Study of experimental pain measures and nociceptive reflex in chronic pain patients and normal subjects

Study of experimental pain measures and nociceptive reflex in chronic pain patients and normal subjects

Pain, 44 (1991) 131-138 0 1991 Elsevier Science Publishers ADONIS 030439599100066Y 131 B.V. 0304-3959/91/$03.50 PAIN 01724 Study of experimental pa...

1MB Sizes 0 Downloads 70 Views

Pain, 44 (1991) 131-138 0 1991 Elsevier Science Publishers ADONIS 030439599100066Y

131 B.V. 0304-3959/91/$03.50

PAIN 01724

Study of experimental pain measures and nociceptive reflex in chronic pain patients and normal subjects F. Boureau, M. Luu and J.F. Doubrke Centre d’Evaluation et de Traitement (Received

de In Douleur, Hapital Saint-Antoine,

27 April 1990, revision received

12 July 1990, accepted

Paris (France)

3 August

1990)

This study evaluates (i) the effect of heterotopic chronic pain on various experimental pain measures, Summary (ii) the relationship between experimental pain measures and chronic pain symptomatology assessment, and (iii) the influence of the various pain aetiologies on experimental pain measures. Fifty-three chronic pain patients were compared to 17 pain-free subjects with the following psychophysical and physiological indices: pain threshold (PTh), pain tolerance (PTol), verbal estimation of intensity and unpleasantness (intensity scale, IS; unpleasantness scale, US), threshold for intensity and unpleasantness (ITh and UTh), lower limb RI11 nociceptive reflex (RIIITh and RI11 frequency of occurrence). Chronic pain syndromes included neuropathic pain (n = 12), iodopathic pain (n = 12), myofascial syndromes (n = 9) headache (n = 9), and miscellaneous pain (n = 11). Chronic pain symptomatology was assessed with a visual analogue scale (VAS), a French MPQ adaptation (QDSA), Beck Depression Inventory (BDI), Spielberger State Trait Inventory (STAI) and Eysenck Personality Inventory (EPI). No significant difference was observed between chronic pain patients and pain-free control groups and between patient subgroups for PTh, PTol and RIIITh. No significant correlation was found between experimental pain measures and clinical pain, anxiety or depression scores. However, the chronic pain patients had a higher threshold for unpleasantness and judged the suprathreshold stimuli significantly less intense and less unpleasant than the control group. These results are discussed in relation to diffuse noxious inhibitory controls and the adaptation level theory of chronic pain experience. Key words: Chronic pain; Electrical stimulation; tolerance

Magnitude estimation:

Introduction

Clinical pain involves perceptual, affective, cognitive, evaluative and behavioral factors 112,281. Clinical pain assessment must rely on subjective or behavioral variables since available measurements of the peripheral or central generators are lacking. In contrast, in the laboratory, noxious stimuli and verbal responses (or other physiological indices) can be quantified. The introduction of experimental pain in the clinical setting may

Correspondence to: Dr. Franqois Boureau, Centre d’Evaluation et de Traitement de la Douleur, H&pita1 Saint Antoine, 184 Rue du Fg. St.-Antoine, 75012 Paris, France.

Nociceptive

reflex; Pain threshold;

Pain

gain more objectivity, permitting, for example, parallel study of treatments on clinical and experimental pain. However, the physiological or psychological mechanisms by which a clinical pain sensation may influence the transmission of experimental pain information or the perceptual processes are not fully understood. Several different physiological and/or psychological factors can be involved in the influence of clinical pain on experimental measures. Various studies [5,10,13] have assessed thresholds in the segmental distribution of the pain, while other studies [4,7,8,13-15,17,19,20,23,27,33,35,36,42] mainly focus on extrasegmental distant effects of clinical pain. In the clinical conditions, one might expect cutaneous sensibility dysfunction (i.e., hyperalgesia, allodynia.. .) in the

