Abstract
Objective
Movement dysfunction in Parkinson’s disease (PD) is well documented but its cause remains unclear. This study examined whether motor symptoms are caused by sensory impairment, executive dysfunction or other factors.
Methods
We examined the integration of tactile and motor programmes using functional magnetic resonance imaging (fMRI) in early stage PD patients compared with neurologically normal controls, to investigate neural mechanisms underlying movement dysfunction.
Results
Twenty-two control participants exhibited activation in an extensive network of cortical regions involved in tactile–motor integration. Decreased activation in somatosensory and motor regions during performance of passive tactile and movement tasks was found in 21 PD patients. The extrastriate visual cortex exhibited greater activation in controls during performance of the tasks. In contrast, greater activation in bilateral prefrontal regions was found in PD patients.
Conclusions
These results indicate that the extrastriate visual cortex is a multisensory region playing an important role in sensorimotor integration. The findings suggest that PD patients exhibit decreased activation in the extrastriate visual cortex, possibly related to dysfunctional integrative processing, and compensatory activity in the prefrontal cortex.
Introduction
Parkinson’s disease (PD) is a progressive neurological disorder characterized by tremor, rigidity and slowness of movement; it is associated with progressive neuronal loss of the substantia nigra and other brain structures. 1 Movement symptoms involved in PD may be caused by interference of motor programme execution, which results from dysfunctional basal ganglia motor cortex circuits and abnormalities in peripheral afferent inputs or central processing.2,3 The successful execution of a voluntary movement critically depends on peripheral sensory feedback. Sensorimotor integration is the process by which sensory input is integrated by the central nervous system (CNS) and then used to assist in motor programme execution. 4 Compared with neurologically normal people, patients with PD are more reliant on external sensory information for motor initiation and execution, suggesting that impaired proprioception may play a role in PD symptomatology. 5
Abbruzzese and Berardelli 5 proposed that some types of movement disorder are characterized by a lack of inhibition at multiple levels of the CNS and by defective sensorimotor organization. 6 In addition, it has been demonstrated that patients with PD, especially those with early stage disease, exhibit severely depressed frontal responsiveness to sensory stimuli, as tested with sensory evoked potentials. 7 Forss and Jousmäki proposed that the secondary somatosensory cortices are also important in sensorimotor integration. 8 An increasing body of evidence indicates that, compared with neurologically normal people, in patients with PD, the supplementary motor area (SMA) commonly exhibits hypoactivation, whereas other cortical motor regions (such as the cerebellum, premotor area [PMA] and parietal cortex) exhibit hyperactivation while performing motor tasks.9,10
A fundamental question in motor research is how motor planning operations and sensory feedback are implemented in the human brain. 11 Studies have indicated that several separate sensory systems are involved in motor planning, including the haptic, visual and auditory systems.11,12 To prepare for and accomplish an action successfully, information from these systems must somehow be integrated. For people with PD, a deficiency may exist in the sensory–motor integration process, 12 and the prefrontal lobe may constitute the neural substrate of this dysfunction. 13 Lack of a definite correlation with clinical features means that it remains unclear whether abnormalities of sensorimotor integration play a major role in the development of akinesia.
We used multitask functional (f) magnetic resonance imaging (MRI) to investigate which brain regions are involved in the integration of tactile–motor information, and to determine whether activity in these regions differs when patients with early PD are compared with normal controls.
Patients and methods
Participants
This study was carried out at The First Affiliated Hospital of Zhejiang University, Hangzhou, China between July 2007 and September 2012. All procedures were approved by the Institutional Review Board of Zhejiang University. All participants gave written informed consent before taking part in the study.
The diagnosis of early PD was made by X.Z., an experienced neurologist, based on medical history, standard physical and neurological examinations, response to standard levodopa or dopaminergic therapy, laboratory tests and MRI scans to exclude other diseases. Patients were examined only after their medication had been withdrawn for ≥15 h. Patients were assessed using the Unified Parkinson’s Disease Rating Scale, 14 the Hoehn and Yahr disability scale 15 and the Mini-Mental State Examination (MMSE) while they were off medication.
