at SciVerse ScienceDirect

Parkinsonism and Related Disorders 19 (2013) 43e46

Contents lists available

Parkinsonism and Related Disorders

journal homepage: www.elsevier .com/locate/parkreldis

Impaired pain processing in Parkinson’s disease and its relative association withthe sense of smell

Takashi Hara a, Masaaki Hirayama a,b, Yasuaki Mizutani a, Tetsuo Hama a, Norio Hori a,Tomohiko Nakamura a, Shigenori Kato a, Hirohisa Watanabe a, Gen Sobue a,*

aDepartment of Neurology, Nagoya University Graduate School of Medicine, 65 Tsurumai-cho, Showa-ku, Nagoya 466-8550, JapanbDepartment of Pathophysiological Laboratory Science, Nagoya University Graduate School of Medicine, Japan

a r t i c l e i n f o

Article history:Received 1 March 2012Received in revised form25 June 2012Accepted 26 June 2012

Keywords:Parkinson’s diseasePainPerceptionOlfactory dysfunctionAmygdala

* Corresponding author. Tel.: þ81 52 744 2385; faxE-mail address: sobueg@med.nagoya-u.ac.jp (G. So

1353-8020/$ e see front matter � 2012 Elsevier Ltd.http://dx.doi.org/10.1016/j.parkreldis.2012.06.020

a b s t r a c t

Background and purpose: Many non-motor symptoms are associated with Parkinson’s disease (PD). Ofthese, pain and olfactory disturbance tend to be common premotor symptoms. PD has been shown toexhibit abnormal central pain processing, although underlying mechanisms are not yet fully understood.In order to investigate this further, we assessed PD patients by specific Ad stimulation withintra-epidermal needle electrode and determined olfactory function.Methods: Forty-two patients (18 males and 24 females) with PD and 17 healthy control subjects (8 malesand 9 females) were studied. A thin needle electrode was used to stimulate epidermal Ad fibers, andsomatosensory evoked potentials (SEPs) recorded at the vertex. Olfactory function was evaluated usingthe Odor Stick Identification Test for Japanese (OSIT-J) and its relationship with pain-related SEPs wasinvestigated.Results: There were no significant differences in N1 latencies or P1 latencies although N1/P1 peak-to-peak amplitudes were significantly lower (p < 0.01) in PD patients than in control subjects. In PDpatients, there were significant correlations between N1/P1 amplitudes and disease duration (r ¼ �0.35,p < 0.05), Hoehn-Yahr stage (r ¼ �0.38, p < 0.05) and UPDRS part III (r ¼ �0.42, p < 0.01). Furthermore,the OSIT-J scores correlated with SEP amplitude (r ¼ 0.41, p < 0.01).Conclusion: Pain processing in PD patients was impaired under specific nociceptive stimulation of Adfibers and significant correlation with smell dysfunction was detected. We suggest that this mechanismmay involve the limbic system during PD pathology.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Pain is an important and distressing symptom in Parkinson’sdisease (PD) [1]. However, the specific features of this phenomenonand the mechanisms involved have not yet been fully assessed.While several studies have investigated pain threshold withmechanical heat or cold stimulation in PD patients, such resultsremain controversial [2e4]. In general, it is difficult to evaluate thedegree of pain objectively, while pain-related somatosensoryevoked potentials (SEPs) have been considered to be one of thereliable objective assessments of pain processing [5,6].

Pain-related SEPs may be recorded by laser, heat, electric, andmechanical stimulations [7]. Each stimulus to activate the specificnociceptive receptor system in the skin may evoke specific centralpain processing, though the precise features are not well

: þ81 52 744 2384.bue).

All rights reserved.

understood. Recently, Inui et al. recorded evoked potentials inducedby epidermal electrical stimulation (ES) using a thin needle elec-trode, which can specifically activate the Ad fiber-mediated painmechanism [5]. The authors of this magnetoencephalography(MEG) study subsequently reported that the vertex biphasic SEPcomponent corresponded approximately to the activity of medialtemporal cortex [8]. ES with this particular needle electrode isa convenientmethod for the selective stimulation of Ad fibers, sinceit represents a very simple technique which does not require anyspecial apparatus.