pain area. The corresponding underlying neurophysiologicat mechanism most probably implicates convergence of excitatory information on spinal or supraspinal neurons which could result in lowered thresholds to different stimuli. including pain. applied in the metameric area. In contrast to these intrasegmental changes, other possible extrasegmental effects could be related to mechanisms such as diffuse noxious inhibitory controls (DNICs) [1X]. Heterotopic (i.e., distant) pain-relieving effects, evoked by noxious stimuli applied to widespread areas of the body, have been described in animals [18] and humans f39]. DNICs affect activities from convergent (or wide-dynamic-range) neurons evoked by noxious stimulation and are mediated by a loop which ascends to supraspinal structures within the ventrolateral quadrant and descends within the dorsolateral funiculus [18,39,40]. From a psychological point of view, two different conceptual models have theorized the possible influence of clinical pain on pain measures. Chapman [6] hypothesized that patients with persistent pain, because of social reinforcement, would develop perceptual habits of hype~igilance for somatic distress signals. This exaggerated focus on internal sensations may shift nonpainful sensations to pain. The clinical implications of the hypervigilance model may be related to the stimulus generalization theory of Fordyce [9]. Stimulus generalization states that a particular response which is at first linked to a specific stimulus could after some time generalize to other stimuli (i.e., proprioceptive sensations, stressful situations). According to the hypervigilance and stimulus generalization models, chronic pain patients when receiving an acute experimental stimulus would react with exaggerated reactivity and would have decreased pain threshold and/or pain tolerance in comparison with a control group. In contrast with the above hypothesis, Rollman [31] suggested that pain patients would evaluate experimental pain within the context of their previous experience with pain, and proposed an adaptation level theory. According to this model [16,31,32], pain judgments are based on comparisons with other pain levels. This suggests that subjects experiencing chronic pain may utilize their endogenous pain levels as an ‘anchor’ in describing external painful stimuli. This model predicts that, because of their internal discomfort, chronic pain patients should have higher pain threshold and tolerance than controls and should judge external painful stimuli as less severe than would pain-free individuals. Hypervigilance theory and adaptation level theory predict opposite reactivity for chronic pain patients submitted to an experimental stimulus. The present study is mainly concerned with the effects of clinical pain on heterotopic pain measures. Of the 16 related studies reviewed in the literature, 8 [7,14,15,17,20.27,41,42] found that chronic pain experience increased the thresholds to various experimental

pain stimuli, 6 found that chronic pain decreased rather than increased pain threshold [4,19,2?.33.35.36], and 2 found no significant effects [X,13]. Many factors can account for these conflicting results. such as type of stimuli. type of pain disorders. type of pain measures. It has been hypothesized that paradigms with stimuli 01 less clinical relevance make adaptation-level responding more likely [26]. Affective stimuli will elicit adaptationlevel behavior. Because of the conflicting data. we undertook the present study to examine (i) the effect of heterotopic chronic pain on different experimental pain measures including psychological and physiological indices (subjective pain thresholds, pain tolerance. verbal estimation of intensity and unpleasantness of the stimuli, intensity and unpleasantness thresholds and a lower limb nociceptive reflex evoked by electrical stimuli). (ii) the influence of Ihe various chronic pain etiologies on the experimental pain measure changes, and (iii) the relationship between experimental measures and the clinical data (i.e., pain, anxiety and depression levels).

MethtnIs Subjects

The 53 chronic pain patients (CPP, 39 F, 14 M) were recruited from the Saint-Antoine multidisciplinary outpatient pain unit. Inclusion criteria were: pain duration of more than 6 months, clinical pain heterotopic to experimental pain test (i.e., homolateral lower limb), verbal aptitude to correctly complete the verbal assessment, informal agreement to the experimental procedure. Exclusion criteria were: polymedication, strong narcotic intake, lower limb sensory or motor dysfunction. The average age was 49.1 years (range 21-75 years). The mean pain duration was 6.6 years (range 6

TABLE

I

Comparison of chronic pain patients (CPP) and control subjects (CS) according to age. sex ratio (F: M), pain threshold (PTh). pain tolerance (PTol), RI11 threshold (RIIITh). intensity threshold (ITh) and unpleasantness threshold (UTh). Statistical comparison is performed using the Mann and Whitney test. CPP (m

(SD.11

CS

U/U’

Significance

(m tS.D.t)

ITO1

n = 53 49.1 39:14 8.4 (3.9) 17.7 (6.4)

n =I7 31.1 6:ll 8.0 (3.4) 18.6 (4.5)

439,‘461 385/515

N’S NS

RIIlTh

n = 48 12.3 (4.5)

n =I7 12.9 (2.8)

318/446

NS

ITh UTh

n = 32 6.5 (4.0) 8.5 (4.8)

n =17 6.0 (3.0) 5.8 (2.0)

249/294 197/347

NS P <: 0.05

Age F:M PTh

133

months-30 years). The diagnoses of the patients were: neuropathic pain (nerve lesion, n = 7; post-herpetic pain, n = 4; phantom pain, n = l), idiopathic pain (pelvic and abdominal pain, n = 7; facial pain, n = 5) tension headache (n = 2) or mixed headache (n = 7) myofascial pain syndromes (fibromyalgia, n = 5; localized myofascial pain syndrome, n = 4), miscellaneous (post-surgical, n = 3; post-traumatic, n = 3; reflex sympathetic dystrophy, n = 2; Tables I and II). Patients were expected to take as few medications as possible. It seemed, however, unrealistic and probably unethical to expect a satisfactory wash-out, particularly for long half-life drugs such as antidepressants. No patients were taking strong narcotics or neuroleptics, 22 patients were taking peripheral analgesics (n = 13) or weak opioids (propoxyphene or codeine, n = 9) 22 patients were taking antidepressants, 19 were taking anxiolytics, 17 were taking miscellaneous other drugs (anticonvulsants, n = 5; antimigraine drugs, n = 3). Ten patients were not taking any drugs. Twenty-one patients were taking only 1 medication, 21 were taking 2 or 3 medications. A single patient had more than 3 different medications. As part of the routine evaluation, the chronic pain patients completed a comprehensive pencil and paper assessment including pain visual analogue scale (VAS), a French adaptation of the McGill Pain Questionnaire,

