77
Journal of Parkinson’s Disease 4 (2014) 77–87 DOI 10.3233/JPD-130307 IOS Press
Research Report
Prepulse Inhibition is Associated with Attention, Processing Speed, and 123 I-FP-CIT SPECT in Parkinson’s Disease Marielle Zoetmuldera,b,∗ , Heidi B. Biernata , Miki Nikolicc , Lise Korboa , Lars Fribergd and Poul J. Jennumb a Department
of Neurology, Bispebjerg University Hospital, Copenhagen NV, Denmark Center for Sleep Medicine, Department of Clinical Neurophysiology, Glostrup Hospital, Glostrup, Denmark c Department of Clinical Neurophysiology, Glostrup Hospital, Glostrup, Denmark d Department of Clinical Physiology and Nuclear Medicine, Bispebjerg University Hospital, Copenhagen NV, Denmark b Danish
Abstract. Background: Prepulse inhibition is a measure of sensorimotor gating, which reflects the ability to filter or ‘gate’ irrelevant information. Prepulse inhibition is dramatically altered in basal ganglia disorders associated with dysfunction in the midbrain dopaminergic system, and corresponding cognitive information processing deficits such as slowed processing speed. Parkinson’s disease is characterised by the degeneration of the midbrain dopaminergic system and is associated with cognitive dysfunction, including slowed information processing. Although sensorimotor processes in Parkinson’s disease have been extensively studied in relation to motor function, less is known about the potential role of sensorimotor processes in cognitive function. Objective: We investigated the relationship between prepulse inhibition, cognition and nigrostriatal dysfunction, as measured with 123 I-FP-CIT-SPECT scanning, in patients with Parkinson’s disease. Methods: 38 Parkinson patients were assessed with prepulse inhibition, neuropsychological tests, and neurological investigation. A subset of these patients underwent 123 I-FP-CIT-SPECT scanning. Results: Patients with a higher level of prepulse inhibition performed better on cognitive measures tapping attention and processing speed than patients with a lower level of prepulse inhibition. Furthermore, there were significant correlations between prepulse inhibition and 123 I-FP-CIT uptake in the striatum. Conclusions: Our results suggest that the level of prepulse inhibition is related to the efficiency of information processing in Parkinson’s disease, and to the density of dopamine transporters in the striatum. Keywords: Prepulse inhibition, Dopamine system, SPECT, Parkinson’s disease
INTRODUCTION ∗ Correspondence
to: Marielle Zoetmulder, Glostrup Hospital, Danish Center for Sleep Medicine, Department of Clinical Neurophysiology, Nordre Ringvej 57, 2600 Glostrup, Denmark. Tel.: +45 22672707; E-mail: [email protected].; Bispebjerg Hospital, Department of Neurology, Bispebjerg Bakke 23, 2400 Copenhagen NV, Denmark. Tel.: +45 35312845; E-mail: marielle [email protected].
Prepulse inhibition (PPI) of the startle reflex is a well characterised and extensively studied model of early stage information processing that can be measured in vertebrates, rodents, nonhuman primates, and humans [1, 2]. PPI is the attenuation of the amplitude of the startle reflex in response to sudden intense stimuli (pulse)
ISSN 1877-7171/14/$27.50 © 2014 – IOS Press and the authors. All rights reserved
78
M. Zoetmulder et al. / PPI, Cognition, and Dopamine in PD
if preceded by a weaker sensory stimulus (prepulse). The amount of inhibition in PPI is used as an operational measure of sensorimotor gating [3]. Deficits in PPI have been observed in basal ganglia disorders associated with dysfunction in the midbrain dopaminergic system, and corresponding cognitive information processing deficits. These disorders include schizophrenia [4, 5], Tourette syndrome [6, 7] and Huntington’s disease [8–10]. The midbrain dopaminergic system seems to play a critical role in the regulation and modulation of PPI. Down-regulation of transcription factors essential for the expression of the midbrain dopaminergic phenotype, such as Nurr1, results in decreased expression of tyrosine hydroxylase in dopaminergic neurons in the substantia nigra pars compacta and ventral tegmental area, and decreased density of dopamine transporters (DAT) [11–14]. The resulting reduction of dopamine levels in the dorsal and ventral striatum and the reduced density of dopamine transporters have been associated with PPI deficits [15, 16]. Moreover, animals with lesions of the dorsal, dorsomedial or ventral striatum show PPI deficits [17–20]. Parkinson’s disease is a neurodegenerative disorder characterized by motor symptoms, and non-motor symptoms including cognitive dysfunction, such as slowed information processing and attention deficits. In PD, degenerative processes target midbrain dopamine-producing neurons in the substantia nigra and other dopaminerich areas [21]. Reduced levels [22, 23] and genetic alterations of midbrain dopamine transcription factors, such as Nurr1, have been found in Parkinson’s disease [24–27]. In addition, progressive decline in the density of the dopamine transporter is characteristic of the pathogenesis of PD [28]. Sensorimotor processes in Parkinson’s disease have been extensively studied in relation to motor function and dopamine [29, 30]. However, sensorimotor processes have been less extensively studied in relation to cognitive function and striatal dopamine in Parkinson’s disease. Impairment of information processing speed, attention, working memory and executive function are considered core deficits in PD [31–33]. These cognitive deficits have been associated with reduced density of the dopamine transporter in striatum [34, 35]. Striatal dopamine function has been shown to promote cognitive flexibility by gating rapidly changing sensory information [36–38]. Disruption of gating processes in the striatum may lead to inappropriate cognitive and behavioural responses [39]. Human studies have found that healthy subjects with high PPI have faster
execution times and better planning abilities than subjects with low PPI [40]. In addition, animals with a low level of PPI appear to be less able to modulate behaviour based upon new changing information than animals with a high level of PPI [41]. These studies suggest a possible relationship between PPI, cognitive function and the presynaptic components of the striatal dopamine system in patients with PD. We conducted a study with the objective to investigate the relationship between PPI, cognitive function, and the dopamine transporters, measured by 123 I-FP-CIT brain SPECT, in patients with PD. Our hypothesis was that better sensorimotor gating, as reflected by greater PPI, will be associated with higher 123 I-FP-CIT uptake in striatum and enhanced performance in cognitive function, including processing speed and attention.
