Ò

PAIN xxx (2014) xxx–xxx

www.elsevier.com/locate/pain

Quantitative sensory testing and pain tolerance in patients with mild to moderate Alzheimer disease compared to healthy control subjects Christina Jensen-Dahm a,⇑, Mads U. Werner b, Jørgen B. Dahl c, Troels Staehelin Jensen d, Martin Ballegaard e, Anne-Mette Hejl a, Gunhild Waldemar a a

Danish Dementia Research Centre, Department of Neurology, Rigshospitalet, Copenhagen University Hospital, Denmark Multidisciplinary Pain Center, Neuroscience Center, Rigshospitalet, Copenhagen University Hospital, Denmark Department of Anaesthesia, Rigshospitalet, Copenhagen University Hospital, Denmark d Danish Pain Research Center, Department of Neurology, Aarhus University Hospital, Denmark e Department of Clinical Neurophysiology, Rigshospitalet, Copenhagen University Hospital, Denmark b c

Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.

a r t i c l e

i n f o

Article history: Received 19 July 2013 Received in revised form 23 December 2013 Accepted 30 December 2013 Available online xxxx Keywords: Alzheimer disease Cold pressor test Dementia Elderly Pain Quantitative sensory testing

a b s t r a c t Patients with Alzheimer disease (AD) report pain less frequently than their cognitively intact peers. It has been hypothesized that pain processing is altered in AD. The aim of this study was to investigate agreement and reliability of 3 pain sensitivity tests and to examine pain threshold and tolerance in patients with AD. We examined 29 patients with mild to moderate AD and 29 age- and gender-matched healthy control subjects with quantitative sensory testing, ie, assessments of detection threshold (warmth detection threshold [WDT]) and pain threshold (heat pain threshold [HPT], pressure algometry, cold pressor test), and assessments of tolerance (pressure algometry, cold pressor test). All procedures were done twice on day 1, 1 hour apart, and repeated on day 2. We found no difference between groups for WDT (patient vs control subjects: mean [95% confidence interval]: 35.5°C [33.4°C to 37.6°C] vs 35.4°C [34.3°C to 36.5°C], P = .8) or HPT (41.2°C [40.0°C to 42.4°C] vs 42.3°C [41.1°C to 43.5°C], P = .24). We observed comparable thresholds for pressure algometry (median [25% to 75% interquartile range]: 120 kPa [100 to 142 kPa] vs 131 kPa [113 to 192 kPa], P = .10), but significantly lower tolerance in AD patients (213 kPa [188 to 306 kPa] vs 289 kPa [262 to 360 kPa], P = .008). No differences were found for the cold pressor test. The study demonstrated good replicability of the sensory testing data with comparable data variability, for both groups, which supports the use of these methods in studies of patients with mild to moderate AD. Contrary to previous studies, we observed a reduced pain tolerance in patients with mild to moderate AD, which suggests that the reduced report of pain cannot be explained by reduced processing of painful stimuli. Ó 2014 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.

1. Introduction Epidemiological studies have described a reduced report of pain [1,18,34] and lower analgesic consumption in patients with dementia [17,23,27]. One study reported that patients with early Alzheimer disease (AD) report less intense pain compared to cognitively intact elderly people [35], suggesting that pain perception or pain reporting is disturbed already in the early stages of the disease. These findings have raised the question of whether AD leads to a change in pain experience due to an altered pain processing related to neurodegenerative changes [36]. This issue has only ⇑ Corresponding author. Address: Danish Dementia Research Centre, Department of Neurology, Rigshospitalet, Copenhagen University Hospital, Blegdamsvej 9 # 7621, 2100 Copenhagen, Denmark. Tel.: +45 35457661; fax: +45 35452446. E-mail address: [email protected] (C. Jensen-Dahm).

been studied in a handful of experimental studies, and so far no consistent findings have been demonstrated. Benedetti et al. did not identify any differences in pain threshold between AD patients and healthy control subjects using electric and ischemic stimuli, but the tolerance to pain was increased [4]. Cole et al. used pressure pain stimuli and found an increased threshold for just noticeable pain [9,10]. These contradictory findings might be attributable to methodological differences in regard to pain induction, pain assessment, and severity of AD. Another explanation is that it is unclear whether the methods are appropriate in patients with AD. Patients with AD have impairment of short-term memory and may have difficulties understanding simple instructions. Consequently, some of the discrepancies found in the literature in regard to pain perception in AD patients may be due to unintentional use of assessment methods that are not reliable in patients

0304-3959/$36.00 Ó 2014 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.pain.2013.12.031

Please cite this article in press as: Jensen-Dahm C et al. Quantitative sensory testing and pain tolerance in patients with mild to moderate Alzheimer Ò disease compared to healthy control subjects. PAIN (2014), http://dx.doi.org/10.1016/j.pain.2013.12.031