TABLE

the Questionnaire Douleur St. Antoine (QDSA) which permits calculation of a total score (QDSA-T), a sensory score (QDSA-S) and an affective score (QDSA-A) [l-3], the Beck Depression Inventory (BDI), the Spielberger State Trait Inventory (STAI S and T) and the Eysenck Personality Inventory (EPI E, N, L). All the patients suffered from persistent continuous pain which was present at the same time as the experimental pain measurement. All the patients suffered from clinical pain in a heterotopic topography, distant from the site for the experimental pain testing locus (Table II). Seventeen normal volunteers (6 F, 11 M) were recruited from among colleagues and students as control subjects (CS). All were medication-free at the time of evaluation. The mean age was 31.1 years (range 20-43 years). General psychophysical procedure After being carefully briefed about the experimental procedure, the subjects gave informed consent. During the experimental session the subjects were sitting in a comfortable armchair to obtain good muscular relaxation. Simultaneous recordings of reflex activities and verbal description were performed. The session was composed of preliminary training lasting 10 min. During a first session, stimulation was delivered in 10

II

Comparison of patient subgroups (neuropathic pain, idiopathic pain, headache, myofascial syndromes and miscellaneous pain) for age, pain duration, pain topography, visual analogue scale (VAS), Eysenck Personality Inventory (extraversion, neuroticism and lie, EPI E, N and L), Questionnaire Douleur Saint Antoine (total, sensory and affective, QDSA T, S and A), Spiegelberger State Anxiety Inventory (STAI-S), Beck Depression Inventory (BDI). RI11 threshold (RIIITh). pain threshold (PainTh) and pain tolerance (PainTol). Statistical comparison is performed by ANOVA. Neurop. (n = 12)

Idiop. (n = 12)

Headache (n = 9)

Myofasc. (n = 9)

Misc. (n = 11)

F test

54.8 8:4

46.7 11:l

43.0 6:3

50.8 514

49.1 9~2

NS

1.6

7.1

7.5

5.6

6 3 _ 1 1 1

1 4 7 _ _

_ 9 _

7 3 1

_

_ _ _

8 2 _

2 1 _ _

VAS EPI-E EPI-N EPI-L

7.0 11 12.1 5.7

6.9 9.7 17.4 4.9

5.4 7.8 15.3 3.8

QDSA-T QDSA-S QDSA-A STAI-S BDI

31.7 17.8 13.9 49.5 14.4

36.5 15.3 21.2 50.6 15.7

31.7 12.3 19.4 46.3 15.1

6.2 11.3 13 5 39.2 20.1 19.1 50.6 11.2

7.8 9.3 14.3 5.9 42.8 22.5 20.3 47.5 16.1

NS NS NS NS NS

RIIITh PainTh PainTol

11.7 8.5 17.2

12.5 7.4 17.9

11.3 7.2 17.6

12.8 7.4 15.1

13.9 11.0 20.5

NS NS NS

Age (years) F:M Pain duration (years) Pain topography Upper limb Head Abdomen/pelvis Low back Cl. lower limb Other

13

P = 0.006

6 2

P = 0.01 P = 0.01 NS NS

1.14

consecutive ascending series with a 1 mA step (Methods of limits). Muscular reflex activities {RIII reflex, see later) were recorded for each stimulation. the subjects describing the evoked sensation using the painfulness scale (PS). This procedure was completed by the 53 chronic pain patients and control subjects. A second session was added for the late 32 chronic pain patients. In this session. the electrical stimulus was delivered at 5 fixed intensity levels ranging from 3 to 15 mA (Method of constant stimuli) [21]. Each intensity was delivered 10 times. The series was constructed so that sequential effects could be controlled. Instructions were given to clarify the distinction between the intensity and the unpleasantness dimension of pain. Each stimulation was described simultaneously for intensity and unpleasantness using an intensity scale (IS) and an unpleasantness scale (US). The RI11 threshold (RIIITh) was calculated with both limits and constants methods. The overall experimental session did not exceed 60 min.