Patients with Hoehn and Yahr 15 stages 1 or 2 PD were eligible for study inclusion. These early PD patients, selected by the neurologist in the outpatient department, were asked whether they wanted to participate in this study. After patients signed the informed consent forms, they had to agree to routine MRI scanning of brain to rule out any other abnormal conditions. All patients were recruited from the First Affiliated Hospital, Zhejiang University. Neurologically normal subjects participated in this study as controls. All of the healthy volunteers were from Zhejiang Province, China, and were recruited over the internet. They were matched with the patients for age and sex. Control subjects were aged between 30 and 75 years, had no history of neurological or psychiatric disorders and were screened for dementia and depression using the MMSE before study entry, as well as undergoing routine MRI scanning of the brain to exclude any abnormalities. All participants were right-handed, as assessed by the Edinburgh handedness inventory. 16
Study design
The multitask paradigm included tactile stimulation, right-hand motor tasks and tactile–motor integration tasks, described below. These tasks were presented in the following order: passive tactile stimulation task; motor task; integration task. Each task unit consisted of a 30-s performance period followed by a 30-s rest (control) period (Figure 1(a)). There were three epochs, with a total testing period of 4.5 min and a 5-min control period. The total fMRI time for 19 blocks was 9.5 min. All participants were fully trained on how to perform the tasks before the scanning session. During practice, errors were recorded and feedback was provided, to inform participants whether their finger movements were correct or incorrect. Training continued until all participants performed all tasks correctly.
(a) Schematic representation of the tasks procedure in a study to investigate neural mechanisms underlying movement dysfunction, using functional magnetic resonance imaging (fMRI) in early stage Parkinson’s disease (PD) patients compared with neurologically normal controls. The cycle was repeated three times. (b) Tasks cue display seen by participants during three tasks.
Passive tactile stimulation task
The stimuli in this task consisted of two commercially available, round wooden wheels: one had a smooth rim; the other had a serrated rim. Both wheels were 2.80 cm in diameter and 1.20 cm thick. A block design paradigm was used in which tactile stimulation was applied in some blocks. The duration of blocks was 30 s. During this stimulation task, an experimenter manually applied the two tactile stimuli (smooth or serrated) by rolling one of the wheels over the participant’s right index fingerpad. The stimulation was applied for 2 s per trial, with a 1-s interval, rolling the wheel in a fixed orientation. The stimuli were alternated in a pseudorandom order, i.e. the order of stimuli seems to be random, but in fact it was set in advance by the experimenter, with the aim of obtaining a genuine response. There were 10 trials within each block and participants were never allowed to see the stimuli.
Motor task
The procedure for this initiative motor task was the same as that for the tactile task, except that a standard finger tapping movement replaced the tactile task. Subjects tapped their own fingers using the thumb and index fingers, during simultaneous fixation on visual cues.
Tactile–motor integration task
During the tactile–motor task, an experimenter manually applied the wheel stimulus to the participant’s right index fingerpad for 2 s per trial, alternating in a pseudorandom order. Participants were instructed to do nothing if they felt the smooth rim, and to tap the thumb and index fingers once if they felt the serrated rim.
fMRI procedure
Participants were scanned using a 1.5T GE Signa EXCITE MRI scanner (GE Healthcare, Milwaukee, WI, USA) with an 8-channel head coil. Each scanning session began with a high-resolution T1-weighted three-dimensional volume acquisition for anatomical localization (voxel size, 1 × 1 × 1 mm3). This was followed by acquisitions of echoplanar T2*-weighted images with blood oxygenation level-dependent (BOLD) contrast (echo time, 35 ms; flip angle, 90°). A total of 190 volumes were acquired per fMRI session. Each volume contained 22 slices of 5 mm thickness (matrix size, 64 × 64 pixels; voxel size, 2.35 × 2.35 × 7 mm3).
Participants lay supine in the scanner with the right arm extended and the right hand supinated. The arm was comfortably supported by foam padding, which also functioned to minimize the transfer of gradient coil vibration to the upper extremity. Foam blocks, as well as chin and forehead straps, were used to reduce head movement. Earplugs were worn to protect the participant’s hearing.
Using the block design paradigm described above, each functional run contained 19 stimulation blocks, each with a duration of 30 s. Nine experimental and 10 control blocks were presented in an alternating order (Figure 1(a)). Each stimulation block contained 10 3-s trials. Thus, there were 30 trials for each condition per run.