MATERIALS AND METHODS Participants Thirty-eight patients with PD were included in the study. They were recruited from the Movement Disorder Clinic at Bispebjerg Hospital, and from the patient support group of Parkinson’s disease in Denmark. Patients were put through a research battery consisting of PPI, neuropsychological tests, neurological examination, routine blood analysis and a hearing test. A subset (N = 23) of these patients, for whom no selection criteria were applied, underwent 123 I-FP-CIT-SPECT scanning. The average time for each patient to complete all examinations was 7 months (SD 5.20 months), the duration being determined by the clinic’s waiting time. Neuropsychological examination and PPI were performed the same day in the morning in fixed order. Inclusion criteria were: idiopathic PD, age between 45 and 75 years, Mini Mental State Examination >25. Exclusion criteria were dementia, a history of psychiatric disorder and family history of mental illness in first-degree relatives, abuse of alcohol and drugs, hearing loss (>40 dB(A) at 1000 Hz), head trauma and use of acetylcholinesterase inhibitors. Patients were receiving optimal medical therapy for Parkinson’s disease. Patients were included according to the UK Parkinson’s Disease Society Brain Bank criteria, and were evaluated using the Unified Parkinson’s Disease Rating Scale (UPDRS) and Hoehn and Yahr rating scale for PD (0–5). The patients gave oral and written consent, according to the Declaration of Helsinki. The study was approved by the local ethics committee (reference code: H-A-2007-0120).
M. Zoetmulder et al. / PPI, Cognition, and Dopamine in PD
A
B
C
D
79
Fig. 1. This figure shows prepulse inhibition of the auditory blink reflex. It explains the effect of the prepulse-pulse (PP-P) inter-stimulus-intervals (ISI) on the following EMG-response of the orbicularis occuli muscle. The figure shows four different prepulse-pulse ISI’s: A: 30 ms 85 dB prepulse-pulse trial; B: 60 ms 85 dB prepulse-pulse trial; C: 120 ms 85 dB prepulse-pulse trial; D: 300 ms 85 dB prepulse-pulse trial; ms = milli seconds. The EMG-response is measured in micro-Volt (V).
Prepulse inhibition An electromyographic startle system (EMG SRHLAB, San Diego Instruments, San Diego, CA, USA) was used to examine the eye blink component of the acoustic startle response. Two miniature silver/silver chloride electrodes were placed over the right orbicularis oculi muscle, measuring EMG-responses, and a ground electrode was placed behind the right ear over the mastoid. Acoustic stimuli were administered binaurally through Sony MDR-V6 headphones. Subjects sat upright and were instructed to relax, stay awake and look at a fixating point straight ahead. Pulses consisted of 40 ms, 115 dB white noise bursts, and prepulses consisted of 20 ms, 75 and 85 dB white noise bursts over 70 dB background noise. The trials with the 85 dB white noise burst were included for further assessment in this study, as studies usually employ these characteristics [2]. Three acoustic startle-eliciting stimuli against a background of 70 dB broadband white noise were administered before each session to evaluate the quality of the signal and adjust the amplifier level. The recording period began with
a 3-min acclimatisation period of 70 dB broadband white noise, which continued throughout the entire session as background noise. The trials began with four startle-eliciting stimuli of 115 dB white noise for 40 ms (pulse-alone), followed by six blocks of 11 trials. Each block consisted of three pulse-alone stimuli, and eight prepulse-pulse trials. Four lead intervals (onset to onset) were used (30, 60, 120 and 300 ms). For each interval there were six trials with the 75 dB prepulse and six trials with the 85 dB prepulse. This study focuses on the 85 dB-PPI trials. The mean inter-trial interval was 15 s. The session ended with four pulsealone trials. None of the subjects had more than three (out of six) trials per trial type discarded in any single session. Bandpass filters were set between low (30 Hz) and high (1000 Hz). A scoring program was computed in MATLAB to analyse the EMG-responses visually and manually. This program corrected for differences in amplification. Baseline was calculated by averaging the individual EMG-signals of the 6 trials (of e.g. 120 ms inter-stimulus-interval) with MATLAB. This averaging was based on the first 100 ms (interval from
80
M. Zoetmulder et al. / PPI, Cognition, and Dopamine in PD Table 1 Demographic and clinical data scores (mean ± SD) of the entire PD group (n = 38) and the subset of PD patients assessed with 123 I-FP-CIT (n = 23) Demographic and clinical data Age (years) Sex (M/F) Age at disease onset (years) Education (years) Duration of disease (years) On-UPDRS-III score (0–108) Stage of disease (Hoehn & Yahr) LED levodopa mg LED agonists mg LED levodopa+agonist mg Total LED mg Beck’s depression scale PPI 30 ms 85 dB PPI 60 ms 85 dB PPI 120 ms 85dB PPI 300 ms 85dB mSPECT-striatum mSPECT-putamen mSPECT-caudate Mean-lowest left or right caudate Mean-lowest left or right putamen Mean-highest left or right caudate Mean-highest left or right putamen
PD patients (N = 23)
PD patients (N = 38)
P
62.70 ± 7.86 13/10 57.19 ± 8.54 10.32 ± 1.56 5.67 ± 4.21 23.60 ± 10.13 2.09 ± 0.80 365 ± 448 163 ± 115 528 ± 500 583 ± 521 6.72 ± 3.85 43.68 ± 23.71 56.50 ± 23.77 67.47 ± 21.50 54.94 ± 21.60 2.79 ± 1.11 1.89 ± 0.86 3.88 ± 1.44 3.59 ± 1.31 1.54 ± 0.72 4.17 ± 1.59 2.22 ± 1.04
62.84 ± 6.62 18/20 57.60 ± 7.62 9.88 ± 1.65 5.66 ± 3.95 22.80 ± 8.90 2.03 ± 0.72 351 ± 385 144 ± 113 495 ± 419 552 ± 436 6.84 ± 4.83 42.96 ± 23.97 54.12 ± 22.85 66.56 ± 20.76 55.85 ± 20.55 N/A N/A N/A N/A N/A N/A N/A
NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS
LED: Levodopa equivalent; PPI: Prepulse inhibition; UPDRS: Unified Parkinson’s disease rating scale. P ≤ 0.05; NS: not significant.