2

Ò

C. Jensen-Dahm et al. / PAIN xxx (2014) xxx–xxx

with AD. Only 1 study has evaluated some aspects of reliability (ie, coefficient of variation) [15], and at present it is unknown whether the methods used in prior studies were appropriate as no studies have thoroughly investigated the reliability or agreement of the methods used. In order to bring this research forward, it is essential to thoroughly examine test-retest reliability and agreement to clarify whether the methods used are appropriate. Second, in order to be able to compare with other studies, it is important to use methods for which standardized protocols have been published [32,43]. The primary aim of our study was to estimate test-retest reliability and agreement of different pain sensitivity models using quantitative sensory testing, ie, assessments of thermal and mechanical thresholds, and assessments of tolerance to cold and pressure stimuli, in patients with AD. If the reliability and agreement of 1 or several of the models were judged to be acceptable, the secondary aim was to investigate the effects of mild to moderate AD on pain processing. 2. Patients and methods 2.1. Subjects The protocol was approved by the Regional Committees on Health Research Ethics of the Capital Region of Denmark (protocol: H-4-2010-099) and the Data Protection Agency (journal number: 2007-58-0015/30-0862). Twenty-nine patients were recruited among outpatients from the Memory Clinic at Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark. All patients fulfilled the 10th revision of the International Statistical Classification of Diseases and Related Health Problems and Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition criteria for dementia and had a diagnosis of probable AD according to the McKhann criteria [25], which was given as a consensus diagnosis by a group of dementia specialists. The patients included were examined with a cranial computed tomography or magnetic resonance imaging (MRI), blood sample screening, and cognitive tests as part of their initial workup. Patients were able to give informed consent, and all had a caregiver who was willing to participate. We choose patients with a Mini Mental State Examination (MMSE, see description later) score between 16 and 26 points (both limits included) and a Clinical Dementia Rating (see description below) of 0.5 to 2 (both limits included), which is equal to mild to moderate dementia. The MMSE limits had been chosen as the authors believed that patients within this range would be able to cooperate with the tests and give informed consent (for further information see description of MMSE later). We included 29 age- and gender-matched healthy control subjects. The control subjects were recruited from a group of elderly people who had previously participated in studies at the memory clinic, where they had been cognitively tested and had been found not to have dementia or mild cognitive impairment. All participants gave informed consent to the study protocol. Both patients and control subjects were excluded if they had significant psychiatric comorbidity, prior or present alcohol abuse, used daily analgesics (ie, paracetamol, nonsteroidal anti-inflammatory drugs, opioids, gabapentin/pregabalin, or tricyclic antidepressants) or had a disorder that would interfere with pain perception and pain report, such as diabetes, peripheral or central neuropathy, a chronic pain disorder, or current pain condition. They also were excluded if they had significant medical comorbidity or previously had a transient ischemic attack or stroke. Patients with a mixed diagnosis of vascular dementia and AD were excluded. At baseline, both patients and control subjects had a neurological examination and were excluded if they had any symptoms or signs of any neurological or inflammatory disease that could interfere with pain perception.

2.2. Measures 2.2.1. Evaluation of global cognitive function To evaluate the participants’ cognitive status, an MMSE was used. The MMSE is a screening instrument that is a brief standardized method to assess mental status. The score ranges from 0 to 30, with higher scores indicating better cognitive performance [14]. The MMSE is highly dependent on educational level [11], and in Denmark, where there is generally a high level of education, the optimal cutoff for dementia in population-based studies has been determined to 26 [19]. To further evaluate the patients, we also administered the Addenbrookes Cognitive Examination (ACE). The ACE is a brief test battery that includes the MMSE, but expands on cognitive domains such as memory, language, and visuospatial function, and includes a test of verbal fluency [24]. It takes 10 minutes to administer and requires no specialized equipment. ACE has been validated in Danish with an optimal cutoff of 85 of 86 for dementia [40]. The Clinical Dementia Rating is a numeric scale used to quantify the severity of dementia (ie, its stage). It uses a structured interview protocol to assess the patient’s cognitive and functional performance in 6 areas: memory, orientation, judgment and problem solving, community affairs, home and hobbies, and personal care. Scores in each of these are combined to obtain a composite score ranging from 0 through 3, with 3 indicating a severe stage of dementia [5]. 2.2.2. Activities of Daily Living (ADL) function The ADL function was assessed using the Functional Activities Questionnaire and the Instrumental Activities of Daily Living Scale [30], which is a questionnaire evaluating 10 different activities of daily living with a maximum score of 30 (dependent on help). 2.2.3. Depressive symptoms The participants were screened for signs of a major depression with the Geriatric Depression Scale, 15 items. The Geriatric Depression Scale is a self-reported questionnaire specifically developed as a screening instrument for the presence of depressive symptoms in older populations [45]. The maximum possible score is 15. As most of the patients were unable to complete the questionnaire on their own, it was completed as an interview. 2.2.4. Reaction time Reaction time was measured using: http://getyourwebsitehere.com/jswb/rttest01.html, showing a traffic light. The subject was instructed to press the button when the light changed from red to green. The light changed with a random interval (up to 7 seconds). The participants were allowed to try the test before recording the reaction time, in order for them to feel comfortable using the test. Ten measurements were made, and a mean value was presented. 2.3. Quantitative sensory testing We used important elements from the standardized protocol for quantitative sensory testing published by the German Research Network on Neuropathic Pain [32] and the cold pressor test, for which guidelines for use in children have been published [43]. We chose to use only part of the protocol as patients with AD are not able to cooperate with long testing sessions, and we wished to be able to repeat the test. We chose methods that were easy to understand and to cooperate with, as this is crucial in doing any kind of examination on patients with AD. Additionally, we were interested in examining different pain modalities induced by either mechanical or thermal stimuli, and we considered the test stimuli suited for demonstrating activation of a range of nociceptors.

Please cite this article in press as: Jensen-Dahm C et al. Quantitative sensory testing and pain tolerance in patients with mild to moderate Alzheimer Ò disease compared to healthy control subjects. PAIN (2014), http://dx.doi.org/10.1016/j.pain.2013.12.031