Two conditions were contained in each scanning session: these were defined as the ‘rest’ and ‘task’ conditions. Each condition lasted 30 s and was repeated three times in each session. In the rest condition, participants were instructed to relax and focus on the screen in front of them. When participants performed tasks they were also instructed to focus on the screen. Immediately before each block, participants were presented with a visual cue with a black background, on the screen. The sequence and timing of stimuli were guided by preprogrammed instructions, displayed to the participants and experimenter on a computer screen using E-prime software (Psychology Software Tools, Sharpsburg, MD, USA). All stimulus cues were projected on a screen positioned outside the scanner at a distance of 2 m from the gantry (Figure 1(b)). Participants viewed the projected screen via a mirror mounted on the head coil. During scanning, no feedback was provided to communicate to participants whether their finger movements were correct or incorrect.
Imaging and statistical analyses
The BOLD signal obtained during performance of tactile, movement and tactile–movement integration trials was compared with that obtained during the corresponding period in the control condition. Intergroup analyses were also conducted.
Image analyses were performed with SPM2 software (Wellcome Institute of Cognitive Neurology, London, UK). Functional images were aligned to the first image of each session using a rigid-body transformation procedure. After spatial normalization, images were then transformed into Talairach space 17 and smoothed with a 7-mm full-width at half maximum Gaussian filter. First- and second-level analyses were performed. In the first-level, data were analysed for each participant separately on a voxel-by-voxel basis using the principles of the general linear model, extended to allow the analysis of fMRI data as a time series. 18 For within-group analyses, a one-sample t-test model was used to identify the brain activity associated with each task (P < 0.001, without correction for multiple comparisons), using a cluster size >20 voxels as a spatial extent threshold. As this step is dependent on the amplitude of head movements, all data in which the head movement was over 5 voxels were abandoned, to avoid the need for corrections that may indeed give a false activation. This procedure has been performed by others.19,20
For intergroup comparisons, two-sample t-tests (P < 0.05, uncorrected) were used to explore the differences between PD patients and normal controls during each task. Locations of activated areas for different conditions were displayed by superimposing them on the Montreal Neurological Institute (MNI) template using SPM2. All co-ordinates were reported in Talairach space converted from MNI space, based on an algorithm (www.mrccbu.cam.ac.uk/imaging/mnispace.html). Specific areas within the cortex were identified as numbered Brodmann areas (BA).
In addition, age data were presented as mean ± SD.
Results
Participants
Demographic and clinicopathological characteristics of people with early stage Parkinson’s disease (PD) (Hoehn and Yahr [H&Y] stages 1 and 2), 15 included in a study to investigate the neural mechanisms underlying movement dysfunction using functional magnetic resonance imaging (fMRI) in early PD patients and controls.
UPDRS, Unified Parkinson’s Disease Rating Scale, 14 MMSE, Mini-Mental State Examination; F, female; M, male; CAB, cabergoline; LD, L-dopa; PMX, pramipexole; PRG, pergolide; SLG, selegiline.
Tactile tasks
In the control group, the tactile stimulus was associated with significantly increased activation in the bilateral somatosensory cortical areas, in BA 2 on the right and BA 40 on the left, relative to the resting condition (Table 2, Figure 2a). A number of visual regions – sometimes referred to as the middle temporal gyrus, V521–23 and the left occipital lobe (BA 17/18) – were also activated (Table 2). The cerebellum also exhibited activation increases (Table 2, Figure 2b).