0–100 ms) with EMG and background noise, before a stimulus was presented. Afterwards these 6 ‘means’ were averaged to one baseline estimate for the given inter-stimulus-interval. The amplitude was measured as the maximum value in the analysis-window which ranged from 100 ms–600 ms. The amplitude was also a mean of the 6 trials per inter-stimulus-interval (30, 60, 120, 300 ms). Mean startle magnitude was calculated by averaging responses on pulse-alone trials for the initial four pulse-alone trials. An example of raw EMG-responses is given in Fig. 1. PPI (%) was calculated using the formula: [(Amplitude pulse-alone–Amplitude prepulse– pulse)/Amplitude pulse-alone]×100. Trials with excessive baseline activity 100 ms before the expected response were excluded to minimize confounding of possible spontaneous and voluntary eyeblink activity upon startle. Patients were excluded from the analysis for low startle response if their startle magnitude was zero in at least 50 per cent of the first four pulse-trials, or if more than three of the six prepulse-pulse trials were discarded. 123 I-FP-CIT
brain SPECT
123 I-FP-CIT
brain SPECT scans were carried out with a Prism XP 3000 (Picker/Phillips) triple-headed
Table 2 Neuropsychological raw scores (mean ± SD) in the entire PD group Neuropsychological raw scores Rapid visual processing Symbol digit modalities test Stroop word Stroop colour-word Trail-making A Trail-making B Stockings of Cambridge, problems solved Intra-extra dimensional set shift, stages completed Rey auditory verbal learning test, immediate recall Rey auditory verbal learning test, delayed recall Danish national reading test Addenbrooke’s cognitive examination
PD group (N = 38) 0.89 ± 0.05 47.79 ± 9.69 84.66 ± 13.15 36.05 ± 8.96 38.50 ± 11.45 97.89 ± 40.42 7.71 ± 1.88 8.31 ± 0.89 41.97 ± 9.26 8.69 ± 2.93 32.61 ± 9.50 91.19 ± 4.66
Raw score data, uncorrected for age and education are reported.
gamma camera equipped with low-energy, ultra-high resolution fan beam collimators. The SPECT-scanning was performed according to guidelines [42]. The SPECT-scanning started exactly 3 hours after patients had received an intravenous injection of 185 MBq 123 IFP-CIT, when steady state within the striatum had been achieved. 123 I-FP-CIT SPECT scans were analysed with an automatic 3D quantification tool (Exini DAT,
M. Zoetmulder et al. / PPI, Cognition, and Dopamine in PD Table 3 Factor structure of the neuropsychological tests-scores in PD (n = 38) Rapid visual processing Symbol digit modalities test Stroop word Stroop colour-word Trail-making A Trail-making B Stockings of Cambridge, problems solved Intra-extra dimensional set shift, stages completed Rey auditory verbal learning test, immediate recall Rey auditory verbal learning test, delayed recall
Factor 1
Factor 2
Factor 3
0.650 0.610 0.596 0.815 −0.689 −0.835 0.073
0.015 0.487 0.461 −0.024 −0.169 −0.271 0.126
0.405 0.247 0.071 0.013 −0.268 −0.056 0.825
0.205
0.183
0.762
0.255
0.783
0.323
0.039
0.947
0.068
Factor loadings greater than 0.500 in bold.