Ò

C. Jensen-Dahm et al. / PAIN xxx (2014) xxx–xxx

The participants were tested on 2 different days. During day 1, the participants were tested with the test battery twice (session 1 and session 2). The sequence of pain stimuli was: thermal stimulation, pressure algometry, followed by the cold pressor test. After the first test session, there was a 30-minute break to ensure that there was 1 hour between sessions, ie, the second session had the same temporal and content sequence as the first. The participants were asked to come back for a third test session on a separate day (day 2), when the test was repeated once. For all 3 tests, standardized, written instructions had a priori been defined in order to ensure similar instruction of both patients and control subjects. There was only 1 examiner (C.J.D.). 2.3.1. Thermal stimulation The patients were comfortably lying in a supine position during the testing. Before starting the test, the skin temperature was measured to ensure that the temperature was P32°C. The contact heat stimuli were given by a thermal stimulator (Medoc TSA-2001; Medoc Ltd., Ramat Yishai, Israel); contact heat stimuli were delivered by a Peltier element-based thermode with an active surface area of 9 cm2 (3.0  3.0 cm). The probe was applied to the skin with a firm but gentle pressure necessary to ensure that the active thermode area was securely in contact with the skin during the thermal testing. The test was performed on the volar side of the lower left arm, with the distal margin of the thermode 10 cm above the wrist. The protocol for determining warmth detection threshold (WDT) and heat pain threshold (HPT) was modified to accommodate for cognitive difficulties and potentially delayed reaction times. We used the method-of-limits approach, and in contrast with the standardized protocol we lowered the heating rate to 0.5°C/s in order to account for delayed reaction time. Baseline temperature was 32°C, and cutoff temperature was 50°C. The stimuli were delivered with a random interval of 4 to 6 seconds. Before each stimulus, the participants were informed about the sensory modality to assess in order to ensure that the participants did not forget this. The participants were asked to press a switch at the point when they first felt warmth sensation (WDT), which terminated the thermal stimulation, returning the temperature to baseline. Heat pain threshold was defined as the point at which the thermal sensation turned to pain. In assessments of WDT and HPT, a repeat series of 5 stimuli was applied, in which the first 2 stimuli were considered practice trials. After HPT assessment, 2 suprathreshold stimuli each of 5 seconds’ duration were applied. The stimuli were 0.5°C and 2.5°C above HPT, and a ramp of 2°C/s was used. The participant was asked to rate the pain on a horizontally held colored analogue scale (CAS), which have been shown to have good reliability in patients with mild to moderate dementia [37] and have been recommended in a recent guideline [16]. At the baseline visit, the participants’ ability to understand the scale had been tested, and all were able to understand and explain the scale correctly. 2.3.2. Pressure pain stimuli Pressure pain thresholds (PPT) and pressure pain tolerance (PPTo) were measured with a manual pressure algometer (Somedic AB, Hörby, Sweden) using a felt-tipped probe with a circular surface of 1 cm2. Details of the algometer have been described elsewhere [8]. The pressure was applied at the middle phalanx of the left index finger and in order to accomodate for cognitive difficulties and potentially delayed reaction times we lowered the ramp rate from 50 kPa/s to 10 kPa/s. The cutoff pressure was 400 kPa. PPT was defined as the point at which the pressure sensation turned to pain and PPTo as the point at which the subject felt the pain as intolerable. The subjects were instructed to press a switch when these points were reached, terminating the stimulus.

3

Participants were allowed to familiarize with the test before starting. The PPT was measured as an average of 3 measurements, whereas PPTo was only measured once. 2.3.3. Cold pressor test (CPT) A recirculating water cooler (model 11371P, VWR International, Radnor, Pennsylvania) with a bath volume of 13 L was used. The water temperature was maintained at 10°C (±0.25°C) because small temperature deviations have been shown to significantly affect pain threshold and tolerance assessments [26]. The participant was asked to submerge their right hand in the water covering the wrist, and the hand was not allowed to touch the sides or the bottom of the bath. The subject was instructed to state when the cold stimulus became painful, indicating the CPT pain threshold, and to withdraw the hand when the pain became intolerable, indicating CPT pain tolerance. Threshold and tolerance were recorded during the same trial. The cutoff value of the tolerance test was 3 minutes. 2.3.4. Evaluation of methods The authors evaluated the feasibility and repeatability of the 3 tests. The feasibility was evaluated as the participants’ ability to understand and cooperate with the tests. The repeatability (test-retest) of the 3 sessions was evaluated within days and between days. Repeatability concerns the variation in repeated measurements made under the same conditions in the same subject and is divided into 2 components [12]: reliability [3] and agreement [6]. Reliability is the consistency of a measure or the strength of the relationship and assesses the degree to which test scores are consistent from one test administration to the next. In our case the values are continuous, and we calculated an intraclass correlation coefficient ICC(2,1) that is closely related to the weighted kappa measure of agreement [2]. If there is good reliability, the ICC will approach 1.0. For interpretation purposes, ICCs were categorized as slight/poor (0.2 to 0.4), moderate (>0.4 to 0.6), substantial (>0.6 to 0.8), and almost perfect (>0.8) [2]. Calculation of ICCs is based on an anticipated normal distribution of data, which was not the case for PPTo and cold pressor tolerance as there was a ceiling effect. We therefore also calculated a weighted kappa (quadratic weights) for ordinal data (data split into quartiles). Agreement quantifies how close 2 measurements, eg, on 2 sessions or days, made on the same subject are (sometimes also termed precision). A certain lack of agreement between 2 measurements is inevitable; what matters are the magnitude by which the 2 measurements disagree. One way of quantifying the difference between the measurements is by the limits of agreement: mean difference ± 1.96  SD. These values define the range within which most differences between measurements are expected to lie. A systematic difference between the 2 measurements is called bias. This method assumes that the SD of the method’s differences is uniform throughout the range of measurements, which was not the case in our study. A number of the plots were funnel-shaped, ie, the test-retest difference increased with magnitude of the mean values. Instead of logarithmic transformation of the data, we calculated the ratios (ratio [95% confidence interval]) between the measurements. If there was perfect agreement, the ratios would be close to 1.0. [3–7]. Data are presented as ratio [95% confidence interval]. 2.4. Statistics Before initiating the study, a sample size calculation was performed. The level of significance (a) was chosen to be 0.05 and a power of 0.80 (b = 0.2). The calculation was based on normative HPT values in the elderly [13], and the smallest difference between groups we wanted to detect in regard to HPT was 5% or 2.2°C.

