The authors have declared that no competing interests exist.
Conceived and designed the experiments: AP GY RK RC. Performed the experiments: AP LN. Analyzed the data: AP LN KS RK YHC. Contributed reagents/materials/analysis tools: RK YHC RC AP. Wrote the manuscript: AP KS LN GY. Critically reviewed the manuscript: GY. Designed figures and tables: AP (Figs. 2, 3, 4, 5), LN (Figs. 1, 6). Revised the manuscript: AP.
Current address: Department of Dermatology, Temple University School of Medicine, Philadelphia, Pennsylvania, United States of America
Previous brain imaging studies investigating the brain processing of scratching used an exogenous intervention mimicking scratching, performed not by the subjects themselves, but delivered by an investigator. In real life, scratching is a conscious, voluntary, controlled motor response to itching, which is directed to the perceived site of distress. In this study we aimed to visualize in real-time by brain imaging the core mechanisms of the itch-scratch cycle when scratching was performed by subjects themselves. Secondly, we aimed to assess the correlations between brain patterns of activation and psychophysical ratings of itch relief or pleasurability of scratching. We also compared the patterns of brain activity evoked by self-scratching vs. passive scratching. We used a robust tridimensional Arterial Spin Labeling fMRI technique that is less sensitive to motion artifacts: 3D gradient echo and spin echo (GRASE) - Propeller. Active scratching was accompanied by a higher pleasurability and induced a more pronounced deactivation of the anterior cingulate cortex and insula, in comparison with passive scratching. A significant involvement of the reward system including the ventral tegmentum of the midbrain, coupled with a mechanism deactivating the periaqueductal gray matter (PAG), suggests that itch modulation operates in reverse to the mechanism known to suppress pain. Our findings not only confirm a role for the central networks processing reward in the pleasurable aspects of scratching, but also suggest they play a role in mediating itch relief.
Scratching is the natural response to itch and, by definition, inseparable from it. The act of scratching not only diminishes itch, but can be rewarding and addictive. The itch-scratch cycle is a complex phenomenon involving sensory, motor and emotional components. It is well-known that the urge to scratch can be remarkably intense, since the reward provided by scratching brings itch relief and associated feelings of pleasurability. Recent studies have shown that rating scratching as a pleasurable experience is correlated with the intensity of the underlying itch, both in patients with chronic itch [
Previous brain imaging studies have employed BOLD fMRI to examine the effects evoked by scratching. However, these studies used “passive scratching” performed by an investigator using a device, such as a cytology brush [
We opted to include both modalities of scratching in our study in order to dissect the characteristic features of the itch-scratch cycle, when scratching was controlled and performed by subjects themselves in response to the itch they perceived. In this regard, passive scratching provided a control condition, offering only the mechanical stimulation that reduces itch, but missing the intricate, inner feedback loops which would close within the same person’s brain, when placed at the center and in control of the action, instead at the receiving end. We therefore specifically aimed to contrast these two conditions, to better illuminate the intimate rapport between sensory perception and motor action, and to observe how this interplay would be reflected in the mediation of itch relief. It is known in the field of reward that there are different operational modes: one which is dependent on one’s own actions (a contingent form of reward) and one that is exogenously provided and passively received. Thus, one of our aims was to uncover potentially meaningful differences between these two forms of relief. An additional rationale to include both forms of scratching in the present study was that previous research papers on this topic have used a passive form of scratching. Therefore, we intended to see how our results would compare to the previous reports and decided to study both active and passive forms of scratching, within the confines and internal controls of the same study, using the same methodology.
All subjects responded to itch stimulation, with an average itch intensity of 7.2±1.3 (on a scale of 0-10). Visual Analog Scale (VAS) ratings of the perceived itch intensity, of the pleasurability of scratching and the magnitude of itch relief provided by active scratching or passive scratching are presented in
Ratings were recorded immediately at the end of each fMRI scan. (*** p<0.001; ** p<0.01).