EXINI Diagnostics AB, Lund, Sweden). The quantification module automatically calculates specific uptake in the striatum, putamen and caudatus, and the putamen/caudate ratio for each patient’s brain images. It delineates the cerebral cortex as the background region. Cognitive assessment The same day as patients were assessed with PPI, we administered a neuropsychological battery that was designed to include measures sensitive to functional deficits of the striatal-frontal pathway and dopamine transporter [43, 44]. Attention and processing speed were rated by performance on several tasks: the symbol digit modalities test, in which the subject matches numbers with symbols; the trail-making test, in which the subject draws lines connecting consecutive numbers or numbers that alternate with letters; and the Stroop interference test, in which the subject reads colour names printed in incongruent ink colours and has to suppress the tendency to say the word instead of the colour; the CANTAB rapid visual processing test, which is a serial target detection task that requires sustained attention and working memory for its execution. Memory was rated by performance on the Rey auditory verbal learning test, in which the subject has to learn and recall lists of unrelated words immediately and after a time-delay. Executive function was rated by performance on the CANTAB stockings of Cambridge, which is a problem-solving test, and intra-extra dimensional set shift, an analogue of the Wisconsin card sorting test. Premorbid verbal intelligence was assessed with the Danish version of the national adult reading test (DART) [45]. In addition, patients were
81
assessed with Addenbrooke’s cognitive examination (ACE), which measures a patient’s global cognitive performance [46]. Data analysis and statistics Due to problems of multicollinearity among the neuropsychological variables, the PPI variables, and the four SPECT variables, preliminary analyses were performed on the original variables to minimize multicollinearity and reduce the number of variables considered in further statistical analysis. As a first step, factor analysis with varimax rotation was applied to the 10 neuropsychological measures in the 38 subjects to identify those scores expressing a similar proportion of the total variance. Factor analysis identified three cognitive factors, as detailed in the results section in Table 3. A conventional threshold of 0.5 was applied to factor loadings (expressing the factor variable correlation). The four PPI-85 dB trials (30 ms 85 dB, 60 ms 85 dB, 120 ms 85 dB, and 300 ms 85 dB) were positively correlated (r = 0.40–0.71 in the PD group; N = 38. After Bonferroni testing (0.05/6; P = 0.0083), all correlations remained significant, see Fig. 3. Therefore, these four PPI values were averaged in each subject, and a mean PPI value (mPPI-85 dB) was computed. The four SPECT variables (caudate and putamen uptake in each side) were also highly correlated (r = 0.71–0.87), as found in another study of Nobili et al. (2010), because these variables express the global nigrostriatal function in a given subject, normalized for background uptake. However, as the caudate is more associated with cognitive function, and putamen more with motor function, the uptake values were averaged for caudate nucleus and putamen in each subject, and a mean uptake for caudate (mSPECT-caudate) and putamen (mSPECT-putamen) was computed. To analyse the relationship between PPI and cognitive measures, Pearson correlation coefficient was used. We applied Bonferroni correction of the threshold of statistical significance in the case of multiple comparisons. A subset of patients (N = 23) were assessed with 123 IFP-CIT brain scanning. The relationship between PPI and cognitive variables, and 123 I-FP-CIT uptake, was assessed with Kendall’s tau-b. To investigate the predictive value of single variables for cognitive measures further a regression analysis with stepwise exclusion was performed. Variables in Table 2 were normally distributed and not skewed, although the Stroop colourword variable showed kurtosis (Z = 2.27), exceeding the Z-score of 1.96, at the 5% significance-level. Statistical analysis was performed using SPSS version 17
82
M. Zoetmulder et al. / PPI, Cognition, and Dopamine in PD
(SPSS Inc., Chicago, IL, USA). A 5% significance level was adopted. RESULTS Demographic and clinical characteristics of all 38 PD patients are listed in Table 1, and the means and confidence intervals of the PPI inter-stimulus-trials are further depicted in Fig. 2. Neuropsychological tests and composite scores The neuropsychological raw scores of the PD group are presented in Table 2. Factor analysis yielded three factors, see Table 3. The first factor represented RVP, SDMT, Stroop colour-word, and the trail-making tests A and B. This factor included time-based tests, tapping attention and processing speed (composite processingspeed). The second factor included subscores of the memory test RAVLT (composite memory). The third factor included the SOCps and the IEDsc, which are measures associated with executive function (composite executive). Relationship between mPPI-85 dB and cognitive composite scores
indicates that patients with less PPI were significantly slower, or had worse performance, on tests included in the composite processing-speed than patients with higher PPI, whereas we found no significant differences between mPPI-85 dB and the composite memory and executive scores. Furthermore, the composite mPPI-85 dB was significantly correlated to ACE scores (Pearson r = .362, P < 0.016). Given the significant correlation between mPPI85 dB and the composite processing speed, we performed a correlation analysis with Bonferonni testing to analyse which of the PPI-85 dB interstimulus trials accounted for this finding. This analysis showed a direct correlation between the composite processing speed and PPI-30 ms 85 dB (Pearson r = 0.370, P = 0.014), PPI-60 ms 85 dB (Pearson r = 0.669, P = 0.000, see Fig. 4), and PPI-120 ms 85 dB (Pearson r = 0.354, P = 0.020), while no significant correlation was found for the PPI-300 ms 85 dB. After Bonferroni testing (0.05/4 = 0.0125), only the PPI 60 ms 85 dB remained significantly correlated, see Fig. 4. The different measures of Levodopa equivalent dosage were not significantly associated with PPI-60 ms 85 dB. Relationship between PPI 60 ms 85 dB, SPECT, and composite score processing-speed
123 I-FP-CIT
In the entire group, a significant correlation was found between mPPI-85 dB and the composite processing-speed (Pearson r = 0.47, P = 0.002), which
Fig. 2. Percent prepulse inhibition (PPI) of the acoustic startle response in patients with Parkinson’s disease. Data represent the mean PPI for the four prepulse-pulse inter-stimulus-intervals, and the error bars describe the confidence interval.