The method for stimulating the surat nerve and recording the RI11 reflex activity from biceps femoris was similar to that used in other studies 13’7,381. The sural nerve was stimulated behind the lateral malleolus through a pair of surface electrodes on degreased skin. The electrical stimulus was delivered by a constant intensity stimulator and consisted of a 20 msec train of 6 rectangular pulses (1 msec duration each). A warning signal predicted the stimulation. The time interval between two successive stimulations varied with an approximate mean interval of 10 sec. Reflexes responses were recorded from the ipsilateral biceps femoris muscle using a pair of surface electrodes on the degreased skin overlying the muscle. The latency of a reflex activity corresponding to a RI11 response ranged from 90 to 180 msec. The response was considered present when an amplitude greater than 50 PV was observed. The RI11 threshold (RIIITh) was calculated as the intensity which gave 50% of RI11 responses. The RI11 threshold calculation was based on the RIII frequency of occurrence [37,38] and not on the 10% RI11 maximal amplitude 1391. The frequency of occurrence of the RI11 response was calculated for each intensity level.

For the late 32 chronic pain patients. verbal magnitude estimation of stimuli was also performed with two 7-category scales: a sensory or intensity scale (IS) and an affective or unpleasantness scale (US). The IS and US were respectively constructed with sensory/intensity and affective descriptors. The ordinal structure of these two scales was validated in our laboratory 1211. Items of the IS were: 7 -‘nothing.’ 6 =‘just noticeable,’ 5 = ‘ tactile sensation.’ 4 = ‘light pricking.’ 3 = ‘moderate pricking,’ 2 =‘strong pricking,’ I =‘very strong pricking.’ Items of the US were: 7 =‘nothing,’ 6 = ‘not unpleasant.’ 5 =‘weakly unpleasant,’ 4 =‘rather unpleasant.’ 3 = ‘clearly unpleasant,’ 2 = ‘extremely unpleasant,’ and 1 =‘unbearable.’ An intensity threshold (ITh) was defined as the electrical intensity level which gave 50% of ‘light pricking’ response. Similarly, an unpleasantness threshold (UTh) was defined as the electrical intensity which gave 50%, of ‘weakly unpleasant’ response. For each stimulation intensity level, the mean scores for IS and US were calculated.

Results Comparison of chronic pain patients and control subjects The mean RIIITh was 12.3 (S.D. = 4.8; range 5-30 mA; n = 45) for chronic pain patients and 12.9 for control subjects (S.D. = 2.8; range 6.8-17.2 mA, n = 17; Table I and Fig. I). No significant difference was found between pain patients and control groups (U/U’ = 318/446). RIIITh calculated by limits and constants methods were highly correlated, the Spearman coefficients being 0.95 and 0.73, respectively, for chronic patient group and control subjects. The RIIITh could not be calculated for 7 patients: in 4 patients the reflex response could not be recorded even for high intensity stimulation and in 3 patients the maximum frequency of occurrence did not reach the 50% level for the highest tolerated electrical intensity (PTol). For this particular

NS

NS

Pain verbal responses were assessed with a 7-category pain scale. The English translation of the painfulness scale (PS) is: 0 =‘nothing,’ 1 =‘just noticeable,’ 2 = ‘non-painful,’ 3 = ‘weakly painful,’ 4 = ‘rather painful,’ 5 = ‘extremely painful,’ 6 = ‘unbearable, do not increase any more.’ Pain threshold (PTh) was calculated as the intensity level which gave 50% of ‘weakly painful’ responses. Pain tolerance (PTol) was determined as the maximal intensity judged as ‘unbearable, do not increase any more.’

/ /

! / / 10 i

Lil ITh

Fig. 1. Comparison of chronic pain patients (CPP) and control subjects (CS) according to intensity threshold (ITh) and unpleasantness threshold (UTh). pain threshold (PTh), RI11 threshold (RlfIThf and pain tolerance (PTol). Statistical comparison is performed using the Mann and Whitney test ( * P < 0.05).

135

,oo%

UNPLEASANTNESS SCALE

R ii! OCCURRENCE

I

M *-*

CPP cs

NS :

n;29

” .I7

I’

‘1

-

CPP

n-32

21

.-.

cs

“.I7 **

T

3.

.:/

41 5’ 61 7’ o-

6

3

9

15 mA

12

Fig. 2. Mean RIII occurrence (%) for each electrical stimulation mtensity for chronic pain patients (CPP, n = 29) and control subjects (CS, n = 17, NS, not significant).