Please cite this article in press as: Jensen-Dahm C et al. Quantitative sensory testing and pain tolerance in patients with mild to moderate Alzheimer Ò disease compared to healthy control subjects. PAIN (2014), http://dx.doi.org/10.1016/j.pain.2013.12.031

Ò

4

C. Jensen-Dahm et al. / PAIN xxx (2014) xxx–xxx

Based on these numbers, the estimated sample size was 28 in each group. All data were tested for normal distribution using the ShapiroWilk test of normality, analyses of skewness and kurtosis, and histograms. In case of nonnormal distribution, simple log-transformations were tried, and if unsuccessful, the numbers are given as median (25% to 75% interquartile range) and otherwise as mean [95% confidence interval]. Differences between groups were evaluated using a 2-way analysis of variance (ANOVA) for repeated measures, including group and session as covariates. For suprathreshold stimuli, the difference between CAS ratings for 3 sessions were evaluated using a random effects model as the CAS ratings are correlated. A significance level of P < .05 was used. All data were analyzed using SAS 9.1.3 (SAS Institute Inc., Cary, NC).

subjects. For HPT the mean values for the 3 sessions were 41.2°C [40.0°C to 42.4°C] for patients and 42.3°C [41.1°C to 43.5°C] for control subjects. Group differences were evaluated using a 2-way ANOVA for repeated measures, which showed no differences for WDT (P = .8) or HPT (P = .24), but for both WDT and HPT there was a significant increase at session 2 (P = .015). Additionally no difference in thermal thresholds, corrected for reaction time, was observed. Pain ratings during the suprathreshold stimuli are shown in Table 4. For all 3 sessions we found a significant difference between CAS ratings for HPT +0.5°C and HPT +2.5°C (P < .0001). There was no effect of group for any of the 3 session (session 1: P = .32; session 2: P = .12; session 3: P = .48).

3. Results

Results from the 3 sessions are presented in Table 2. Table 3 presents the limits of agreement and ICC. The estimated biases ranged between 1% to 4% for patients and between 4% to 9% for control subjects for same-day measurement. For separate-day measurements, the estimated biases ranged between 18% to 24% for patients and 4% to 5% for control subjects. For same-day measurements, the ICC indicated a substantial reliability for the patients (threshold: 0.72; tolerance: 0.79), but a substantial to almost perfect reliability for the healthy control subjects (threshold: 0.84; tolerance: 0.79). The ICC calculated for separate-day measurements indicated a moderate reliability for both patients (threshold: 0.34; tolerance: 0.61) and control subjects (threshold: 0.50; tolerance: 0.46). Weighted kappa values were closely related to the ICCs for PPTo. The median value for the 3 sessions for pressure pain threshold was 120 kPa (100 to 142 kPa) for patients and 131 kPa (113 to 192 kPa) for control subjects. For PPTo the median values were 213 kPa (188 to 306 kPa) for patients and 289 kPa (262 to 360 kPa) for control subjects. Group differences were evaluated using a 2-way ANOVA for repeated measures, which showed no significant difference for PPT (P = .10), but a significant interaction between group and session (P = .021) was seen. However, a significantly lower PPTo was observed for the patients compared to the control subjects (P = .008) with no effect of session. These findings were robust even when correcting for reaction times.

Baseline demographics are shown in Table 1. The patients used more drugs, but this was due to treatment for AD (ie, antidementia medication and selective serotonin reuptake inhibitor). All subjects were able to understand and cooperate with the tests. All subjects completed day 1, but 6 patients and 3 control subjects were unavailable for day 2. Reasons for dropout were illness that occurred after day 1 (3 patients, 2 control subjects). For 2 patients, the caregiver was unable to take the patient to the last visit; for the remaining (1 patient, 1 control), we were unable to get in contact with them. 3.1. WDTs and HPTs Results from the 3 sessions are displayed in Table 2. Table 3 presents the limits of agreement and ICC. The ratios were close to 1.0, with the estimated biases ranging between 1% for the patients and 2% to 4% for the control subjects. The ICC for WDT indicated a moderate to substantial agreement for the patients (same day: 0.72; separate days: 0.45), and a moderate agreement for control subjects (same day: 0.56; separate days: 0.62). For HPT there was a moderate to substantial reliability for patients (same day: 0.70; separate days: 0.50), but an almost perfect reliability (same day: 0.86; separate days: 0.84) for the control subjects. The mean values for the 3 sessions for WDT were 35.5°C [33.4°C to 37.6°C] for the patients and 35.4°C [34.3°C to 36.5°C] for control Table 1 Demographics.

Age Sex,% female Assessment of cognitive function Mini Mental State Examination Addenbrookes Cognitive Examination Depressive symptoms (Geriatric Depression Scale) Activities of daily living (Functional Activities Questionnaire and Instrumental Activities of Daily Living Scale) Number of drugs [95% CI] SSRI,% (N) Antidementia medication (anticholinergic drugs) Reaction time (ms) Interval between day 1 and day 2 (in days) Numbers are given as mean ± SD. NS = nonsignificant. P < .001.

*

Patients (n = 29)

Healthy control subjects (n = 29)

70.9 ± 6.9 51.7%

70.5 ± 5.5 51.7%

22.4 ± 1.18 66.2 ± 10.9 1.5 ± 1.7

29.7 ± 0.5 – 0.6 ± 0.9

14.03 ± 6.11

–

2.9 [1.7– 3.7] 27.6% (8) 96.6% (28)

1.5 [1.2–2.1]* 3.4% (1) 0

655 ± 315 13.9 ± 2.3

351 ± 82* 10.3 ± 1.8, NS

3.2. PPT and PPTo

3.3. CPT Results from the 3 sessions are displayed in Table 2. Table 3 displays the limits of agreement and ICC. The estimated biases ranged from 4% to 29% for both groups, but were similar for the 2 groups. For patients, the ICC for CPT pain threshold indicated a substantial reliability for same-day measurements (0.73) but fair reliability for separate-day measurements (0.32), which was fair to moderate for the control subjects (same day: 0.31; separate: 0.52). For the CPT pain tolerance, we found a slight to poor reliability for same-day measurements (ICC: patients: 0.12; control subjects: 0.18) and substantial to strong reliability for both groups (ICC: patients: 0.68; control subjects: 0.70) for separate-day measurements. Weighted kappa values were closely related to ICCs for CPT pain tolerance. The median values for the 3 sessions for CPT pain threshold were 14.9 seconds (11.7 to 20.5 seconds) for patients and 11.8 seconds (9.4 to 20.0 seconds) for control subjects. For CPT pain tolerance, the median values for the 3 sessions were: 31.3 seconds (21.2 to 50.7 seconds) for patients and 28.1 seconds (20.2 to 45.6 seconds) for control subjects. Group differences were evaluated using a 2-way ANOVA for repeated measures, which showed no significant differences between groups in regard to CPT pain threshold (P = .89) or CPT pain tolerance (P = 1.0). These findings also were robust when correcting for reaction times.