To evaluate the relationship between sensorimotor gating, cognitive function and the nigrostriatal dopaminergic system, we looked at the subset of patients who, in addition to PPI and cognitive testing, had undergone 123 I-FP-CIT SPECT (N = 23), for demographic and clinical data see Table 1. We found that the PPI-60 ms 85 dB inter-stimulus-interval was correlated with the composite processing-speed in this subgroup (Kendall’s tau-b = 0.534, P =
Journal of Parkinson’s Disease 4 (2014) 77–87 DOI 10.3233/JPD-130307 IOS Press
Research Report
Prepulse Inhibition is Associated with Attention, Processing Speed, and 123 I-FP-CIT SPECT in Parkinson’s Disease Marielle Zoetmuldera,b,∗ , Heidi B. Biernata , Miki Nikolicc , Lise Korboa , Lars Fribergd and Poul J. Jennumb a Department
of Neurology, Bispebjerg University Hospital, Copenhagen NV, Denmark Center for Sleep Medicine, Department of Clinical Neurophysiology, Glostrup Hospital, Glostrup, Denmark c Department of Clinical Neurophysiology, Glostrup Hospital, Glostrup, Denmark d Department of Clinical Physiology and Nuclear Medicine, Bispebjerg University Hospital, Copenhagen NV, Denmark b Danish
Abstract. Background: Prepulse inhibition is a measure of sensorimotor gating, which reflects the ability to filter or ‘gate’ irrelevant information. Prepulse inhibition is dramatically altered in basal ganglia disorders associated with dysfunction in the midbrain dopaminergic system, and corresponding cognitive information processing deficits such as slowed processing speed. Parkinson’s disease is characterised by the degeneration of the midbrain dopaminergic system and is associated with cognitive dysfunction, including slowed information processing. Although sensorimotor processes in Parkinson’s disease have been extensively studied in relation to motor function, less is known about the potential role of sensorimotor processes in cognitive function. Objective: We investigated the relationship between prepulse inhibition, cognition and nigrostriatal dysfunction, as measured with 123 I-FP-CIT-SPECT scanning, in patients with Parkinson’s disease. Methods: 38 Parkinson patients were assessed with prepulse inhibition, neuropsychological tests, and neurological investigation. A subset of these patients underwent 123 I-FP-CIT-SPECT scanning. Results: Patients with a higher level of prepulse inhibition performed better on cognitive measures tapping attention and processing speed than patients with a lower level of prepulse inhibition. Furthermore, there were significant correlations between prepulse inhibition and 123 I-FP-CIT uptake in the striatum. Conclusions: Our results suggest that the level of prepulse inhibition is related to the efficiency of information processing in Parkinson’s disease, and to the density of dopamine transporters in the striatum. Keywords: Prepulse inhibition, Dopamine system, SPECT, Parkinson’s disease
INTRODUCTION ∗ Correspondence
to: Marielle Zoetmulder, Glostrup Hospital, Danish Center for Sleep Medicine, Department of Clinical Neurophysiology, Nordre Ringvej 57, 2600 Glostrup, Denmark. Tel.: +45 22672707; E-mail: [email protected].; Bispebjerg Hospital, Department of Neurology, Bispebjerg Bakke 23, 2400 Copenhagen NV, Denmark. Tel.: +45 35312845; E-mail: marielle [email protected].
Prepulse inhibition (PPI) of the startle reflex is a well characterised and extensively studied model of early stage information processing that can be measured in vertebrates, rodents, nonhuman primates, and humans [1, 2]. PPI is the attenuation of the amplitude of the startle reflex in response to sudden intense stimuli (pulse)
ISSN 1877-7171/14/$27.50 © 2014 – IOS Press and the authors. All rights reserved
78
M. Zoetmulder et al. / PPI, Cognition, and Dopamine in PD
if preceded by a weaker sensory stimulus (prepulse). The amount of inhibition in PPI is used as an operational measure of sensorimotor gating [3]. Deficits in PPI have been observed in basal ganglia disorders associated with dysfunction in the midbrain dopaminergic system, and corresponding cognitive information processing deficits. These disorders include schizophrenia [4, 5], Tourette syndrome [6, 7] and Huntington’s disease [8–10]. The midbrain dopaminergic system seems to play a critical role in the regulation and modulation of PPI. Down-regulation of transcription factors essential for the expression of the midbrain dopaminergic phenotype, such as Nurr1, results in decreased expression of tyrosine hydroxylase in dopaminergic neurons in the substantia nigra pars compacta and ventral tegmental area, and decreased density of dopamine transporters (DAT) [11–14]. The resulting reduction of dopamine levels in the dorsal and ventral striatum and the reduced density of dopamine transporters have been associated with PPI deficits [15, 16]. Moreover, animals with lesions of the dorsal, dorsomedial or ventral striatum show PPI deficits [17–20]. Parkinson’s disease is a neurodegenerative disorder characterized by motor symptoms, and non-motor symptoms including cognitive dysfunction, such as slowed information processing and attention deficits. In PD, degenerative processes target midbrain dopamine-producing neurons in the substantia nigra and other dopaminerich areas [21]. Reduced levels [22, 23] and genetic alterations of midbrain dopamine transcription factors, such as Nurr1, have been found in Parkinson’s disease [24–27]. In addition, progressive decline in the density of the dopamine transporter is characteristic of the pathogenesis of PD [28]. Sensorimotor processes in Parkinson’s disease have been extensively studied in relation to motor function and dopamine [29, 30]. However, sensorimotor processes have been less extensively studied in relation to cognitive function and striatal dopamine in Parkinson’s disease. Impairment of information processing speed, attention, working memory and executive function are considered core deficits in PD [31–33]. These cognitive deficits have been associated with reduced density of the dopamine transporter in striatum [34, 35]. Striatal dopamine function has been shown to promote cognitive flexibility by gating rapidly changing sensory information [36–38]. Disruption of gating processes in the striatum may lead to inappropriate cognitive and behavioural responses [39]. Human studies have found that healthy subjects with high PPI have faster
execution times and better planning abilities than subjects with low PPI [40]. In addition, animals with a low level of PPI appear to be less able to modulate behaviour based upon new changing information than animals with a high level of PPI [41]. These studies suggest a possible relationship between PPI, cognitive function and the presynaptic components of the striatal dopamine system in patients with PD. We conducted a study with the objective to investigate the relationship between PPI, cognitive function, and the dopamine transporters, measured by 123 I-FP-CIT brain SPECT, in patients with PD. Our hypothesis was that better sensorimotor gating, as reflected by greater PPI, will be associated with higher 123 I-FP-CIT uptake in striatum and enhanced performance in cognitive function, including processing speed and attention.