patient subgroup, the mean PTol (13.3 mA, range 6-27 mA) was lower than for other pain patients. The RI0 occurrences for 3, 6, 9, 12 and I5 mA intensity levels were not significantly different between CPP and CS. Fig. 2 indicates that the RIII occurrence at 15 mA tends to be lower for CPP than for CS but the difference did not reach statistical significance. The mean PTh was 8.4 & 3.9 mA for CPP (n = 53) and 8.0 + 3.4 mA for CS (n = 17). No significant difference was found between CPP and CS for PTh (Table I and Fig. 1). The mean PTol was 17.7 f 6.4 mA for CPP and 18.8 + 4.5 mA for CS. No significant difference was found between CPP and CS for PTol. Figs. 3 and 4 show the average verbal estimation for intensity and unpleasantness for different levels of stimulation: 3,6,9, 12 and 15 mA. For the 15 mA level, the mean scores of CPP (n = 32) for intensity (3.0 rt 0.9) and unpleasantness (4.6 f 0.9) were significantly different than for CS (n = 17) (respectively 2.3 + 0.6, W = 245, P < 0.03 and 3.5 + 0.8, W = 212, P -C0.01). There was no significant difference for intensity and unpleasantness scores between the two compared groups for other intensity levels. INTENSITY SCALE 1 2

1

0

-

*.-*

3

CPP CS

3

6

9

12

---, 15 mA

Fig. 4. Mean scores for unpleasantness scale (US) for each electrical stimulation level (m.4) for chronic pain patients (CPP, n = 29) and control subjects (CS, n = 17; * * P < 0.01).

The mean ITh was 6.5 + 4.0 mA for CPP and 6.0 + 3.0 mA for CS. There was no significant difference between CPP and CS. The mean UTh was 8.5 rt 4.8 mA for CPP and 5.8 _t 2.0 mA for CS. The difference was statistically significant (U/U’ = 197/347, P < 0.05; Fig. 1). Relation between experimental pain measures and chronic pain symptomatology RIIITh was highly correlated with PTh (0.62) and PTol (0.77). No correlation was observed between experimental pain measures (RIIITh, PTh or PTol) and clinical data related to pain (VAS, QDSA-T, QDSA-S, QDSA-A), anxiety (STAI-S, STAI-T), depression (BDI) and EPI scores (E. N, I). No significant difference according to sex was observed for CPP (RIIITh = 12.1 + 3.3 mA for males and 12.6 + 5.3 mA for females). The incidence of medication intake was also examined; for the 13 medicationfree patients RIIITh (12.3 + 3.8 mA) was not significantly different from other patients RIIITh (12.5 rsI:3.8 mA). The influence of etiology on experimental pain measures ANOVA showed that there was no significant difference between the patient groups for RIIITh, PTh and PTol.

nz32

n-17

6

Discussion

9

12

15 mA

Fig. 3. Mean scores for intensity scale (IS) for each electrical stimulation Ievel (mA) for chronic pain patients (CPP. n = 29) and control subjects (CS, n = 17, * P < 0.05).

We had expected more significant differences between chronic pain patients and control subjects. The main result is that no significant difference was observed between chronic pain patients and pain-free control groups and between patient subgroups for pain and tolerance thresholds and RI11 reflex threshold. No significant correlation was found between experimental pain measures and clinical pain, anxiety or depression scores. However, it is interesting to observe that chronic

pain patients had a higher threshold for unpleasantness and judged the higher stimuli significantly less severe for intensity and unpleasantness than did the control group. The present study encountered some difficulties. We can hypothesize than our chronic pain patient sample is not representative of a chronic pain population since we selected patients with a drug intake as low as possible. As a group, the sample analyzed in this study is very similar according to the chronic pain symptomatology assessment to a non-selected group of consecutive patients examined in our center [2]. The chronic pain group and the normal control group are also different according to average age. This factor may have influenced our results. However, the RIII thresholds observed in our study are very similar to the results observed in previous studies using similar experimental conditions (37,381. One other limitation could be the problem of missing data for high intensity stimulations. When we were studying fixed intensity stimulation from 3 to 15 mA, some patients had a pain tolerance lower than the highest intensity delivered (15 mA). These patients would be expected to have a high response since this intensity is above the pain tolerance. For ethical reasons, the study of stimuli at such intensity was discontinued. This problem was not encountered with control subjects, but was observed with 6 chronic pain patients. This factor may have influenced the results for the 15 mA intensity levels. However, when we extrapolated the missing data with corresponding highest subject responses the statistical significance was still observed. Apart from these limitations, our results provide new data related to the influence of heterotopic clinical pain on pain measures. The overall chronic pain patient sample did not show any significant difference on subjective pain threshold, pain tolerance and RI11 threshold in comparison with control subjects. These negative results may contrast with other studies which show either increased or decreased pain or tolerance thresholds. This discrepancy may be explained by numerous factors: acute versus chronic pain, pain etiology, type of nociceptive stimulus. Another study with 8 patients with acute sciatica found the RI11 threshold in usual range (10 mA) and no significant difference between normal and painful sides [40]. Lastgue’s maneuver performed on the painful side resulted in a large reduction of the nociceptive reflex either elicited on the normal side (80% depression) or on the painful side (74% depression). This study performed in patients with sciatica demonstrates that an acute neurological pain triggered by Lasegue’s maneuver can result in a profound depression of the nociceptive reflexes to sural nerve stimulation. Tactile, cold. warm and heat thresholds were studied in oral postoperative acute pain and no significant difference was observed for thresholds