In this section, we describe the complex patterns of activation observed during the continuous brain processing of actively scratching an itch (presented in
Activations induced by actively scratching an itch (yellow), main areas activated by itch (red-orange), and major areas of deactivations induced by both forms of scratching are shown (with active scratching in blue, passive scratching in green). Standard Talairach space coordinates (x, y, z). The color tones displayed correspond to Z score values as depicted in the color bars (at the same scale). R = Right; L = Left.
B) Deactivations induced by active scratching (yellow) were represented in the hippocampus, amygdala, ACC and only slightly overlapped with the deactivations induced by passive scratching in the ACC (red). Standard Talairach space coordinates (x, y, z). The color tones displayed correspond to Z score values as depicted in the color bars (at the same scale). R = Right; L = Left.
The deactivations evoked by active scratching an itch were significantly represented in the following areas: ACC (Right, R), amygdala (R), cerebellum (anterior lobe, culmen; R), frontal medial cortex (R), frontal pole (bilateral), hippocampus (R), inferior frontal gyrus (Left, L), insula (R), and paracingulate gyrus (bilateral) (
Actively scratching an itch (blue) deactivated the ACC, the ventromedial prefrontal Cortex, midbrain (VTA), while the effect of passive scratching was found only in the ACC (green). Standard Talairach space coordinates (x, y, z). The color tones displayed correspond to Z score values as depicted in the color bars (at the same scale). R = Right; L = Left.
At the level of
In order to gain further insight into the intricate mechanism of self-scratching, we have performed the following higher-level contrast analyses. We analyzed brain activation patterns in respect to activations evoked by itch itself and compared active with passive scratching in the presence or absence of itch. The areas that were significantly deactivated by scratching an itch either passively or actively are shown in
In order to identify the specific features of scratching an itch in comparison with scratching alone (in the absence of itch), we performed a higher level contrast analysis between these two conditions. Interestingly, the primary motor cortex, the premotor cortex and somatosensory area S1 were more strongly activated by scratching in the presence of itch compared to scratching without an itch. Areas related to emotion (ACC), craving (nucleus accumbens, NAc) and motivation (putamen) were emphasized in the context of scratching an itch. An overview summarizing the effects induced by the two forms of scratching in various cerebral regions of interest is presented in
Regression analyses of brain activity versus pleasurability ratings showed that areas deactivated while actively scratching an itch coincided with areas significantly correlated with pleasurability. These areas are presented in
Z score | ||||
---|---|---|---|---|
VTA | -2 | -16 | -8 | 2.91 |
2 | -16 | -14 | 2.39 | |
-2 | -24 | -14 | 3.96 | |
VTA (rmTgm) | 2 | -24 | -20 | 5.58 |
Raphé nucleus | 4 | -26 | -8 | 4.33 |
-2 | -26 | -8 | 6.08 | |
Red Nucleus | 6-8 | -20 | -8 | 3.04 |
4 | -22 | -8 | 3.16 | |
-4 | -24 | -8 | 2.75 | |
PAG | 4 | -34 | -8 | 3.00 |
-2 | -30 | -8 | 2.95 | |
-4 | -38 | -8 | 3.86 | |
-2 | -32 | -8 | 4.31 | |
-4 | -30 | -8 | 5.52 | |
-2 | -30 | -6 | 3.71 | |
0 | -28 | -6 | 7.44 | |
2 | -28 | -6 | 2.72 | |
Sb. nigra | -14 | -22 | -8 | 2.51 |
-14 | -24 | -8 | 5.80 | |