Please cite this article in press as: Jensen-Dahm C et al. Quantitative sensory testing and pain tolerance in patients with mild to moderate Alzheimer Ò disease compared to healthy control subjects. PAIN (2014), http://dx.doi.org/10.1016/j.pain.2013.12.031

Ò

5

C. Jensen-Dahm et al. / PAIN xxx (2014) xxx–xxx Table 2 WDT, HPT, PPT, PPTo, CPT pain threshold, and CPT pain tolerance in patients and healthy control subjects. Day 1

Day 2

Session 1

WDT (°C) HPT (°C) PPT (kPa) PPTo (kPa) CPT threshold (seconds) CPT tolerance (seconds)

Session 2

Session 3

Patients

Control subjects

Patients

Control subjects

Patients

Control subjects

35.4 [34.2–36.6] 40.9 [39.7–42.1] 112 (91–154) 241(177–282) 15.4 (11.9–23.4) 31.7 (22.5–61.3)

34.9 [33.8–36.0] 42.0 [40.7–43.1] 132 (97–182) 288 (237–378) 11.8 (9.8–21.5) 26.1 (19.7–55.5)

35.6 [34.3–36.9] 42.0 [40.8–43.2] 132 (99–166) 212 (168–318) 12.6 (10.1–19.3) 28.9 (20.3–60.8)

36.3 [35.1–37.1] 42.7 [4152–43.9] 126 (88–179) 287 (223–359) 12.6 (8.8–16.8) 27.9 (19.6–42.4)

35.6 [34.5–36.7] 40.5 [39.3–41.6] 129 (86–136) 238 (177–281) 12.5 (10.7–16.2) 29.6 (19.9–39.0)

35.0 [33.9–36.1] 42.1[40.9–43.3] 168 (124–224) 326 (253–393) 12.7 (8.6–15.3) 28.0 (18.6–37.5)

WDT and HPT are normally distributed and values are mean [95% confidence interval]). For PPT, PPTo, CPT pain threshold, and CPT pain tolerance, numbers are given as median (25%–75%) because they were nonnormally distributed. WDT = warmth detection threshold; HPT = heat pain threshold; PPT = pressure pain thresholds; PPTo = pressure pain tolerance; CPT = cold pressor test.

Table 3 Limits of agreement (ratio [95% CI]) and ICC for WDT, HPT, PPT, PPTo, CPT pain threshold, and CPT pain tolerance in patients and healthy control subjects. Session 1 vs session 2

Session 1 vs session 3, day 2

Patients

Control subjects

Patients

Control subjects

WDT Limits of agreement ICC

1.00 [0.81–1.18] 0.72 [0.48–0.86]

0.96 [0.89–1.04] 0.56 [0.04–0.81]

1.01 [0.86–1.16] 0.45 [0.046–0.72]

0.98 [0.89–1.07] 0.62 [0.30–0.81]

HPT Limits of agreement ICC

0.98 [0.83–1.12] 0.70 [0.45–0.85]

0.98 [0.89–1.07] 0.86 [0.70–0.93]

1.01 [0.83–1.19] 0.50 [0.11–0.75]

0.99 [0.89–1.09] 0.84 [0.67–0.92]

PPT Limits of agreement ICC

0.99 [0.23–1.75] 0.72 [0.49–0.86]

1.09 [0.40–1.78] 0.84 [0.69–0.92]

1.24 [ 0.28–2.76] 0.34 [ 0.048–0.65]

1.05 [ 0.59–2.69] 0.50 [0.15–0.74]

PPTo Limits of agreement ICC Weigthed kappa

1.04 [0.26–1.82] 0.79 [0.59–0.89] 0.72 [0.52–0.92]

1.04 [0.58 –1.50] 0.79 [0.59 – 0.89] 0.69 [0.53–0.86]

1.18 [ 0.88–3.24] 0.61 [0.27–0.81] 0.67 [0.43–0.90]

1.04 [0.28–1.80] 0.46 [0.08–0.72] 0.28 [ 0.10–0.65]

CPT threshold Limits of agreement ICC

1.15 [0.21–2.10] 0.73 [0.49–0.87]

1.29 [ 0.26–2.84] 0.31 [ 0.050–0.61]

1.06 [0.67–1.46] 0.32 [ 0.15–0.66]

1.23 [0.15–2.45] 0.52 [0.18–0.76]

CPT tolerance Limits of agreement ICC Weigthed kappa

1.04 [0.26–1.82] 0.12 [ 0.014–0.42] 0.18 [ 0.02–0.39]

1.15 [0.23–2.08] 0.18 [ 0.10–0.47] 0.19 [0.01–0.37]

1.23 [ 0.38–2.83] 0.68 [ 0.37–0.86] 0.67 [0.41–0.92]

1.24 [0.06–2.41] 0.70 [0.44–0.85] 0.85 [0.74–0.95]

CI = confidence interval; ICC = intraclass correlation coefficient; WDT = warmth detection threshold; HPT = heat pain threshold; PPT = pressure pain thresholds; PPTo = pressure pain tolerance; CPT = cold pressor test.

Table 4 Pain ratings (CAS 0 to 100) during suprathreshold heat stimuli, 0.5°C and 2.5°C above the individual HPT. Patients (CAS)

Healthy control subjects (CAS)

Session 1 HPT + 0.5°C HPT + 2.5°C

29 (10–33) 30 (15–42)

24 (15–46) 38 (21–56)

Session 2 HPT + 0.5°C HPT + 2.5°C

24 (9–36) 27 (17–54)

21 (14–38) 35 (22–54)

Session 3, day 2 HPT + 0.5°C HPT + 2.5°C

29 (10–41) 33 (13–50)

27 (8–39) 35 (16–60)

Numbers are given as median (25%–75%). CAS = colored analogue scale; HPT = heat pain threshold.