at either the homotopic area or the distant IOCLI~ / 131. These studies related to acute pain may be discusacd according to the ‘pain inhibits pain’ phenomenon and diffuse noxious inhibitory controls [IX]. In animals and humans [18.29,40] it has been observed that heterotopic nociceptive stimuli produced by experimental pain reduced lower limb nociceptive reflex and associated pain sensations in normal humans. The study with sciatica [40] illustrated the influence of a phaaic acute pain on experimental measures. It is worth noting. however. that the control value of the sub.ject in a relaxed position tloes not exhibit significant modification in comparison to usual values. In our chronic pain group we did not attempt to increase the level of the clinical pain to study an effect on the reflex measures. Since phasic acute pain has been shown to modify reflex activities in other studies, we suggest that the observed discrepancy between our results and other studies [39.40] is related to the tonic properties of chronic pain or to too low intensity levels to induce a ‘pain inhibits pain’ phenomenon. DNIC does not seem to play any significant role on the threshold values. at least at the chronic stage of a clinical pain. One interesting result is the selective increase of the unpleasantness threshold for chronic pain patients in comparison with control subjects. This result supports the adaptation theory, which is supported by other experiments [7,14,15,17.20,27,41,42]. Pain thresholds for 3 set thermal stimuli applied to the volar forearm have been studied in 8 patients suffering from chronic low back pain and for 6 pain-free control subjects [36]. The mean threshold for chronic pain patients was significantly greater than the threshold for the pain-free subjects. The difference was investigated in a verbal sensory and affective descriptor scaling procedure. The difference was observed for unpleasantness but not for sensory intensity responses. Thresholds to radiant heat stimuli were compared in low back pain patients and non-patient controls [27]. The pain patients had substantially higher pain thresholds than controls and showed poorer discrimination during the signal detection task at lower intensity levels. These results were replicated in another study 171. The results of both studies [7,27] have been explained in terms of adaptation level theory [16]. It is interesting to observe that in our study the adaptation processes mainly concerned the affective component of pain since unpleasantness threshold as well as verbal estimation for unpleasantness were modified. The discomfort associated with the experimental stimulation is reduced because it is compared with the much greater discomfort of the clinical pain sensation. However, all studies are not in accord with an adaptation level theory for chronic pain perception. It has been reported that patients with myofascial pain dysfunction syndrome had lower pain thresholds for con-

137

stant pressure than normal controls, were less able to discriminate varying intensities of pressure stimulation and demonstrated a greater tendency to report pain as compared to normal control subjects [23]. The comparison of 20 patients with fibrositis with a group with rheumatoid arthritis and normal controls indicated a lower pain threshold and tolerance for the fibrositis with a pressure dolorimeter but not for electrical stimulation or constant pressure. TWO studies [4,35] compared chronic low back pain patients and control subjects undergoing a cold pressor test and observed that chronic low back pain patients demonstrate poorer acute pain tolerance and report higher acute pain. All these studies support a hype~i~lance model [6] or a stimulus generalization theory [9,34]. At the present time it is hard to explain the discrepancy in the literature. One attractive hypothesis is that the nature of the experimental pain stimulus may trigger either a hypervigilant response or an adaptation response (261. Our results may support this hypothesis. The 20 msec electric shock trains used in our study, usually described as a pricking sensation, may appear to chronic pain subjects very different from their internal persistent pain discomfort. The discrepancy in the literature may be explained by pain of various etiologies. For example, it was found that patients with organic causes for their pain had higher pain complaint thresholds than psychiatric patients [25]. It is worth noting that in our study we were not able to detect a difference between chronic pain subgroups on the experimental pain measures. One important issue is the value of indices such as pain threshold and pain tolerance as a measure of ‘pain’ index [28]. These indices have been criticized since their variation may confound different sensory, affective or cognitive factors. One interesting result is that the verbal estimation with intensity and affective scales may reflect a more specific aspect of the pain experience. Our study may indicate that these indices may be more sensitive since for high intensity levels we observed that chronic pain patients showed lower scores for intensity and unpleasantness. This approach is interesting but difficult, since according to the variation of pain tolerance, subjects with low tolerance could not receive the high intensity stimuli. This has not to our knowledge been discussed in the literature. We do not have a methodological approach to rule out this important difficulty. Experimental pain measures were not related to the clinical data, i.e., pain severity, depression or anxiety levels, type of pain etiology. Chronic clinical pain does not appear to modify significantly, at least in our study, the major experimental pain measures. Since clinical pain per se does not significantly modify experimental pain measures, it is even more interesting to introduce experimental pain in the clinical setting and conduct

parallel studies [29,30]. The effect of a treatment on experimental pain measures may not reflect an indirect effect due to a clinical pain-relieving effect but rather a direct effect of the treatment on the experimental pain measures. This approach is interesting to explore clinical pain mechanisms further since various authors have observed that the change in experimental pain induced by a treatment might serve as a confirmatory or prognostic index of success ]5,11,20,22]. In contrast, other authors have observed a significant clinical effect without modification of experimental pain measures 1241.