VPM | -14 | -22 | 0 | 2.86 |
-16 | -20 | 2 | 2.50 | |
VPL | -18 | -22 | 0 | 2.37 |
VL | 12 | -12 | 0 | 3.45 |
Mediodorsal nc. | 12 | -20 | 8 | 2.64 |
-6 | -12 | 8 | 2.33 | |
-8 | -12 | 8 | 2.41 | |
-8 | -16 | 8 | 5.67 | |
Ventral Anterior nc. | 14 | -6 | 10 | 2.63 |
VPM | -12 | -18 | 6 | 2.49 |
Pulvinar | -16 | -24 | 8 | 6.40 |
-20 | -26 | 8 | 5.58 | |
-14 | -26 | 8 | 4.63 | |
-20 | -24 | 8 | 3.46 | |
Mammillary body | 14 | -18 | 0 | 3.49 |
6 | -12 | -14 | 3.07 | |
14 | -16 | 0 | 2.82 | |
12 | -20 | 0 | 3.93 | |
Hippocampus |
-22 | -42 | 2 | 4.61 |
-22 | -40 | 2 | 5.61 | |
Parahippocampal gyrus | -24 | -42 | -6 | 5.03 |
ACC | -4 | 32 | 8 | 2.96 |
ACC | 2 | 18 | -8 | 3.71 |
-4 | 30 | -8 | 2.66 | |
PCC (BA 30) | 4 | -48 | 18 | 4.30 |
PCC (BA 23) | 4 | -56 | 18 | 5.31 |
PCC (BA 29) | 4 | -46 | 18 | 3.42 |
PCC (BA 31) | 10 | -68 | 18 | 3.48 |
S2 | 36 | -6 | 18 | 2.91 |
Putamen | 22 | 0 | 10 | 3.58 |
30 | -6 | 10 | 2.84 | |
Precuneus | 14 | -66 | 22 | 2.72 |
20 | -62 | 22 | 2.33 | |
16 | -64 | 22 | 2.56 | |
Cuneus (BA 18) | 8 | -70 | 22 | 2.33 |
Cuneus | 6 | -80 | 42 | 3.12 |
Lingual gyrus | -26 | -64 | 2 | 3.54 |
Fusiform gyrus | -14 | -74 | 2 | 3.17 |
Occipital pole | 8 | -88 | 22 | 6.21 |
12 | -88 | 22 | 3.29 | |
Sup. frontal gyrus | 22 | 64 | -8 | 5.06 |
Middle frontal gyrus | -34 | 58 | -8 | 6.50 |
Inf. frontal gyrus | -48 | 20 | -8 | 2.52 |
Talairach standard space coordinates (x, y, z). Z score > 2.3; p< 0.05. BA = Brodmann area; ACC = anterior cingulate cortex; PCC = posterior cingulate cortex; VTA = ventral tegmental area; PAG = periaqueductal gray; VPL= ventral posterolateral nucleus; VPM = ventral posteromedial nucleus; S2 = secondary somatosensory area; VL = ventrolateral nucleus; rmTgm = rostromedial tegmentum.
VAS ratings of pleasurability and itch relief (calculated as the difference in itch ratings before and after scratching) were used as covariates of interest. PAG = periaqueductal gray; VTA = ventral tegmental area; SN= substantia nigra; ACC = anterior cingulate cortex; PCC = posterior cingulate cortex; rmTgm = rostromedial tegmentum. Talairach standard space coordinates (x, y, z). Areas significantly correlated with a Z score > 2.3, p < 0.05 are displayed. The color tones displayed correspond to Z score values as depicted in the color bars (at the same scale). R = Right; L = Left.
Regions which were activated (red) or deactivated (blue) by scratching are depicted. __(solid line) = positive correlation; ---- (dashed line) = inverse correlation with pleasurability ratings. Box size for each area is proportional to the magnitude of Z scores obtained in the regression analysis.
We have also performed a regression analysis of brain areas whose activity correlated with the decrease in the ratings of itch perception, during actively and passively scratching an itch (
Z score | ||||
---|---|---|---|---|
Thalamus | -8 | -8 | 14 | 2.46 |
Pulvinar | -14 | -24 | 14 | 3.03 |
-4 | -8 | 12 | 2.87 | |
Ventral lateral nucleus | -8 | -10 | 6 | 2.46 |
4 | -14 | 12 | 2.99 | |
-4 | -10 | -14 | 3.23 | |
Mediodorsal nc. | 6 | -14 | 12 | 3.50 |
10 | -18 | 12 | 2.89 | |
-4 | -10 | 12 | 2.79 | |
Cuneus | 14 | -78 | 14 | 3.87 |
Lingual gyrus | 4 | -80 | 12 | 4.21 |
Putamen | 26 | -2 | 12 | 2.40 |
Caudate body | 14 | -2 | 12 | 2.78 |
PCC | -4 | -68 | 14 | 2.33 |
The ratings of itch relief were calculated as the difference between initial itch intensity and actual itch after scratching was performed. Talairach standard space coordinates (x, y, z). Z scores > 2.3, p < 0.05. PCC = posterior cingulate cortex.