MATERIALS AND METHODS Participants Thirty-eight patients with PD were included in the study. They were recruited from the Movement Disorder Clinic at Bispebjerg Hospital, and from the patient support group of Parkinson’s disease in Denmark. Patients were put through a research battery consisting of PPI, neuropsychological tests, neurological examination, routine blood analysis and a hearing test. A subset (N = 23) of these patients, for whom no selection criteria were applied, underwent 123 I-FP-CIT-SPECT scanning. The average time for each patient to complete all examinations was 7 months (SD 5.20 months), the duration being determined by the clinic’s waiting time. Neuropsychological examination and PPI were performed the same day in the morning in fixed order. Inclusion criteria were: idiopathic PD, age between 45 and 75 years, Mini Mental State Examination >25. Exclusion criteria were dementia, a history of psychiatric disorder and family history of mental illness in first-degree relatives, abuse of alcohol and drugs, hearing loss (>40 dB(A) at 1000 Hz), head trauma and use of acetylcholinesterase inhibitors. Patients were receiving optimal medical therapy for Parkinson’s disease. Patients were included according to the UK Parkinson’s Disease Society Brain Bank criteria, and were evaluated using the Unified Parkinson’s Disease Rating Scale (UPDRS) and Hoehn and Yahr rating scale for PD (0–5). The patients gave oral and written consent, according to the Declaration of Helsinki. The study was approved by the local ethics committee (reference code: H-A-2007-0120).
M. Zoetmulder et al. / PPI, Cognition, and Dopamine in PD
A
B
C
D
79
Fig. 1. This figure shows prepulse inhibition of the auditory blink reflex. It explains the effect of the prepulse-pulse (PP-P) inter-stimulus-intervals (ISI) on the following EMG-response of the orbicularis occuli muscle. The figure shows four different prepulse-pulse ISI’s: A: 30 ms 85 dB prepulse-pulse trial; B: 60 ms 85 dB prepulse-pulse trial; C: 120 ms 85 dB prepulse-pulse trial; D: 300 ms 85 dB prepulse-pulse trial; ms = milli seconds. The EMG-response is measured in micro-Volt (V).
Prepulse inhibition An electromyographic startle system (EMG SRHLAB, San Diego Instruments, San Diego, CA, USA) was used to examine the eye blink component of the acoustic startle response. Two miniature silver/silver chloride electrodes were placed over the right orbicularis oculi muscle, measuring EMG-responses, and a ground electrode was placed behind the right ear over the mastoid. Acoustic stimuli were administered binaurally through Sony MDR-V6 headphones. Subjects sat upright and were instructed to relax, stay awake and look at a fixating point straight ahead. Pulses consisted of 40 ms, 115 dB white noise bursts, and prepulses consisted of 20 ms, 75 and 85 dB white noise bursts over 70 dB background noise. The trials with the 85 dB white noise burst were included for further assessment in this study, as studies usually employ these characteristics [2]. Three acoustic startle-eliciting stimuli against a background of 70 dB broadband white noise were administered before each session to evaluate the quality of the signal and adjust the amplifier level. The recording period began with
a 3-min acclimatisation period of 70 dB broadband white noise, which continued throughout the entire session as background noise. The trials began with four startle-eliciting stimuli of 115 dB white noise for 40 ms (pulse-alone), followed by six blocks of 11 trials. Each block consisted of three pulse-alone stimuli, and eight prepulse-pulse trials. Four lead intervals (onset to onset) were used (30, 60, 120 and 300 ms). For each interval there were six trials with the 75 dB prepulse and six trials with the 85 dB prepulse. This study focuses on the 85 dB-PPI trials. The mean inter-trial interval was 15 s. The session ended with four pulsealone trials. None of the subjects had more than three (out of six) trials per trial type discarded in any single session. Bandpass filters were set between low (30 Hz) and high (1000 Hz). A scoring program was computed in MATLAB to analyse the EMG-responses visually and manually. This program corrected for differences in amplification. Baseline was calculated by averaging the individual EMG-signals of the 6 trials (of e.g. 120 ms inter-stimulus-interval) with MATLAB. This averaging was based on the first 100 ms (interval from
80
M. Zoetmulder et al. / PPI, Cognition, and Dopamine in PD Table 1 Demographic and clinical data scores (mean ± SD) of the entire PD group (n = 38) and the subset of PD patients assessed with 123 I-FP-CIT (n = 23) Demographic and clinical data Age (years) Sex (M/F) Age at disease onset (years) Education (years) Duration of disease (years) On-UPDRS-III score (0–108) Stage of disease (Hoehn & Yahr) LED levodopa mg LED agonists mg LED levodopa+agonist mg Total LED mg Beck’s depression scale PPI 30 ms 85 dB PPI 60 ms 85 dB PPI 120 ms 85dB PPI 300 ms 85dB mSPECT-striatum mSPECT-putamen mSPECT-caudate Mean-lowest left or right caudate Mean-lowest left or right putamen Mean-highest left or right caudate Mean-highest left or right putamen
PD patients (N = 23)
PD patients (N = 38)
P
62.70 ± 7.86 13/10 57.19 ± 8.54 10.32 ± 1.56 5.67 ± 4.21 23.60 ± 10.13 2.09 ± 0.80 365 ± 448 163 ± 115 528 ± 500 583 ± 521 6.72 ± 3.85 43.68 ± 23.71 56.50 ± 23.77 67.47 ± 21.50 54.94 ± 21.60 2.79 ± 1.11 1.89 ± 0.86 3.88 ± 1.44 3.59 ± 1.31 1.54 ± 0.72 4.17 ± 1.59 2.22 ± 1.04
62.84 ± 6.62 18/20 57.60 ± 7.62 9.88 ± 1.65 5.66 ± 3.95 22.80 ± 8.90 2.03 ± 0.72 351 ± 385 144 ± 113 495 ± 419 552 ± 436 6.84 ± 4.83 42.96 ± 23.97 54.12 ± 22.85 66.56 ± 20.76 55.85 ± 20.55 N/A N/A N/A N/A N/A N/A N/A
NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS
LED: Levodopa equivalent; PPI: Prepulse inhibition; UPDRS: Unified Parkinson’s disease rating scale. P ≤ 0.05; NS: not significant.