Acknowledgement We are grateful comments.

to R.H.

Gracely

for his helpful

References 1 Boureau, F. and Paquette, C.. Translated versus reconstructed McGill Pain Questionnaires: a comparative study of two French forms. In: R. Dubner, F.G. Gebhart and M.R. Bond (Eds.), Proc. of the Vth World Congress on Pain, Elsevier, Amsterdam, 1988, pp. 395-402. 2 Boureau, F., Luu, M., Doubrtre, J.F. and Gay, C., Elaboration dun questionnaire d’auto-evaluation de la douleur par liste de qualificatifs. Comparaison avec le McGill Questionnaire de Melzack, Therapie, 39 (1984) 119-139. 3 Boureau, F., Doubrere, J.F.. Luu, M. and Combes. A., Chronic pain assessment. In: Scherpereel et al. (Eds.), The Pain Clinic II, VNU Science Press, Utrecht, 1987, pp. 11-21. 4 Brands, A.M.E.F. and Schmidt, A.J.M., Learning processes in the persistence behavior of chronic low back pain patients with repeated acute pain stimulation. Pain, 30 (1987) 329-337. 5 Calfaghan, M., Stembach, R.A., Nyquist, J.K. and Timmermans, G., Changes in somatic sensitivity during transcutaneous electrical analgesia, Pain, 5 (1978) 115-127. 6 Chapman, C.R., Pain, perception and illusion. In: R.A. Stembach (Ed.), The Psychology of Pain, Raven Press, New York, 1986. 7 Cohen, M.J.. Naliboff. B.D., Schandler, S.L. and Heinrich, R.L., Signal detection and threshold measures to loud tones and radiant heat in chronic low back pain patients and cohort controls, Pain, 16 (1983) 245-252. 8 Ekblom, A. and Hansson, P., Thermal sensitivity is not changed by acute pain or afferent stimulation, J. Neurol. Neurosurg. Psychiat.. 50 (1987) 1216-1220. 9 Fordyce. W.E., Behavioral conditioning concepts in chronic pain. In: J.J. Bonica et al. (Eds.). Advances in Pain Research and Therapy, Vol. 5, Raven Press, New York, 1983. 10 Francini, F., Maresca, M., Procacci, P. and Zoppi, M., The effects of non-painful transcutaneous electricai nerve stimuiation on cutaneous pain threshold and muscular reflexes in normal men and in subjects with chronic pain, Pain, 11 (1981) 49-63. 11 Garcia-Larrea, L., Sindou, M. and Mauguiere, F., Nociceptive flexion reflexes during analgesic neurostimulation in man, Pain, 39 (1989) 145-156. 12 Gracely, R., R.H., Methods of testing pain m~hanisms in normal man. In: P.D. Wail and R. Melzack @is.), Textbook of Pain, Churchill Livingstone, Edinburgh, 1989, 257-268.