One of the most intriguing neurobiological questions related to scratching an itch is why scratching is so pleasurable and what role it plays in relieving itch perception. In this study, we have discovered that brain responses evoked by active scratching implicate multiple structures previously described to process reward: VTA, NAc, the caudate nucleus and the ventromedial prefrontal cortex (vmPFC). The evoked responses in most of these brain areas were correlated with the pleasurability of scratching, while other areas were associated with itch relief. There was a limited degree of overlap between these regions, which suggests that perceiving itch relief and pleasurability are not identical processes. One of the most interesting results is that most areas correlated with pleasurability were virtually identical with areas of deactivation evoked by actively scratching an itch (
The
A previous study in non-human primates investigating the mechanism of itch inhibition by scratching showed that the effect on itch-specific spinal neurons was state-dependent, being active only in the presence of itch [
Multiple anatomic subdivisions of the ventral and dorsal striatum: the nucleus accumbens, caudate, putamen, previously attributed roles in encoding motivation, expectation or pleasure, were associated with the pleasurability of scratching, a finding which is in good agreement with other studies: the putamen was linked with the relief of itch induced by acupuncture [
A possible connection at the level of midbrain between the activation of motor relays coordinating the voluntary motor response - inherent in scratching motions - and the activation of reward circuits is strongly suggested. The activity observed in the red nucleus and substantia nigra (as well as in the cerebellum) is not surprising, as these areas are classically involved in motor coordination; however, their activation was also strongly correlated with pleasurability. Noteworthy, substantia nigra is a major source of dopamine in the nigrostriatal and mesolimbic circuits. Our results suggest that the association of motor-related activity with pleasurability may have a basis in the implication of midbrain structures. Substantia nigra and VTA could represent the key areas encoding the pleasurability of scratching. Behavioral and pharmacological studies of dopamine pathways have led to the association of the mesolimbic pathway with reward processing and of the nigrostriatal pathway with motor activity; recently, both circuits have been associated with reward [
It is worthy of mention that besides the involvement of midbrain structures, active scratching, which provided a higher pleasurability and itch relief, induced a more extensive deactivation of the anterior cingulate cortex, prefrontal cortex, insula and the lentiform nucleus, other main areas involved in processing reward. This effect was much stronger in comparison with passive scratching.
In addition, our results suggest that thalamus has a significant role in processing the variable sensory information of perceiving itch while scratching. Multiple nuclei in the thalamus were significantly activated by active scratching of itch, and were correlated with pleasurability and itch relief. The activity of VPL, a known station processing itch information projecting to the cortex, was correlated with itch relief. The anterior nucleus of thalamus could play a role in the assessment of itch reduction, due to its activation during active scratching. The pulvinar was significantly activated during self-scratching and its response was correlated with the perceived pleasurability. The medial dorsal nucleus (
Active self-scratching recruited stations of the reward circuit: NAc, VTA, vmPFC, SN, emphasizing the addictive or reinforcing character of this behavior. In contrast, only parts of these circuits were stimulated by passive scratching: ACC and vmPFC. Passive scratching uniquely activated the subthalamic nucleus, a formation that has also been implicated in reward [
In conclusion, the engagement of reward circuits during active scratching may represent a major route by which antipruritic top-down mechanisms are initiated. Our data indicate that the reward is dually perceived as a decrease in itch intensity and a pleasurable input that accompanies scratching. These co-existing, synergistic mechanisms combine to counter the negative emotional-affective impact of itch.