In the present study, we recorded pain-related SEPs by using ESmethodology to activate Ad mediated pain mechanisms in PDpatients and healthy controls. Furthermore, pain symptoms ante-date the onset of motor symptoms in some PD patients [9,10].Olfactory dysfunction is also recognized as non-motor symptom inPD patients [11e13] and is considered as a useful diagnostic markerof preclinical PD because pathological changes of the olfactorysystems begin before motor symptoms develop [14]. Furthermore,

Fig. 1. Comparison of N1/P1 amplitude between PD patients (N ¼ 42) and controlsubjects (N ¼ 17). Symbols to the right of each group represent mean � SDs.

T. Hara et al. / Parkinsonism and Related Disorders 19 (2013) 43e4644

in a recent positron emission tomography study, Baba et al reportedthat odor-identification performance was closely associated withbroad cortical dysfunction of the piriform cortex and amygdala, andsuggested that cognitive deficit in olfactory perception is animportant aspect of hyposmia in PD and that this deficit is causedby altered brain metabolism in the amygdala and piriform cortex[15]. Consequently, we considered that pain processing impairmentmay be correlated to olfactory dysfunction in similar regions, andinvestigated whether there may be an association between pain-related SEPs and olfactory function.

2. Methods

2.1. Subjects

Forty-two patients (18 males and 24 females) with PD and 17 healthy controlsubjects (8 males and 9 females) were studied. Patients were recruited from theNagoya University Hospital, Japan. Patients with PD fulfilled the diagnostic criteriafor PD [16]. Motor performance was assessed using the Hoehn and Yahr (H&Y) scaleand the Unified Parkinson’s Disease Rating Scale (UPDRS) part III-motor examina-tion. Using item 17 from the UPDRS part II-activities of daily living, PD patients wereclassified as ‘with pain’ (point 1e4) or ‘without pain’ (point 0) groups. Patients withsymptoms and signs compatible with an atypical parkinsonian syndrome wereexcluded. Exclusion criteria included clinical findings or nerve conduction evidenceof peripheral neuropathy or of any other disease potentially causing sensoryimpairment. Components of ES related SEPs are known to be changed by cognitivefunction. Consequently, patients with cognitive defects (Mini-Mental State Exami-nation score below 25/30), or taking analgesics or antidepressant treatment, werealso excluded. Each patient’s daily intake of anti-parkinsonian medications (e.g.,levodopa, dopamine agonist, monoamine oxidase type B inhibitor, catechol-O-methyl transferase and amantadine) was recorded. The total daily levodopa equiv-alent dose was calculated for each patient according to published conversionformulas [17]. All PD patients were examined in the on condition.

The Ethics Committee of Nagoya University School of Medicine approved allaspects of this study. Written informed consent for participation was obtained fromall subjects.

2.2. Preferential stimulation of Ad fibers by intra-epidermal needle electrode

We recorded pain-related SEPs using methodology described previously [5] anda custom-made needle electrode (Nihon-Koden Co. Ltd. Tokyo, Japan). Electricalstimulus was a current constant square wave pulse delivered at random intervalsbetween 0.1 Hz. Stimulus duration was 1.0 ms. Current intensity was at a pre-determined level which produced a definite pain sensation in each subject. Westimulated the face, upon the cheek, approximately 3 cm below the infra-orbitalmargin. Stimuli were applied in PD patients to the dominant side of motor symp-toms, and was applied consistently to the right side of control subjects. Stimulationproduced a well-defined pricking pain without definite tactile sensations. Stimulusintensity was under 0.4 mA which does not induce other pain receptors.