I3 Hansson. P., Ekblom, A.. Lmdblom. U. and Marchettmi, P.. Does acute intraoral pain alter cutaneous sensibility? J. Neural. Neurosurg. Psychiat.. 51 (1988) 1032-1036. 14 Hapidou, EL. and De Catanzaro. D.. Sensitivity to cold pressor pain in dysmenorrheic and non-dysmenorrheic women as a function of menstrual cycle phase, Pain, 34 (1988) 277-283. 15 Hazouri. L.A. and Mueller. A.D.. Pain threshold studies on paraplegic patients. Arch. Neurol. Psychol.. 64 (1950) 6077613. 16 Helson. II., Adaptation-Level Theory: an Experimental and Systematic Approach to Behavior, Harper and Row, New York. 1964. 17 Husk&on, EC. and Hart. F.D., Pain threshold and arthritis, Br. Med. J., 28 (1972) 193-195. I8 Le Bars, I>., Dickenson, A.H. and Besson. J.M., Diffuse noxious inhibitory controls (DNIC). I. Effects on dorsal horn convergent neurones in the rat. Pain, 6 (I9791 283-304. I9 Langermank, M.. Jensen. K.. Jensen. T.S. and Olesen, J., Pressure pain thresholds and thermal nociceptive thresholds in chronic tension-type headache, Pain, 3X (1989) 203-210. 20 Lipman. J.J.. Blumenkopf. B. and Parris, W.C.V. Chronic pain assessment using heat beam dolorimetry. Pain, 30 (1987) 59-67. 21 Luu. M.. Bonnel. A.M. and Boureau. F.. Multidimensional experimental pain study in normal man: combining physiological and psycholo@aI indices, In: R. Dubner et al. (Eds.), Proc. Vth World Congress on Pain, Elsevier Science Publishers. Amsterdam 1988. pp. 375-382. 22 Malow, R.M. and Olson. R.E., Changes in pain perception after treatment for chronic pain, Pain, I1 (1981) 65-72. 23 Malow, R.M., Grimm, L. and Olson, R.E., Differences in pain perception between myofascial pain dysfunction patients and normal subjects: a signal detection analysis. J. Psychosom. Res.. 24 (1980) 303-310. 24 Max, M.B., Culnanz. M., Schafer, SC., Gracely, R.H.. Walther, D.J.. Smoller, B. and Dubner, R., Amitriptyline relives diabetic neuropathy pain in patients with normal or depressed mood, Neurology, 37 (1987) 589-596. 25 Merskey, H. and Evans, P.R., Variations in pain complaint threshold in psychiatric and neurological patients with pain, Pain. 1 (1975) 73-79. 26 Naliboff. B.D. and Cohen, M.J., Psychophysical Iaboratory methods applied to clinical pain patients. In: CR. Chapman and J.D. Loeser (Eds.), Issues in Pain Measurement, Raven Press, New York. 1989. pp. 365-386. 27 Nahboff, B.D.. Cohen, M.J., Schandler. S.O. and Heinrich. R.L.. Signal detection and threshold measures for chronic back pain patients, chronic illness patients and cohort controls to radiant heat stimuli. J. Abnorm. Psychol., 90 (1981) 271-274.

2X Price, D.D., Psychological and Neurai Me~hanI~rns of Pain. Raveu Press, New York. 1988. 241 pp. 29 Price, D.D. and Harkins. SW.. Combined use of experimental pain and visual analogue scales in providing standardized measurement of clinical pain. Clin. J. Pain, 3 (1987) 1 --X. 30 Price. D.D., Harkins, SW.. Radii. A. and Price, C.. A simultaneous comparison of fentanyl’s analgesic effects on experimental and climcal pain, Pain, 24 (1986) 197-203. 31 Rollman. G.B.. Signal detection theory pain measures: empirical validation studies and adaptation-level effects, Pain, 6 (1979) 9-21 32 Rollman, G.B.. Measurements of experimental pain in chronic pain patients: meth~ologica1 and individual factors. In: R. Melzack (Ed.), Pain Measurements and Assessment. Raven Press. New York, 1983, pp. Z-258. 33 Sandrini. G., Martignoni. E.. Micieli, G., Alfonsi, E.. Sances, G. and Nappi. G., Pain reflexes in the ctinical assessment of migraine syndromes, Func. Neural.. 4 (1986) 423-429. 34 Schmidt. A.J.M., The behavioral management of pain: a criticism of a response, Pain. 30 (1987) 2X5-291. 35 Schmidt, A.J.M. and Brands, A.M.E.F., Persistence behavior of chronic low back pain patients in an acute pain situation. J. Psychosom. Res.. 3 (1986) 339-346. 36 Scudds, R.A., Rollman. G.B., Harth, M. and McCain, G.A.. Pain perception and personality measures as discriminators in the classification of fibrositis. J. Rheumatol., 14 (1987) 563-$69. 37 WiIIer. J.C., Comparative study of perceived pain and nociceptive flexion reflex in man. Pain, 3 (1977) 69-80. 38 Witler. J.C., Boureau, F. and Athe-Fessard, D., Supraspinal influences on nociceptive flexion reflex and pain sensation in man, Brain Res.. 179 (1979) 61 -68. 39 Willer, J.C.. Roby. A. and Le Bars, D., Psychophysical and electrophysiological approaches to the pain-relieving effects of heterotopic nociceptive stimuli, Brain. 107 (1984) 1095-1112. 40 Wilier, J.C.. Barranquero. A.. Kahn. M.F. and Salliere, D., Pain in sciatica depresses lower limb nociceptive reflexes to sural nerve stimulation. J. Neurol. Neurosurg. Psychiat.. 50 (1987) l-5. 41 Wolskee, P.J. and Gracely, R.H.. Effect of chronic pain on experimental pain response, Am. Pain Sot.. (1980) Abst. 4: 4. 42 Yang, J.C., Richiin, D.. Brand, L.. Wagner, J. and Clark. W.C.. Thermal sensory decision theory indices and pain thresholds in chronic pain patients and healthy volunteers. Psychosom. Med., 47 (1985) 461-468.