0–100 ms) with EMG and background noise, before a stimulus was presented. Afterwards these 6 ‘means’ were averaged to one baseline estimate for the given inter-stimulus-interval. The amplitude was measured as the maximum value in the analysis-window which ranged from 100 ms–600 ms. The amplitude was also a mean of the 6 trials per inter-stimulus-interval (30, 60, 120, 300 ms). Mean startle magnitude was calculated by averaging responses on pulse-alone trials for the initial four pulse-alone trials. An example of raw EMG-responses is given in Fig. 1. PPI (%) was calculated using the formula: [(Amplitude pulse-alone–Amplitude prepulse– pulse)/Amplitude pulse-alone]×100. Trials with excessive baseline activity 100 ms before the expected response were excluded to minimize confounding of possible spontaneous and voluntary eyeblink activity upon startle. Patients were excluded from the analysis for low startle response if their startle magnitude was zero in at least 50 per cent of the first four pulse-trials, or if more than three of the six prepulse-pulse trials were discarded. 123 I-FP-CIT
brain SPECT
123 I-FP-CIT
brain SPECT scans were carried out with a Prism XP 3000 (Picker/Phillips) triple-headed
Table 2 Neuropsychological raw scores (mean ± SD) in the entire PD group Neuropsychological raw scores Rapid visual processing Symbol digit modalities test Stroop word Stroop colour-word Trail-making A Trail-making B Stockings of Cambridge, problems solved Intra-extra dimensional set shift, stages completed Rey auditory verbal learning test, immediate recall Rey auditory verbal learning test, delayed recall Danish national reading test Addenbrooke’s cognitive examination
PD group (N = 38) 0.89 ± 0.05 47.79 ± 9.69 84.66 ± 13.15 36.05 ± 8.96 38.50 ± 11.45 97.89 ± 40.42 7.71 ± 1.88 8.31 ± 0.89 41.97 ± 9.26 8.69 ± 2.93 32.61 ± 9.50 91.19 ± 4.66
Raw score data, uncorrected for age and education are reported.
gamma camera equipped with low-energy, ultra-high resolution fan beam collimators. The SPECT-scanning was performed according to guidelines [42]. The SPECT-scanning started exactly 3 hours after patients had received an intravenous injection of 185 MBq 123 IFP-CIT, when steady state within the striatum had been achieved. 123 I-FP-CIT SPECT scans were analysed with an automatic 3D quantification tool (Exini DAT,
M. Zoetmulder et al. / PPI, Cognition, and Dopamine in PD Table 3 Factor structure of the neuropsychological tests-scores in PD (n = 38) Rapid visual processing Symbol digit modalities test Stroop word Stroop colour-word Trail-making A Trail-making B Stockings of Cambridge, problems solved Intra-extra dimensional set shift, stages completed Rey auditory verbal learning test, immediate recall Rey auditory verbal learning test, delayed recall
Factor 1
Factor 2
Factor 3
0.650 0.610 0.596 0.815 −0.689 −0.835 0.073
0.015 0.487 0.461 −0.024 −0.169 −0.271 0.126
0.405 0.247 0.071 0.013 −0.268 −0.056 0.825
0.205
0.183
0.762
0.255
0.783
0.323
0.039
0.947
0.068
Factor loadings greater than 0.500 in bold.