2.3. Recording of SEPs

An exploring electrode was placed at the Cz (vertex) according to the 10-20International system. The reference electrode was applied to the left earlobe. Wefocused upon evoked potential responses recorded from the Cz. Impedance for allelectrodes remained under 5 kU. Responses were recorded with a 0.5e50 Hzbandpass filter at a sampling rate of 1024 Hz. Analysis time was 1 s followingstimulation. In each stimulus condition, ten responses with approximately 10 srandomized stimulation intervals were collected and averaged in one trial. Inaddition, three trials were recorded over 2 min intervals in order to avoid habitu-ation. During SEP recording, the subjects laid on a bed in awarm and quiet room. SEPcomponents were identified on the basis of their latency and polarity and werelabeled in accordance with a previous report [6]. Peak-to-peak amplitude wasmeasured for the vertex biphasic SEP component (N1/P1). Since the number ofdetectable SEPs in PD is low, and in order to allow comparison with other diseases,we defined non-recordable SEPs as having an amplitude of zero.

Epochs in which signal variations were larger than 80 mV in the EEG wereexcluded from data analysis. After ensuring that our methodology was consistentand produced reproducible data, we conducted three trials which were then aver-aged for analysis.

2.4. Olfactory testing

The odor-identification performance of each subject was measured using theOdor Stick Identification Test for Japanese (OSIT-J, Daiichi Yakuhin, Co., Ltd., Tokyo,Japan), which consists of 12 odorants familiar to Japanese people. This test has been

successfully applied for the assessment of odor-identification ability in Japanese PDpatients and the precise protocol used has been described previously [18].

2.5. Statistical analysis

We calculated the mean and the standard deviation (SD) of all variables for allpatients and control subjects. The difference of age between PD patients andcontrols was determined with the student t-test. We used the c2 test to compare sexdistribution among groups. Stimulus intensity and parameters associated with pain-related SEPs between group differences were analyzed using analysis of variance.Spearman’s rank correlation was used to examine correlations between variousparameters of pain-related SEPs and H&Y scale, UPDRS part III, and OSIT-J score, andstimulus intensity. Differences in pain-related SEP parameters between PD patientswith and without pain were determined using the KruskaleWallis test. Statisticalcomputing was performed with JMP software, version 7. A value of p < 0.05 wasconsidered to define statistical significance.

3. Results

3.1. Clinical characteristics

The age of PD patients (66.7� 6.9 years) and controls (63.7� 6.4years) were not significantly different. Similarly, there was nosignificant difference identified in gender between PD patients andcontrols. Themean disease duration of PDwas 6.5� 5.1 years. In PDpatients, H&Y stagewas 2.3�1.1, and UPDRS partⅢwas 21.6� 10.7.Total daily levodopa equivalent dose was 343 � 331 mg. Twentyone of the 42 PD patients reported to be experiencing pain; OSIT-Jscore was 4.9 � 3.2, compared to 8.9 � 2.3 in control subjects,which was significantly lower (p < 0.01).

3.2. Pain-related SEP recordings

SEPs were not evoked in six PD patients. Stimulus intensitieswere not significantly different between PD patients and controls.Furthermore, stimulus intensities did not differ in PD patients withor without detectable SEPs. There were no significant differences inN1 or P1 latencies between PD patients (N1: 188.7 � 39.5 ms, P1:265.8 � 46.2 ms) and controls (N1: 183.5 � 42.1 ms, P1:255.5 � 53.0 ms). However, N1/P1 amplitudes were significantlylower in PD patients (9.0 � 5.8 mV) than in controls (13.7 � 6.4 mV)

Fig. 2. Pain-related somatosensory evoked potentials for representative PD patients (left) and control subjects (right).

T. Hara et al. / Parkinsonism and Related Disorders 19 (2013) 43e46 45

(p < 0.01) (Fig. 1). Fig. 2 shows waveforms in representative PDpatients and controls. There were significant correlations betweenN1/P1 amplitudes and disease duration (r ¼ �0.35, p < 0.05), H&Ystage (r ¼ �0.38, p < 0.05), and UPDRS part III (r ¼ �0.42, p < 0.01)(Fig. 3). UPDRS part III scores of patients without detectable SEPswere significantly higher than those of patients with detectableSEPs. N1/P1 amplitudes did not correlate with stimulus intensitiesin PD patients and control subjects. Pain-related SEP parametersdid not differ in PD patients with or without pain.