4. Discussion The authors believe this is the first study to demonstrate that it is possible to test pain sensitivity in a reliable way in patients with mild to moderate AD, substantiating that the patients are fully able to understand and cooperate with the thermal and mechanical pain tests. Replicated data assessments showed reproducibility,

ie, agreement and reliability, to be comparable to that of control subjects, which perhaps would not be expected in this population. Pressure algometry and the cold pressor test were associated with some bias, but there were no systematic differences between the 2 groups. Furthermore, to the best of our knowledge this study is the first to use contact heat stimuli in patients with dementia. It is noteworthy that although both mechanical and thermal pain thresholds did not differ between the groups, mechanical pain tolerances, assessed by pressure algometry, were significantly lower in the AD group. The findings were robust also when correcting for differences in reaction time, which may be an important confounder. There was no difference between the 2 groups in pain thresholds on any of the 3 tests. Comparable pain thresholds in AD patients and healthy control subjects were shown in 2 previous studies using pressure and ischemic painful stimuli [4] and heat stimuli [15]. The finding of pain thresholds comparable to matched control subjects is corroborated by the fact that the sensory cortex is largely unaffected until the severe stages of AD [7]. In contrast, other groups have found an increased threshold for just noticeable pain [9] and a markedly diminished nociceptive flexion reflex in patients with dementia [20]. One study looked at the effects of age and mild cognitive impairment and found no effect on pain

Please cite this article in press as: Jensen-Dahm C et al. Quantitative sensory testing and pain tolerance in patients with mild to moderate Alzheimer Ò disease compared to healthy control subjects. PAIN (2014), http://dx.doi.org/10.1016/j.pain.2013.12.031

6

Ò

C. Jensen-Dahm et al. / PAIN xxx (2014) xxx–xxx

ratings or facial responses to electrical stimulation [21]. One explanation for why we did not find a difference in pain thresholds could be that we had insufficient power to detect a difference (ie, a type II error). When we planned our study, we chose 2.2°C as the minimal relevant difference we wished to detect on HPT, and the differences we found were actually smaller. Methodological differences between the studies may also explain the differences, and to our knowledge we are the first to provide extensive examination of test-retest reliability and agreement of psychophysical methods in patients with dementia. However, thermal thresholds are influenced by a number of variables (ie, thermode area, ramp rate, anatomical site, testing glabrous or nonglabrous skin, and skin temperature), and accordingly published reference values vary between 39.7°C ± 1.4°C (left scapula) and 41.9°C ± 2.4°C (left leg) for the elderly depending on the anatomical site [29]. Likewise, pressure pain thresholds are influenced by similar variables, and Neziri et al. found normative values that in most cases were higher than ours, but at different locations [29]. The cold pressor assessments critically depend on the circulating water temperature [26]. Because temperatures below 5°C in the cold pressor test are associated with activation of thermoregulatory baroreflexes, ie, stressinduced hyperalgesia [41,43], the more modest temperature of 10°C was used in this study. To our knowledge no studies using 10°C water have been conducted in elderly subjects. Importantly, we found that our results were comparable to those of previous studies. We found a lower pain tolerance in the patients using pressure algometry. In contrast, Benedetti et al. found an increased tolerance in AD patients [4], but the patients in the study by Benedetti et al. had a more severe degree of dementia (MMSE between 10 and 19). Whether our findings can be extrapolated to patients with more severe degrees of AD is uncertain. In advanced cases of dementia, pain tolerance may not be clearly expressed, either because of problems with comprehension or because of changes in processing of noxious information. Augmented pain responses in the AD population have been shown in a functional MRI study, in which Cole et al. found greater pain-evoked responses in a group of patients with mild to moderate AD [9]. Likewise an increase in facial responses to pain was shown in a mixed population of patient with AD and vascular dementia [22]. An interesting study in healthy adults showed that short-term memory for pain deteriorates after a few seconds [31]. In the study by Benedetti et al., intervals of 10 to 30 seconds were used, and the cognitive challenge therefore may have been too high. Similarly, this could also be the case in the present study and may explain why we did not find a difference in regard to the cold pressor test. Second, the cold pressor test was associated with large variability, and a post hoc power calculation estimated that 4500 patients would be needed in each group to show a probable difference (a = 0.05 and b = 0.2), which lead us to conclude that a type II error and rejection of the null hypothesis are very unlikely. It is of interest to speculate on the mechanism behind the differential response on pain threshold and pain tolerance. Although the former preferentially relates to the sensory-discriminative processing of pain, the latter rather reflects the cognitive-evaluative and affective-emotional aspects of pain. Cognitive impairment, psychological stress, and anxiety are more likely to affect pain tolerance than pain threshold. Previous imaging studies have indicated that areas such as the amygdala, anterior cingulate cortex, and insular cortex are sites that are activated when the emotional aspect of pain is examined [39,44]. The pathological changes in AD patients develop over decades and first affect the transenthorinal cortex and the hippocampus region. The changes spread throughout the limbic system, and coinciding with the diagnosis of AD, the patients have widespread changes of the limbic system with interruption of connections between its components; its influence