14 healthy volunteers: 8 females, 6 males, with ages 19-58 (average age 30.3 ± 9.2) were enrolled. Subjects were free of any skin disease and were not using any systemic or topical medications. A written informed consent was obtained. The protocol was approved by the Internal Review Board of Wake Forest University Health Sciences. The study was conducted in agreement to the principles of Declaration of Helsinki.
Two baseline scans at rest were acquired first. One fMRI scan was acquired separately for passive or active scratching intervention without itch and was used as a distinct control. For the passive scratching series, the participants were exogenously scratched by an investigator (see below). The itch condition alone - without scratching - was imaged separately. Acute experimental itch was induced by applying cowhage spicules on the distal volar side of the right forearm, as previously described [
Active scratching was performed continuously by study participants themselves during the entire scan period (4 min 54 s) using only the tips of the fingers of the left hand to scratch a distal site on the right forearm (2-3 cm above the wrist), while strictly refraining from other movements and minimizing larger movements of the left arm. To easily reach the right forearm (with the left hand), the right hand was placed on the chest in the most comfortable position possible while lying (supine) inside the scanner.
Passive scratching was delivered by a study investigator using a MediPak cytology brush no. 7 (General Medical Corporation, Elkridge, MD) which applied a constant force equivalent to a weight of 29 g, as described previously [
A mixed model analysis accounting for subject as a random factor and adjusted for itch while scratching was performed to compare pleasurability between the active and passive itch while scratching, using PASW 18.0 software (SAS; Chicago, IL). Paired t tests were performed to compare pleasurability for each form of scratching vs. its own control (scratching without itch), and the decrease in itch ratings by active or passive itch with scratching to its itch control. Mixed model analysis with subject as a random factor was used to compare itch relief (as the difference in itch ratings between itch while scratch vs. itch alone) between active vs. passive modalities.
All experiments were carried out on a GE 1.5T TwinSpeed scanner (GE Healthcare, Milwaukee, WI) with an eight-channel phased array receive-only head coil (Invivo Devices, Gainesville, FL) for data collection. PseudoContinuous Arterial Spin Labeling consisted of series of 800 µs, 25 degree Hanning window-shaped RF pulses. Repetition time between pulses was 1736 µs; gradient residual moment was set at 13.6%. Labeling duration was 1600 ms with a post-labeling delay of 1000 ms. Control and label imaging pairs were acquired by changing the phase of alternating RF pulse by 0 degrees and 180 degrees, respectively. Two non-selective inversion pulses were interleaved among the pCASL RF pulses and the post-labeling delay at 673 ms and 2219 ms.
The functional image analysis package FSL (Functional Magnetic Resonance Imaging of the Brain Software Library, Center for FMRIB, University of Oxford, Oxford, UK) was used for image processing and statistical analysis. The Cerebral Blood Flow (CBF) data was movement corrected and spatially smoothed using a 5 mm 3D isotropic Gaussian kernel. Each CBF image was scaled by its mean global intensity (intensity normalization) to minimize variability due to global CBF changes. Next, each subject's CBF images were registered to their structural data using a seven-parameter linear 3D transformation and transformed into standard stereotaxic space as defined by the Montreal Neurological Institute (MNI), using a 12-parameter linear 3D transformation. Standard general linear model-based analyses using fixed-effects models within subjects and random effects models (FLAME1+2) between subjects were performed to analyze the contrasts between the series of interest versus their corresponding control conditions, by paired t test as follows: 1) scratching an itch, either actively or passively, was compared to itch condition itself. 2) Scratching an itch was contrasted with scratching in the absence of itch, for either active or passive scratching. 3) Higher level analysis was performed to compare the effects of active scratching with passive scratching, either in the presence or absence of itch. The average effect for each condition was also studied for the entire group. To identify the areas correlated with itch relief and pleasurability, a multiple regression analysis using the VAS ratings as covariates of interest, respectively, was performed under the General Linear Model (GLM), setting up individual orthogonalization of ratings versus main effect in order to prevent overestimation. All analyses were performed as whole-brain analyses. Statistical significance for whole brain analysis was set at p<0.05, Z score > 2.3.
(DOCX)