(a) Tactile activated map of neural mechanisms underlying movement dysfunction using functional magnetic resonance imaging (fMRI) in early stage Parkinson’s disease (PD) patients compared with neurologically normal controls. In controls, the higher activated area is the postcentral gyrus (arrows). Increased activation is also seen in the left middle temporal gyrus (curved arrow) and cerebellum (arrowhead). (b) Motor tasks activated map in controls and PD patients. In controls, higher activated areas include the right cerebellum, left somatosensory cortex (arrow), bilateral motor cortex (curved arrows) and right frontal gyrus (arrowhead). (c) Tactile-motor integrated tasks activated map in controls and PD patients. In controls, higher activated areas are the bilateral somatosensory cortex (arrows), left extrastriate visual cortex (arrowhead), and the right cerebellum (curved arrow). The colour version of this figure is available at: http://imr.sagepub.com. Within-group analyses of activated brain areas receiving (A) tactile stimulus, (B) movement tasks and (C) and sensory-movement tasks in Parkinson’s disease (PD) patients and matched neurologically normal controls, included in a study to investigate neural mechanisms underlying movement dysfunction using functional magnetic resonance imaging (fMRI) in early stage PD (Hoehn and Yahr stages 1 and 2)
15
. Co‐ordinates given as stereotaxic co-ordinates referring to the atlas of Talairach and Tournoux.
17
Cluster size is the number of voxels. All areas were significant at P < 0.001 (one-sample t-test, uncorrected). L, left; R, right
Intergroup comparisons between Parkinson’s disease (PD) patients and matched controls included in a study to investigate the neural mechanisms underlying movement dysfunction using functional magnetic resonance imaging (fMRI) in early stage PD (Hoehn and Yahr stages 1 and 2) 15 .
Co-ordinates given as stereotaxic co-ordinates, referring to the atlas of Talairach and Tournoux. 17
Two-sample t-tests (P < 0.05, uncorrected) were used to explore differences between PD patients and normal controls during each task. Cluster size is the number of voxels.
All activated areas were significant at P < 0.001 (one-sample t-test, uncorrected).
L, left; R, right.
Motor tasks
Control subjects exhibited stronger signals in the right cerebellum, bilateral motor cortex (BA 4/6), left somatosensory cortex (BA 2/3), and right frontal gyrus (BA 10) compared with PD patients (Table 2, Figure 2b).
Intergroup comparisons showed significantly stronger activation in control subjects in the right extrastriate visual cortex (MT/V5, BA 18) and the caudate nucleus (Table 3). In contrast, greater activation was observed in PD patients in the bilateral frontal lobe (BA 24/9/32) and right thalamus (Table 3).
Tactile–motor integrated tasks
Controls demonstrated an increased signal in the bilateral somatosensory cortex (BA 40/2), left extrastriate visual cortex (BA 19, 37, left middle temporal gyrus/V5), thalamus and right cerebellum while performing tactile–motor integrated tasks (Figure 2C; Table 2).
Intergroup comparisons revealed increased activity in the right occipital lobe (BA 18/19) in control subjects compared with PD patients (Table 3). In contrast, increased activation was observed in PD patients compared with controls in the bilateral frontal lobe (BA 10/47) (Table 3).
Discussion
Results presented in this paper suggested that the right dorsal premotor BAs (PMdr), posterior parietal cortex, cerebellum, left postcentral gyrus and precentral gyrus comprised a sensorimotor integration neural network in neurologically normal subjects because they were activated during integration tasks. These regions could also be part of the striatal–thalamo–cortical loop, since this loop includes several parallel circuits, including the sensorimotor, associative and limbic circuits.10,24,25 The sensorimotor circuit projects somatotopically from the primary sensorimotor area, PMA and SMA to the putamen, then through the thalamus and back to these cortical motor areas.26,27 Studies have suggested that the PMd (especially the PMdr), is involved in integrating spatial information with motor programming to produce precise sequential movement maps that are necessary to facilitate complex movements.28,29 In addition, the posterior parietal cortex is thought to integrate visual, somatic and other information, and provides a sensory representation of extrapersonal space. 26
In PD patients, the sensorimotor integration networks probably include the bilateral middle frontal gyrus (BA 6/10; Table 2), left middle temporal gyrus/V5 and the right occipital lobe (BA 18), since they are activated during integration tasks. Observations in PD patients, including reports of excessive reliance on visual information during movement tasks, have suggested an abnormality in the sensorimotor integration process. Some studies reported that training associations between auditory or visual cues in PD patients produced greater improvements than conventional treatments.30,31 Our present findings support the notion that visual cues can provide a supplementary source of information to improve the efficacy of the visual–motor pathway, since the right middle frontal gyrus (BA 6) was activated. Importantly, the left middle temporal gyrus is involved in movement perception, and the left parietal cortex is involved in temporospatial movement control. 32 We observed abnormal activation in the left middle temporal gyrus when PD patients performed motor tasks.