on the prefrontal cortex is markedly reduced (equivalent to Braak stage IV to V) [7,28], but the sensory cortex is largely unaffected until the severe stages of AD. It also has been shown that patients have pathological changes in the intralaminar nuclei of the thalamus, which are progressively affected by the disease, with severe changes in pathological stages equivalent to clinical AD [33]. The pathological changes in AD have a wide impact on the medial pain system [25,38], which is involved in processing the affective components of pain, but that it should lead to a loss of function (ie, increased tolerance of pain) is not a given. In a functional MRI study, Cole et al. tested the hypothesis that reduced pain report in patients with AD occurred due to selective damage of the medial thalamic and limbic structures involved in affective-motivational pain processing. They found no evidence of diminished pain-related activity in the medial pain regions in AD, but found greater amplitude and duration, indicating that the affective/motivational component of pain was not selectively diminished [9]. A second explanation for why patients with AD may have a lower tolerance of pain, but a similar pain threshold, could be that due to cognitive problems the patients have difficulty handling stressful events such as a painful stimulus, which could lead to an increased level of anxiety. Studies in healthy adults have shown that manipulation of mood and/or emotions strongly affected pain unpleasantness and not pain intensity [42]. This also may explain why we observed a lower tolerance of pain, but no change in pain threshold. One of the strengths of our study is that it included a well-characterized group of patients with AD using a standardized protocol. Replicated assessments were made showing good reproducibility, ie, agreement and reliability, which perhaps would not be expected in this population, strengthening that our findings are not incidental. A limitation is that we included AD patients without concomitant diseases, and our results may not be applicable to all AD patients. Second, both patients and control subjects were elderly, and some participants experienced worsening of existing or occurrence of new disease, which in part explains why we had a dropout of 6 patients and 3 control subjects. Third, we only examined patients with mild to moderate dementia, and our results cannot be translated to patients with severe dementia, in whom other methods for pain assessment may be needed. Fourth, we were only able to show a lower tolerance using pressure algometry, and we cannot be certain that this finding was not incidental (ie, a type I error). 4.1. Conclusions Patients with mild to moderate AD were able to reliably cooperate with standardized thermal and mechanical pain sensitivity tests, compared to an age- and gender-matched control group. Three replicated quantitative sensory assessments demonstrated good data reproducibility, ie, reliability and agreement. The pain thresholds did not differ between AD patients and control subjects, but a significantly lowered mechanical pain tolerance was observed in AD patients. Our results suggest that the reduced verbal report of pain in AD cannot be explained by impaired processing of thermal and mechanical pain stimuli. Alternative explanations may include impaired communication, memory problems, or anosognosia associated with AD. Conflicts of interest G.W. is a board member of the Lundbeck Foundation and has served as a speaker and consultant for Lundbeck Inc. and Pfizer. T.S.J. has received financial support and provided consultant services for the following: Pfizer; Grünenthal; Astellas; Daiichi

Please cite this article in press as: Jensen-Dahm C et al. Quantitative sensory testing and pain tolerance in patients with mild to moderate Alzheimer Ò disease compared to healthy control subjects. PAIN (2014), http://dx.doi.org/10.1016/j.pain.2013.12.031

Ò

C. Jensen-Dahm et al. / PAIN xxx (2014) xxx–xxx

Sankyo. M.U.W. has received unrestricted research grants from Grünenthal og Astellas. Acknowledgments The Danish Dementia Research Centre is supported by grants from the Danish Health Foundation and the Danish Ministry of Health. The authors thank the doctors at the Memory Clinic at Rigshospitalet for referring patients to the study. References [1] Achterberg WP, Gambassi G, Finne-Soveri H, Liperoti R, Noro A, Frijters DH, Cherubini A, Dell’aquila G, Ribbe MW. Pain in European long-term care facilities: cross-national study in Finland Italy and The Netherlands. PAINÒ 2010;148:70–4. [2] Armitage P, Berry G, Matthews JNS. Statistical methods in medical research, statistical methods in epidemiology. 4th ed. Oxford, UK: Blackwell Science; 2002. p. 648–716. [3] Bartlett JW, Frost C. Reliability, repeatability and reproducibility: analysis of measurement errors in continuous variables. Ultrasound Obstet Gynecol 2008;31:466–75. [4] Benedetti F, Vighetti S, Ricco C, Lagna E, Bergamasco B, Pinessi L, Rainero I. Pain threshold and tolerance in Alzheimer’s disease. PAINÒ 1999;80:377–82. [5] Berg L. Clinical dementia rating (CDR). Psychopharmacol Bull 1988;24:637–9. [6] Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res 1999;8:135–60. [7] Braak E, Griffing K, Arai K, Bohl J, Bratzke H, Braak H. Neuropathology of Alzheimer’s disease: what is new since A. Alzheimer? Eur Arch Psychiatry Clin Neurosci 1999;249(Suppl. 3):14–22. [8] Brandsborg B, Dueholm M, Kehlet H, Jensen TS, Nikolajsen L. Mechanosensitivity before and after hysterectomy: a prospective study on the prediction of acute and chronic postoperative pain. Br J Anaesth 2011;107: 940–7. [9] Cole LJ, Farrell MJ, Duff EP, Barber JB, Egan GF, Gibson SJ. Pain sensitivity and fMRI pain-related brain activity in Alzheimer’s disease. Brain 2006;129: 2957–65. [10] Cole LJ, Gavrilescu M, Johnston LA, Gibson SJ, Farrell MJ, Egan GF. The impact of Alzheimer’s disease on the functional connectivity between brain regions underlying pain perception. Eur J Pain 2011;15:511–68. [11] Crum RM, Anthony JC, Bassett SS, Folstein MF. Population-based norms for the mini-mental state examination by age and educational level. JAMA 1993;269: 2386–91. [12] de Vet HC, Terwee CB, Knol DL, Bouter LM. When to use agreement versus reliability measures. J Clin Epidemiol 2006;59:1033–9. [13] Edwards RR, Fillingim RB, Ness TJ. Age-related differences in endogenous pain modulation: a comparison of diffuse noxious inhibitory control subjects in healthy older and younger adults. PAINÒ 2003;101:155–65. [14] Folstein MF, Folstein SE, McHugh PR. ‘‘Mini-mental state’’. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189–98. [15] Gibson SJ, Voukelatos X, Ames D, Flicker L, Helme RD. An examination of pain perception and cerebral event-related potentials following carbon dioxide laser stimulation in patients with Alzheimer’s disease and age-matched control volunteers. Pain Res Manag 2001;6:126–32. [16] Hadjistavropoulos T, Fitzgerald TD, Marchildon GP. Practice guidelines for assessing pain in older persons with dementia residing in long-term care facilities. Physiother Can 2010;62:104–13. [17] Horgas AL, Tsai PF. Analgesic drug prescription and use in cognitively impaired nursing home residents. Nurs Res 1998;47:235–42. [18] Jensen-Dahm C, Vogel A, Waldorff FB, Waldemar G. Discrepancy between selfand proxy-rated pain in Alzheimer’s disease: results from the Danish Alzheimer intervention study. J Am Geriatr Soc 2012;60:1274–8. [19] Korner EA, Lauritzen L, Nilsson FM, Wang A, Christensen P, Lolk A. Mini mental state examination. Validation of new Danish version. Ugeskr Laeger 2008;170: 745–9. [20] Kunz M, Mylius V, Scharmann S, Schepelman K, Lautenbacher S. Influence of dementia on multiple components of pain. Eur J Pain 2009;13:317–25.