3.3. Correlation between pain-related SEPs and odor-identification

Smell identification scores on the OSIT-J correlated positivelywith N1/P1 amplitudes (r ¼ 0.41, p < 0.01) (Fig. 4) but did notcorrelate with N1 latency and P1 latency.

4. Discussion

This study represents the first investigation of pain-related SEPswith epidermal ES evoked by intra-epidermal needle electrodes inPD patients. Current data demonstrates that the amplitude of pain-related SEPs was significantly reduced in PD patients.

Speculation surrounding the functional implication of SEPamplitudes with ES is based upon current knowledge concerningcerebral regions. Inui et al. studied pain perception in healthysubjects by MEG using ES. The onset of N1 in pain-related SEPscorresponded approximately to the onset of activity in medialtemporal cortex (MT). Peak latencies of N1 and P1 corresponded toearly and late MT peak latencies. The entire wave form of N1/P1potential was similar to that of activity in the MT. These authorsconsidered that large parts of N1/P1 potential originated from

Fig. 3. Correlation between the amplitude of pain-related SEPs and diseas

activity within the MT [8]. The role of the MT region surroundingthe amygdala and hippocampus remains controversial. The amyg-dala is thought to contribute to the emotional processing of pain[19], while activation of the hippocampal formation appears to beenhanced when pain is not expected, or when pain stimulus isassociated with anxiety [20]. It is also known that the amygdalaexhibits significant pathological changes in PD, including atrophyand Lewy body formation [21]. We therefore speculate that theobserved reduction in pain-related SEP amplitude may reflectpathological abnormality of the amygdala in PD patients. Futherstudies, which compare PD with other parkisonian syndromes,such as multiple system atrophy are necessary to verify whetherreduction in pain-related SEPs amplitude is specific to PD.

Furthermore, the amplitudes of pain-related SEPs correlatedwith olfactory dysfunction. Olfactory dysfunction is an early clinicalmanifestation of PD [22]. The olfactory nerve fibers connect tomitral and tufted cells of the olfactory bulb, which project via theolfactory tract and tubercle projects to the olfactory cortex. Positronemission tomography studies have shown that odors have theability to activate the amygdala, piriform, orbitofrontal, insular, andcingulate cortices and thalamus [23]. Pathologically, Braak et al.found that Lewy body pathology is initially confined to the medullaoblongata and olfactory nucleus [14]. Moreover, a recent study byHawles et al. proposes that in PD, a neurotropic pathogen enters thebrain via two routes: (a) nasal, with anterograde progression intothe temporal lobe; namely, amygdala, piriform and periamygdalearcortex; and (b) gastric, with retrograde progression into the vagaldorsal motor nucleus of themedulla [24]. Collectively, these reportssuggest that the limbic systemmay be affected in the early phase ofPD patients. Thus, pain-related SEPs may be correlated to olfactorydysfunction in similar regions which are impaired in PD. Our

e duration (left, N ¼ 42), and UPDRSIII (right, N ¼ 38) in PD patients.

Fig. 4. Correlation between amplitude of pain-related SEPs and odor-identification inPD patients (N ¼ 42).

T. Hara et al. / Parkinsonism and Related Disorders 19 (2013) 43e4646

present study is the first to investigate the association betweenpain-related SEPs and olfactory dysfunction in PD patients.

There have been some interesting studies of pain-related SEPs inPD patients. Using CO2 laser-evoked potentials (LEPs), Tinazzi et al.reported abnormal processing of nociceptive inputs in PD [25].These authors also studied Ad pain fibers and recorded scalp CO2LEPs to hand skin in pain-free PD patients. PD patients and normalsubjects showed comparable LEP latencies, although amplitudewas significantly lower in PD patients than in controls. These earlierresults concurred with the results from our present study. CO2 LEPsrepresent a reliable neurophysiological method for assessing thefunction of nociceptive pathways. However, CO2 LEPs requireexpensive laser equipment, which is beyond the financial reach ofmany facilities. The newly-developed needle electrode used in thecurrent study was a very convenient method for the selectivestimulation of Ad fibers, since it was very simple, did not requirespecial apparatus, did not cause bleeding or burns, and minimizeduncomfortable feelings.

