|
||
Life Control | 3.88 | 1.24 |
|
||
Affective Distress | 2.79 | 1.27 |
|
||
General Activities | 2.69 | 0.95 |
|
||
Disability of Arm, Shoulder and Hand (DASH) | 42.98 | 17.62 |
|
||
Pressure Pain Thresholds (kg) | ||
|
||
Right (affected) Hand | 6.93 | 3.83 |
|
||
Left (unaffected) Hand | 8.42 | 4.89 |
|
||
Two Point Discrimination (mm) | ||
|
||
Right (affected) Hand | 10.92 | 2.89 |
|
||
Left (unaffected) Hand | 10.16 | 2.72 |
|
||
Joint Position Sense (errors) | 3.82 | 1.51 |
|
||
Purdue Pegboard Scores | ||
|
||
Right (affected) Hand | 12.44 | 3.48 |
|
||
Left (unaffected) Hand | 13.20 | 2.29 |
|
||
Both Hands | 10.43 | 3.30 |
|
||
Total | 35.31 | 9.50 |
|
||
Assemblies | 21.89 | 7.86 |
|
||
Grip Strength (kg) | ||
|
||
Right (affected) Hand | 23.60 | 13.35 |
|
||
Left (unaffected) Hand | 30.65 | 14.20 |
|
||
Stroop Time (seconds) | 37.25 | 7.99 |
|
||
Motor Imagery Questionnaire (maximum score - 98) | 67.34 | 23.94 |
No difference was found in LRJT accuracy or RT between hands and between feet (see Figure
Left Right Judgement Task (LRJT) performance in participants with musculoskeletal disorders of the wrist/hand. Mean±95% Confidence Intervals.
The best fitting MLR model (F2,56=4.11, p=0.002) included pain medication, MPI General Activities, Two-Point Orientation Discrimination of the Right Hypothenar, and Purdue Pegboard values of the left hand and, after entering Stroop test and motor imagery ability scores, accounted for an additional 20% of explained variance (R2 adjusted) (see Tables
Multiple linear regression models for Left Right Judgement Task Accuracy for the hands.
|
|
|||||||||
---|---|---|---|---|---|---|---|---|---|---|
|
|
|
|
|
|
|
|
|
||
Right (affected) Hand | ||||||||||
|
||||||||||
1 | 0.26 | 0.07 | 0.04 | 13.09 | 0.07 | 2.105 | 2 | 56 | 0.131 | |
|
||||||||||
2 | 0.57 | 0.32 | 0.24 | 11.61 | 0.25 | 4.819 | 4 | 52 | 0.002 | |
|
||||||||||
Left (unaffected) Hand | ||||||||||
|
||||||||||
3 | 0.46 | 0.21 | 0.18 | 15.67 | 0.21 | 7.353 | 2 | 56 | 0.001 | |
|
||||||||||
4 | 0.55 | 0.30 | 0.25 | 15.03 | 0.09 | 3.430 | 2 | 54 | 0.040 |
1. Predictors: (Constant), Motor Imager Questionnaire Visual Motor Imagery, Stroop Time.
2. Predictors: (Constant), Motor Imager Questionnaire Visual Motor Imagery, Stroop Time, Pain Medications, MPI General Activities, two-point orientation discrimination hypothenar right hand, Purdue Pegboard Test left hand.
3. Predictors: (Constant), Motor Imager Questionnaire Visual Motor Imagery, Stroop Time.
4. Predictors: (Constant), Motor Imager Questionnaire Visual Motor Imagery, Stroop Time, MPI general activities, Purdue Pegboard Test left hand.
Coefficients of best fitting MLR LRJT Right Hand Accuracy.
LRJT Accuracy | Confidence Intervals | |||||||
---|---|---|---|---|---|---|---|---|
Unstandardized Coefficients |
Standard Deviation | Standardized Coefficients |
t | p | Lower |
Upper | ||
|
||||||||
|
||||||||
|
|
69.75 | 13.49 | 5.169 | 0.000 | 42.71 | 96.78 | |
|
||||||||
|
-0.06 | 0.24 | -0.04 | -0.252 | 0.802 | -0.55 | 0.43 | |
|
||||||||
|
0.29 | 0.17 | 0.25 | 1.686 | 0.097 | -0.05 | 0.63 | |
|
||||||||
|
|
53.61 | 18.70 | 2.867 | 0.006 | 16.09 | 91.13 | |
|
||||||||
|
0.24 | 0.23 | 0.15 | 1.039 | 0.304 | -0.27 | 0.71 | |
|
||||||||
|
0.07 | 0.17 | 0.06 | 0.381 | 0.705 | -0.28 | 0.40 | |
|
||||||||
|
-8.34 | 4.14 | -0.26 | -2.017 | 0.049 | -16.64 | -0.04 | |
|
||||||||
|
3.36 | 1.85 | 0.24 | 1.818 | 0.075 | -0.35 | 7.07 | |
|
||||||||
|
-1.12 | 0.57 | -0.24 | -1.965 | 0.055 | -2.26 | 0.02 | |
|
||||||||
|
1.39 | 0.73 | 0.24 | 1.815 | 0.075 | -0.15 | 2.92 | |
|
||||||||
|
||||||||
|
||||||||
|
|
74.70 | 16.15 | 4.625 | 0.000 | 42.35 | 107.06 | |
|
||||||||
|
-0.470 | 0.29 | -0.22 | -1.613 | 0.112 | -1.05 | 0.11 | |
|
||||||||
|
0.48 | 0.20 | 0.31 | 2.331 | 0.023 | 0.07 | 0.88 | |
|
||||||||
|
|
43.43 | 21.19 | 2.050 | 0.045 | 0.96 | 85.91 | |
|
||||||||
|
-0.34 | 0.29 | -0.16 | -1.183 | 0.242 | -0.92 | 0.24 | |
|
||||||||
|
0.24 | 0.22 | 0.15 | 1.084 | 0.283 | -0.20 | 0.67 | |
|
||||||||
|
4.73 | 2.33 | 0.26 | 2.030 | 0.047 | 0.06 | 9.40 | |
|
||||||||
|
1.74 | 0.94 | 0.23 | 1.844 | 0.071 | -0.15 | 3.62 |
MIQ VMI: Motor Imagery Questionnaire–Visual Motor Imagery; MPI: West Haven Yale Multidimensional Pain Inventory; TPOD: Two-Point Orientation Discrimination.