7

[21] Kunz M, Mylius V, Schepelmann K, Lautenbacher S. Effects of age and mild cognitive impairment on the pain response system. Gerontology 2009;55: 674–82. [22] Kunz M, Scharmann S, Hemmeter U, Schepelmann K, Lautenbacher S. The facial expression of pain in patients with dementia. PAINÒ 2007;133:221–8. [23] Mantyselka P, Hartikainen S, Louhivuori-Laako K, Sulkava R. Effects of dementia on perceived daily pain in home-dwelling elderly people: a population-based study. Age Ageing 2004;33:496–9. [24] Mathuranath PS, Nestor PJ, Berrios GE, Rakowicz W, Hodges JR. A brief cognitive test battery to differentiate Alzheimer’s disease and frontotemporal dementia. Neurology 2000;55:1613–20. [25] McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology 1984;34:939–44. [26] Mitchell LA, MacDonald RA, Brodie EE. Temperature and the cold pressor test. J Pain 2004;5:233–7. [27] Morrison RS, Siu AL. A comparison of pain and its treatment in advanced dementia and cognitively intact patients with hip fracture. J Pain Symptom Manage 2000;19:240–8. [28] Nelson PT, Braak H, Markesbery WR. Neuropathology and cognitive impairment in Alzheimer disease: a complex but coherent relationship. J Neuropathol Exp Neurol 2009;68:1–14. [29] Neziri AY, Scaramozzino P, Andersen OK, Dickenson AH, Rendt-Nielsen L, Curatolo M. Reference values of mechanical and thermal pain tests in a painfree population. Eur J Pain 2011;15:376–83. [30] Pfeffer RI, Kurosaki TT, Harrah Jr CH, Chance JM, Filos S. Measurement of functional activities in older adults in the community. J Gerontol 1982;37: 323–9. [31] Rainville P, Doucet JC, Fortin MC, Duncan GH. Rapid deterioration of pain sensory-discriminative information in short-term memory. PAINÒ 2004;110: 605–15. [32] Rolke R, Baron R, Maier C, Tolle TR, Treede RD, Beyer A, Binder A, Birbaumer N, Birklein F, Botefur IC, Braune S, Flor H, Huge V, Klug R, Landwehrmeyer GB, Magerl W, Maihofner C, Rolko C, Schaub C, Scherens A, Sprenger T, Valet M, Wasserka B. Quantitative sensory testing in the German research network on neuropathic pain (DFNS): standardized protocol and reference values. PAINÒ 2006;123:231–43. [33] Rub U, Del TK, Del TD, Braak H. The intralaminar nuclei assigned to the medial pain system and other components of this system are early and progressively affected by the Alzheimer’s disease-related cytoskeletal pathology. J Chem Neuroanat 2002;23:279–90. [34] Scherder E, Bouma A, Borkent M, Rahman O. Alzheimer patients report less pain intensity and pain affect than non-demented elderly. Psychiatry 1999;62:265–72. [35] Scherder E, Bouma A, Slaets J, Ooms M, Ribbe M, Blok A, Sergeant J. Repeated pain assessment in Alzheimer’s disease. Dement Geriatr Cogn Disord 2001;12: 400–7. [36] Scherder E, Herr K, Pickering G, Gibson S, Benedetti F, Lautenbacher S. Pain in dementia. PAINÒ 2009;145:276–8. [37] Scherder EJ, Bouma A. Visual analogue scales for pain assessment in Alzheimer’s disease. Gerontology 2000;46:47–53. [38] Scherder EJ, Sergeant JA, Swaab DF. Pain processing in dementia and its relation to neuropathology. Lancet Neurol 2003;2:677–86. [39] Schweinhardt P, Bushnell MC. Pain imaging in health and disease—how far have we come? J Clin Invest 2010;120:3788–97. [40] Stokholm J, Vogel A, Johannsen P, Waldemar G. Validation of the Danish Addenbrooke’s cognitive examination as a screening test in a memory clinic. Dement Geriatr Cogn Disord 2009;27:361–5. [41] Tousignant-Laflamme Y, Page S, Goffaux P, Marchand S. An experimental model to measure excitatory and inhibitory pain mechanisms in humans. Brain Res 2008;1230:73–9. [42] Villemure C, Schweinhardt P. Supraspinal pain processing: distinct roles of emotion and attention. Neuroscientist 2010;16:276–84. [43] von Baeyer CL, Piira T, Chambers CT, Trapanotto M, Zeltzer LK. Guidelines for the cold pressor task as an experimental pain stimulus for use with children. J Pain 2005;6:218–27. [44] Wiech K, Tracey I. The influence of negative emotions on pain: behavioral effects and neural mechanisms. Neuroimage 2009;47:987–94. [45] Yesavage JA. Geriatric Depression Scale. Psychopharmacol Bull 1988;24: 709–11.

Please cite this article in press as: Jensen-Dahm C et al. Quantitative sensory testing and pain tolerance in patients with mild to moderate Alzheimer Ò disease compared to healthy control subjects. PAIN (2014), http://dx.doi.org/10.1016/j.pain.2013.12.031