In addition, our results demonstrate that the amplitudes of pain-related SEPs were associated with disease duration and severity. Ithas previously been reported that significant negative correlationexists between olfactory dysfunction and disease duration andseverity [26]. Collectively, these findings also support that pain-related SEPs may be correlated to olfactory dysfunction in similarregions which are impaired in PD. On the other hand, pain-relatedSEP amplitude did not differ between PD patients with and withoutpain. Therefore, it is not necessarily the case that reduction in SEPamplitude directly reflects abnormality of pain sensation. Tinazziet al. demonstrated that LEP amplitude was significantly lower, andpain rating to laser stimulation significantly increased, in treated PDpatients than in controls. Consequently, these authors speculatedthat LEP might be related to the activation of pain-related inhibitorycircuits [27]. Althoughwe consider that our results may indicate thatpain-related SEPs represent the activity of pain inhibitory systems orthe emotional processing of pain, further study is needed to confirmthe clinical significance of reduced pain-related SEP amplitudes.

In this study,wedidnot assess the impact of dopaminergic drugs.However, Tinazzi et al. showed that in PD patients, L-dopa admin-istration yielded no significant change in LEP amplitude ascompared to the “off” condition. These results suggest abnormalnociceptive input processing in pain-free PD patients, whichappears to be independent of the clinical expression of parkinsonianmotor signs, and are not affected by dopaminergic stimulation [25].

In conclusion, we have shown that pain-related SEP amplitudeswith ES are reduced in PD patients. Furthermore, these reducedamplitudes correlate with olfactory dysfunction. Our resultssuggest that these findings may be associated with pathologicalabnormality of limbic systems.

References

[1] Ford B, Louis ED, Greene P, Fahn S. Oral and genital pain syndromes inParkinson’s disease. Mov Disord 1996;11:421e6.

[2] Djaldetti R, Shifrin A, Rogowski Z, Sprecher E, Melamed E, Yarnitsky D.Quantitative measurement of pain sensation in patients with Parkinsondisease. Neurology 2004;62:2171e5.

[3] Vela L, Lyons KE, Singer C, Lieberman AN. Pain-pressure threshold in patientswith Parkinson’s disease with and without dyskinesia. Parkinsonism Relat D2007;13:189e92.

[4] Brefel-Courbon C, Payoux P, Thalamas C, Ory F, Quelven I, Chollet F, et al. Effectof levodopa on pain threshold in Parkinson’s disease: a clinical and positronemission tomography study. Mov Disord 2005;20:1557e63.

[5] Inui K, Tran TD, Hoshiyama M, Kakigi R. Preferential stimulation of A deltafibers by intra-epidermal needle electrode in humans. Pain 2002;96:247e52.

[6] Valeriani M, Rambaud L, Mauguiere F. Scalp topography and dipolar sourcemodelling of potentials evoked by CO2 laser stimulation of the hand. ElectroenClin Neuro 1996;100:343e53.

[7] Handwerker HO, Kobal G. Psychophysiology of experimentally induced pain.Physiol Rev 1993;73:639e71.

[8] Inui K, Tran TD, Qiu Y, Wang X, Hoshiyama M, Kakigi R. A comparativemagnetoencephalographic study of cortical activations evoked by noxious andinnocuous somatosensory stimulations. Neuroscience 2003;120:235e48.

[9] Defazio G, Berardelli A, Fabbrini G, Martino D, Fincati E, Fiaschi A, et al. Pain asa nonmotor symptom of Parkinson disease: evidence from a case-controlstudy. Arch Neurol 2008;65:1191e4.