Our study revealed decreased visual activation in PD patients during tactile, movement and sensorimotor integration tasks when compared with controls. Extrastriate visual regions correspond to right BA 18, BA 19 and the inferior temporal gyrus (BA 20). Our results also revealed activity in the bilateral middle temporal gyrus/V5 during all three tasks in both groups. We hypothesize that decreased activity in the visual cortex may relate to hypokinetic movements in PD.
The term visual cortex refers to the primary visual cortex (also known as striate cortex or V1) and extrastriate visual cortical areas, including V2, V3, V4 and V5. The primary visual cortex is anatomically equivalent to BA 17. A hierarchical relationship between a primary cortex and secondary cortices is found in several parts of the brain, including the sensory, motor and auditory cortices, which as a whole perform information processing on a more global level. 33 Several fMRI studies have confirmed that these multisensory areas are involved in tactile and movement perception, in normal subjects.21,34,35
In accordance with such findings, our results confirmed an activation in extrastriate visual cortical areas during tactile tasks and, importantly, during movement and integration tasks in neurologically normal controls. However, our findings revealed significantly decreased activity in right BA 18 and BA 19, in patients with early PD compared with neurologically normal controls.
The BAs 18 and 19 are involved in the translation and interpretation of visual impressions transmitted from BA 17. Although simple and complex cells are found in visual association cortex, the region is predominantly populated by hypercomplex (both higher and lower order) neurons, most of which are involved in the determination of precise geometric form, as well as the assimilation of signals transmitted from the primary cortex. 35 In addition to visual input, neurons in the superior portions of BA 19 respond to tactile stimuli. 36 We propose that decreased activation at these loci (BAs 18 and 19) is likely to be involved in the dysfunction of tactile perception in PD and may play a role in sensorimotor integration.
Visual area middle temporal gyrus/V5 is thought to play a major role in the perception of motion, the integration of local motion signals into global percepts and the guidance of some eye movements. 37 Research has demonstrated that neurons in this area are capable of responding to visual information, often in a direction-selective manner, even after V1 has been destroyed or deactivated. 38 Lesion research has confirmed the role of the middle temporal gyrus/V5 in motion perception, with a neuropsychological report of a patient with lesions in this area who was unable to visualize motion, instead seeing the world in a series of static ‘frames’ 39 Our surprising finding of decreased activity in the middle temporal gyrus/V5 in PD patients during the movement task suggests that this region may play a key role in sensorimotor computation and movement dysfunction.
The present study has some limitations. We were unable to get an accurate correlation in the tactile–motor procedure, which is the aim of our next study. The effect of medication was minimized as much as possible, otherwise PD patients would have been unable to fulfil the tasks due to tremor. While high-quality images were needed for this study, PD patients find it difficult to control their movements. For this reason, early stage PD patients were selected. The study began in 2007 and continued for >5 years due to the difficulty in recruiting suitable participants. The software package we used (SPM2) was current at the start of the study, but has been updated in subsequent years. Although the older version lacked consistency and was less compatible with toolboxes than later versions, we believe the results are still credible.
Based on these observations, we propose that extrastriate visual responses to object shapes and perception of motion might be similarly integrated, enhancing the population response to sensorimotor stimuli. Understanding the precise role of this area in perception and cognition will require additional research in future.
To conclude, the current results provide novel insights into the cortical components of sensorimotor integration that may underlie the motor symptoms of PD. In particular, our findings suggest an important role of initial sensory input and movement outcomes related to extrastriate visual cortex activity. Future work should focus on elucidating the precise multisensory role of the extrastriate visual cortex.
Footnotes
Declaration of conflicting interest
The authors declare that there are no conflicts of interest.
Funding
This study was supported by the Zhejiang Provincial Natural Science Foundation of China (Grant No.LY13H180005), Public Interest Research on Social Development Projects of Zhejiang Province (Grant No. 2012c33051), and the Foundation of the Education Department of Zhejiang Province of China (Grant No. Y200909841).
Acknowledgements
We thank Dr Lifei Ma, GE MR modality application specialist, and Dr Xiaoping Xie, Physics Department of Zhejiang University, for their help with data analysis for this study.