The best fitting MLR model (F4,54=5.71, p=0.001) included pain medication, MPI General Activities, and Purdue Pegboard values of the left hand and after entering Stroop Test and Motor Imagery Ability scores accounted for only an additional 4% of explained variance (R2 adjusted) (see Tables
The best fitting MLR model for LRJT Right Hand RT (F2,56=4.42, p=0.017) included only the variables Stroop Time and Gender (see Tables
Best fitting multiple linear regression LRJT Right Hand Reaction Time.
|
|||||||||
---|---|---|---|---|---|---|---|---|---|
|
|
|
|
|
|
|
|
|
|
1 | 0.41 | 0.17 | 0.14 | 0.63 | 0.17 | 5.738 | 2 | 56 | 0.005 |
1. Predictors: (Constant), Gender, Stroop Time.
Coefficients of best fitting MLR LRJT Right Hand Reaction Time.
|
|||||||
---|---|---|---|---|---|---|---|
|
|
|
|
|
|
|
|
(Constant) | 0.74 | 0.43 | 1.726 | 0.090 | -0.13 | 1.59 | |
|
|||||||
Gender | 0.33 | 0.17 | 0.24 | 1.961 | 0.055 | -0.01 | 0.66 |
|
|||||||
Stroop Time | 0.03 | 0.01 | 0.39 | 3.117 | 0.003 | 0.01 | 0.05 |
No statistically significant model could be produced with LRJT Left Hand RT entered as the dependent variable.
Multiple linear regression models using LRJT Feet Accuracy and RT as the dependent variables explained (R2 adjusted) 27-35% of the variance and the Stroop time and MIQ-VMI scores accounted for 78% and 86% of the explained variance of these models.
LRJT performance was compared for the data of two groups, those who took pain medication (PainMeds) (n=13) (10 participants: acetaminophen, 2 participants: Lyrica, and 1 participant: Tramadol) on the day of the evaluation and those who did not (NoPainMeds) (n=48). A difference in LRJT accuracy between the two groups was found for the right hand (see Table
Nonparametric test results for Left Right Judgement Task performance accuracy and reaction time in participants who were taking pain medication and not taking pain medication on the day of the evaluation.
|
|
|||||||
---|---|---|---|---|---|---|---|---|
|
|
|
|
|
||||
|
|
|
|
|
|
|
|
|
|
0.001 |
0.003 |
0.191 | 0.009 |
0.066 | 0.880 | 0.730 | 0.092 |
|
||||||||
|
125.5 | 143.5 | 386.0 | 460.0 | 208.5 | 303.5 | 292.5 | 407.5 |
|
||||||||
|
-3.31 | -2.99 | 1.31 | 2.61 | -1.84 | -0.15 | -0.35 | 1.69 |
Left Right Judgement performance accuracy and reaction time in participants who were taking pain medication (PainMeds) and not taking pain medication (NoPainMeds) on the day of the evaluation.
LRJT hands accuracy
LRJT feet accuracy
LRJT hands reaction time
LRJT feet reaction time
There was no difference in age, gender, or symptom duration between these two groups. After controlling for multiple comparisons, motor functions and Stroop test times were significantly different between groups. Purdue Pegboard Both Hands score was lower in the PainMed group (
Several variables including DASH scores and having constant pain demonstrated trends for differences between groups but were not significant after controlling for multiple comparison tests performed. Self-reported disability DASH scores were higher in the PainMeds group (
The LRJT is utilized in rehabilitation as a method of treatment. However, there is variability in study results evaluating LRJT in participants with pain associated with musculoskeletal injuries and conditions and clinical measures associated with LRJT have not been investigated. We hypothesized that LRJT performance would be related to cognitive factors. We found that motor imagery ability and a measure of cognitive function, the Stroop test scores, explained a significant portion of the explained variance in the linear regression models of the LRJT. Secondly, we hypothesized that sensory, motor, and pain related factors would be specifically associated with the presentation of images of the right affected hand and would explain the majority of the variance in the model. Importantly, sensory and motor processes, the taking of pain medication, and participation in social, work, leisure, and household activities were responsible for 86% of the explained variance in the linear regression model for LRJT accuracy in the right, affected hand only. Novel and unexpected findings are that participants who indicated that they had taken pain medication on the day of the evaluation performed more poorly in the LRJT and that activities and participation were positively associated with better LRJT performance in the affected hand only.