[10] O’Sullivan SS, Williams DR, Gallagher DA, Massey LA, Silveira-Moriyama L,Lees AJ. Nonmotor symptoms as presenting complaints in Parkinson’s disease:a clinicopathological study. Mov Disord 2008;23:101e6.

[11] Becker G, Muller A, Braune S, Buttner T, Benecke R, Greulich W, et al. Earlydiagnosis of Parkinson’s disease. J Neurol 2002;249(Suppl. 3):40e8. III.

[12] Berg D. Marker for a preclinical diagnosis of Parkinson’s disease as a basis forneuroprotection. J Neural Trans Suppl 2006:123e32.

[13] Siderowf A, Stern MB. Preclinical diagnosis of Parkinson’s disease: are wethere yet? Curr Neurol Neurosci Rep 2006;6:295e301.

[14] Braak H, Del Tredici K, Rub U, de Vos RA, Jansen Steur EN, Braak E. Staging ofbrain pathology related to sporadic Parkinson’s disease. Neurobiol Aging2003;24:197e211.

[15] Baba T, Takeda A, Kikuchi A, Nishio Y, Hosokai Y, Hirayama K, et al. Associationof olfactory dysfunction and brain. Metabolism in Parkinson’s disease. MovDisord 2011;26:621e8.

[16] Hughes AJ, Daniel SE, Kilford L, Lees AJ. Accuracy of clinical diagnosis ofidiopathic Parkinson’s disease: a clinico-pathological study of 100 cases.J Neurol Neurosur and Psychiatry 1992;55:181e4.

[17] Tomlinson CL, Stowe R, Patel S, Rick C, Gray R, Clarke CE. Systematic review oflevodopa dose equivalency reporting in Parkinson’s disease. Mov Disord2010;25:2649e53.

[18] Oka H, Toyoda C, Yogo M, Mochio S. Olfactory dysfunction and cardiovasculardysautonomia in Parkinson’s disease. J Neurol 2010;257:969e76.

[19] Bornhovd K, Quante M, Glauche V, Bromm B, Weiller C, Buchel C. Painfulstimuli evoke different stimulus-response functions in the amygdala,prefrontal, insula and somatosensory cortex: a single-trial fMRI study. Brain2002;125:1326e36.

[20] Ploghaus A, Narain C, Beckmann CF, Clare S, Bantick S, Wise R, et al. Exacer-bation of pain by anxiety is associated with activity in a hippocampalnetwork. J Neurosci 2001;21:9896e903.

[21] Harding AJ, Stimson E, Henderson JM, Halliday GM. Clinical correlates ofselective pathology in the amygdala of patients with Parkinson’s disease.Brain 2002;125:2431e45.

[22] Doty RL, Deems DA, Stellar S. Olfactory dysfunction in parkinsonism: a generaldeficit unrelated to neurologic signs, disease stage, or disease duration.Neurology 1988;38:1237e44.

[23] Savic I, Gulyas B, Larsson M, Roland P. Olfactory functions are mediated byparallel and hierarchical processing. Neuron 2000;26:735e45.

[24] Hawkes CH, Del Tredici K, Braak H. Parkinson’s disease: the dual hit theoryrevisited. Ann N Y Acad Sci 2009;1170:615e22.

[25] Tinazzi M, Del Vesco C, Defazio G, Fincati E, Smania N, Moretto G, et al.Abnormal processing of the nociceptive input in Parkinson’s disease: a studywith CO(2) laser evoked potentials. Pain 2008;136:117e24.

[26] Berendse HW, Roos DS, Raijmakers P, Doty RL. Motor and non-motorcorrelates of olfactory dysfunction in Parkinson’s disease. J Neurol Sci 2011;310:21e4.

[27] Tinazzi M, Recchia S, Simonetto S, Defazio G, Tamburin S, Moretto G, et al.Hyperalgesia and laser evoked potentials alterations in hemiparkinson:evidence for an abnormal nociceptive processing. J Neurol Sci 2009;276:153e8.