EXINI Diagnostics AB, Lund, Sweden). The quantification module automatically calculates specific uptake in the striatum, putamen and caudatus, and the putamen/caudate ratio for each patient’s brain images. It delineates the cerebral cortex as the background region. Cognitive assessment The same day as patients were assessed with PPI, we administered a neuropsychological battery that was designed to include measures sensitive to functional deficits of the striatal-frontal pathway and dopamine transporter [43, 44]. Attention and processing speed were rated by performance on several tasks: the symbol digit modalities test, in which the subject matches numbers with symbols; the trail-making test, in which the subject draws lines connecting consecutive numbers or numbers that alternate with letters; and the Stroop interference test, in which the subject reads colour names printed in incongruent ink colours and has to suppress the tendency to say the word instead of the colour; the CANTAB rapid visual processing test, which is a serial target detection task that requires sustained attention and working memory for its execution. Memory was rated by performance on the Rey auditory verbal learning test, in which the subject has to learn and recall lists of unrelated words immediately and after a time-delay. Executive function was rated by performance on the CANTAB stockings of Cambridge, which is a problem-solving test, and intra-extra dimensional set shift, an analogue of the Wisconsin card sorting test. Premorbid verbal intelligence was assessed with the Danish version of the national adult reading test (DART) [45]. In addition, patients were
81
assessed with Addenbrooke’s cognitive examination (ACE), which measures a patient’s global cognitive performance [46]. Data analysis and statistics Due to problems of multicollinearity among the neuropsychological variables, the PPI variables, and the four SPECT variables, preliminary analyses were performed on the original variables to minimize multicollinearity and reduce the number of variables considered in further statistical analysis. As a first step, factor analysis with varimax rotation was applied to the 10 neuropsychological measures in the 38 subjects to identify those scores expressing a similar proportion of the total variance. Factor analysis identified three cognitive factors, as detailed in the results section in Table 3. A conventional threshold of 0.5 was applied to factor loadings (expressing the factor variable correlation). The four PPI-85 dB trials (30 ms 85 dB, 60 ms 85 dB, 120 ms 85 dB, and 300 ms 85 dB) were positively correlated (r = 0.40–0.71 in the PD group; N = 38. After Bonferroni testing (0.05/6; P = 0.0083), all correlations remained significant, see Fig. 3. Therefore, these four PPI values were averaged in each subject, and a mean PPI value (mPPI-85 dB) was computed. The four SPECT variables (caudate and putamen uptake in each side) were also highly correlated (r = 0.71–0.87), as found in another study of Nobili et al. (2010), because these variables express the global nigrostriatal function in a given subject, normalized for background uptake. However, as the caudate is more associated with cognitive function, and putamen more with motor function, the uptake values were averaged for caudate nucleus and putamen in each subject, and a mean uptake for caudate (mSPECT-caudate) and putamen (mSPECT-putamen) was computed. To analyse the relationship between PPI and cognitive measures, Pearson correlation coefficient was used. We applied Bonferroni correction of the threshold of statistical significance in the case of multiple comparisons. A subset of patients (N = 23) were assessed with 123 IFP-CIT brain scanning. The relationship between PPI and cognitive variables, and 123 I-FP-CIT uptake, was assessed with Kendall’s tau-b. To investigate the predictive value of single variables for cognitive measures further a regression analysis with stepwise exclusion was performed. Variables in Table 2 were normally distributed and not skewed, although the Stroop colourword variable showed kurtosis (Z = 2.27), exceeding the Z-score of 1.96, at the 5% significance-level. Statistical analysis was performed using SPSS version 17
82
M. Zoetmulder et al. / PPI, Cognition, and Dopamine in PD
(SPSS Inc., Chicago, IL, USA). A 5% significance level was adopted. RESULTS Demographic and clinical characteristics of all 38 PD patients are listed in Table 1, and the means and confidence intervals of the PPI inter-stimulus-trials are further depicted in Fig. 2. Neuropsychological tests and composite scores The neuropsychological raw scores of the PD group are presented in Table 2. Factor analysis yielded three factors, see Table 3. The first factor represented RVP, SDMT, Stroop colour-word, and the trail-making tests A and B. This factor included time-based tests, tapping attention and processing speed (composite processingspeed). The second factor included subscores of the memory test RAVLT (composite memory). The third factor included the SOCps and the IEDsc, which are measures associated with executive function (composite executive). Relationship between mPPI-85 dB and cognitive composite scores
indicates that patients with less PPI were significantly slower, or had worse performance, on tests included in the composite processing-speed than patients with higher PPI, whereas we found no significant differences between mPPI-85 dB and the composite memory and executive scores. Furthermore, the composite mPPI-85 dB was significantly correlated to ACE scores (Pearson r = .362, P < 0.016). Given the significant correlation between mPPI85 dB and the composite processing speed, we performed a correlation analysis with Bonferonni testing to analyse which of the PPI-85 dB interstimulus trials accounted for this finding. This analysis showed a direct correlation between the composite processing speed and PPI-30 ms 85 dB (Pearson r = 0.370, P = 0.014), PPI-60 ms 85 dB (Pearson r = 0.669, P = 0.000, see Fig. 4), and PPI-120 ms 85 dB (Pearson r = 0.354, P = 0.020), while no significant correlation was found for the PPI-300 ms 85 dB. After Bonferroni testing (0.05/4 = 0.0125), only the PPI 60 ms 85 dB remained significantly correlated, see Fig. 4. The different measures of Levodopa equivalent dosage were not significantly associated with PPI-60 ms 85 dB. Relationship between PPI 60 ms 85 dB, SPECT, and composite score processing-speed
123 I-FP-CIT
In the entire group, a significant correlation was found between mPPI-85 dB and the composite processing-speed (Pearson r = 0.47, P = 0.002), which
Fig. 2. Percent prepulse inhibition (PPI) of the acoustic startle response in patients with Parkinson’s disease. Data represent the mean PPI for the four prepulse-pulse inter-stimulus-intervals, and the error bars describe the confidence interval.
To evaluate the relationship between sensorimotor gating, cognitive function and the nigrostriatal dopaminergic system, we looked at the subset of patients who, in addition to PPI and cognitive testing, had undergone 123 I-FP-CIT SPECT (N = 23), for demographic and clinical data see Table 1. We found that the PPI-60 ms 85 dB inter-stimulus-interval was correlated with the composite processing-speed in this subgroup (Kendall’s tau-b = 0.534, P =