The LRJT is believed to involve implicit motor imagery where the participant makes an initial impression of laterality and then mentally imagines moving their hand in the same position as the image, and then either confirming or rejecting their initial impression of laterality [
The belief that the LRJT involves implicit motor imagery is based upon at least two experimental findings. Imaging studies involving the LRJT demonstrate a similar pattern of activation as motor imagery [
The ability to perform the LRJT also requires complex mental processes. This is supported by imaging studies that demonstrate the activation of distributed cortical structures including those involved in working memory/attention such as the dorsolateral prefrontal cortex [
Although the majority of studies utilizing the LRJT have not controlled for motor imagery ability and cognitive factors such as concentration/attention, the present results suggest that such control is necessary when attempting to understand the different processes involved in LRJT performance including improvement in the task and differences between groups with and without pain.
A measure of sensory function, TPOD, was also included in the linear regression model of LRJT performance accuracy in the right affected hand only. Stanton et al. (2013) previously found a correlation between two-point discrimination thresholds and LRJT accuracy in participants with back pain, but not in subjects with knee osteoarthritis [
In a previous study we found a stronger relationship between LRJT performance and Purdue Pegboard Test scores in the healthy control group [
Two interesting findings were the inclusion of Pain Medications and MPI General Activities subscale in the linear regression models for LRJT right (affected) hand accuracy. The regression model and subsequent nonparametric tests found that participants who reported taking pain medication on the day of assessment performed more poorly on the LRJT Hand accuracy. It is possible that taking the pain medication was simply a function of increased pain scores and that pain severity is associated with the poorer LRJT performance. However, the link between pain severity and LRJT performance is unclear with several studies finding no association [
The participants who took pain medications demonstrated several differences with the participants who had not taken medication. Participants who took pain medication had greater pain severity, poorer motor function, and Stroop test scores and describing their pain as constant was close to statistical significance. Differences between nociceptive and neuropathic pain on central nervous system changes have previously been attributed to the differences between these two types of pain and the belief that neuropathic pain is more constant and unrelenting [
LRJT Hand accuracy performance was also positively correlated with the MPI subset of general activities. This subset is comprised of 18 questions related to the participation in household, work, leisure, and outdoor activities. There was no correlation between MPI General Activities and pain measures. Although speculative, increased activities and participation may help to maintain the integrity of the Body Schema of the injured area through use. Another possible explanation is that participants involved in greater activities and participation have higher self-efficacy. Self-efficacy is defined as the confidence in performing/managing a particular behaviour and in overcoming barriers [
All experiments were performed in a single setting. The study included participants who were experiencing pain associated with musculoskeletal injuries and conditions of the right dominant hand and may not be generalizable for the left hand. The participants who had taken pain medication was a small sample and a larger study would help to confirm these results. Although the TPOD is a more rigorous method of evaluation as it decreases the nonspatial cues that are associated with the two-point discrimination task, it has not been assessed for reliability and therefore results involving this measure of tactile acuity should be interpreted with caution. The adjusted R2 values for the multiple regression models did not explain the majority of the variance and therefore other variables are also implicated in the LRJT performance not included in the models.
The study has several important implications for rehabilitative research and practice involving the LRJT. The LRJT appears to be a multidimensional task that is related to sensorimotor but also cognitive processes. LRJT accuracy in the right affected hand of participants with pain was related to measures of cognitive, sensory, and motor function. These differences in sensory, motor, and cognitive function need to be addressed when attempting to understand differences in LRJT performance between groups. Differences in LRJT performance between a subset of participants suggest that taking pain medication, higher pain severity, impaired cognitive function, and decreased motor performance may be indicators of altered sensorimotor integration and highlight persons that may benefit from cognitively oriented rehabilitation strategies in addition to conventional rehabilitative care.
Disability of the Arm, Shoulder and Hand
Joint Position Sense
Left Right Judgement Task
Motor Imagery Questionnaire
Motor Imagery Questionnaire–Visual Motor Imagery
Multiple linear regression
Participants who did not take pain medications on day of evaluation
West Haven Yale Multidimensional Pain Inventory
Participants who took pain medications on day of evaluation
Purdue Pegboard Test
Pressure pain threshold
Reaction time
Two-Point Orientation Discrimination.
The data used to support the findings of this study are available from the corresponding author upon request.
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
All authors declare that we have no conflicts of interest and disclose that we have no financial and personal relationships with other people or organisations that could inappropriately influence this work.