| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Department of Psychology (T.D.B., T.T.B.), Wake Forest University, Winston-Salem, North Carolina; and Cedars Sinai Medical Center (C.D.S.), Los Angeles, California.
Address reprint requests to: Terry D. Blumenthal, PhD, Department of Psychology, Wake Forest University, Winston-Salem, NC 27109. Email: blumen{at}wfu.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONSUMMARYACKNOWLEDGMENTSREFERENCES |
|---|
METHODS: Perceptual and pain thresholds for electrical shocks to the arm were determined in 20 adults. Participants then rated the painfulness of 25 electrical shocks that were 1.5 times the pain threshold (mean shock intensity, 
160 V) and either presented alone or preceded (at 40–60 ms) by weak electrical prepulses equal to or 25% above the perceptual threshold.
RESULTS: Prepulses significantly reduced the pain produced by the intense shocks. Individuals with the lowest pain thresholds experienced the greatest pain reduction with prepulses. In these more sensitive individuals, the most effective prepulses reduced perceived pain by 26% across the entire test session and by 54% in the initial block of five shocks.
CONCLUSIONS: Prepulses may be useful in diminishing the pain associated with the therapeutic electrical shocks used to treat cardiac arrhythmias.
Key Words: implantable cardioverter/defibrillator • prepulse inhibition • pain • startle.
Abbreviations: ANOVA = analysis of variance; ICD = implantable cardioverter/defibrillator; T = perception threshold; VAS = visual analog scale.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUMMARYACKNOWLEDGMENTSREFERENCES |
|---|
Abrupt, intense stimuli can trigger the startle reflex, and both the magnitude of the reflex and the perceived intensity of the stimulus can be reduced if the startling stimulus is preceded by a weak prepulse delivered 30 to 500 ms earlier (9–12). For example, acoustic prepulses can reduce both the magnitude of the startle reflex elicited by an intense noise or air puff and the perceived intensity of the startling stimuli (11, 12). Electrical prepulses have been reported to inhibit the startle reflex elicited by intense electric stimuli (13).
On the basis of these observations, we hypothesized that electrical prepulses could reduce the pain of a sudden, intense electrical shock. The major aim of the present study was to test this hypothesis for cutaneous shocks as a first step in evaluating the possible utility of prepulses as a method to reduce the pain of therapeutic electrical shocks.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUMMARYACKNOWLEDGMENTSREFERENCES |
|---|
Instrumentation
Prepulses and shocks were delivered through two electrodes placed 5 to 6 cm apart and taped to the skin overlaying the biceps muscle on the nondominant arm. We delivered shocks to the arm rather than the torso to prevent the possibility of initiating a cardiac arrhythmia. Although the risk of inducing ventricular arrhythmias may be minimized by synchronizing shocks to the QRS complex of the electrocardiogram, synchronization may fail. Also, QRS-synchronized shocks have nearly optimal timing to induce atrial arrhythmias. Shocks delivered to the arm create a sufficiently weak electrical field in the chest so that there is no risk of initiating arrhythmias. Electrical pulses with a duration of 0.5 ms were produced by a Biopac MP100 stimulator and a StimIsoB stimulus-isolation unit. The experimenter controlled the amplitude of the stimulus over a range of 0 to 200 V. The interval between shocks was approximately 20 seconds. To record eye-blink response as a measure of startle, two recording electrodes were taped to the skin overlaying the orbicularis oculi muscle below the left eye, with an interelectrode distance of 1.5 cm. A third (reference) electrode was taped to the left temple area. Eye-blink responses (periorbital electromyographs from orbicularis oculi) were recorded using a Coulbourn Bioamplifier with filters passing 90 to 250 Hz to a Biopac Systems MP100 interface.
Procedure
Perception threshold.
First we determined the perception threshold for a stimulus to the arm. The amplitude of the initial stimulus was 10 V with a duration of 0.5 ms. The stimulus amplitude was increased by 2-V increments until the participant reported feeling something. Then the stimulus amplitude was reduced by 2 V per trial until the participant no longer reported feeling the stimulus. Stimulus amplitude was then increased again by 2-V increments until the participant reported feeling something. This value was defined as the perception threshold (T) ( Table 1).
|
Effect of prepulses on perception of shock pain.
We then delivered a sequence of five blocks of five constant-amplitude, painful shocks for a total of 25 shocks. The shock intensity was set at a value equal to 1.5 times the pain threshold but no higher than 200 V. Thus, shock amplitude was 200 V whenever the pain threshold was >133 V.
Each trial block included a control shock without a prepulse and four test shocks with prepulses. We used two prepulse amplitudes, each delivered at two prepulse intervals (time from onset of prepulse to onset of painful shock). The two prepulse amplitudes were perceptual threshold (T) and 1.25 times perceptual threshold (1.25T). The two prepulse intervals were 40 and 60 ms. The order of stimulus conditions (control, 40 ms/1.0T, 60 ms/1.0T, 40 ms/1.25T, 60 ms/1.25T) was randomized independently for each trial block.
After each trial, the participant was asked to assess the perceived intensity of the intense shock using a 100-mm VAS that ranged from "can’t feel it" to "intolerable pain." VAS ratings were measured in millimeters from the left end of the scale. The VAS value on each trial was converted to a "relative pain score," which was the VAS rating expressed as a percentage of the total range for that participant: (Px - Pmin)/(Pmax - Pmin), where Px = the rating on this trial, Pmin = the minimum rating for all trials, and Pmax = the maximum rating for all trials. We also recorded eye-blink electromyographs for each trial.
Statistical Analyses
Relative pain scores were analyzed by ANOVA. VAS values were not recorded on 14 of 500 trials (2.8%). Almost half of these were from a single participant who terminated the study after 19 trials; the remaining 8 missing trials occurred in four other participants. Because any single missing cell resulted in the loss of the entire participant from a within-participant analysis, several strategies were taken to minimize the statistical impact of these missing cells. In the most conservative strategy, values were collapsed within a stimulus condition across the five trial blocks to yield an average VAS rating for each stimulus condition across the entire test. These data were analyzed by an ANOVA with repeated measures on stimulus condition (no prepulse, 40 ms/1.0T, 40 ms/1.25T, 60 ms/1.0T, 60 ms/1.25T). To permit assessment of the effects of repeated stimulus presentations on pain and prepulse effects, missing cells were replaced with the average value for that stimulus condition for that participant. Data were then analyzed by ANOVA with repeated measures on stimulus condition and trial block. Similar results were obtained if missing values were replaced by a value derived by extrapolation/interpolation of the previous and subsequent trials of the same stimulus condition for that participant. Finally, exploratory analyses were undertaken by dividing the participants into two groups using a median split of pain thresholds, based on the known impact of pain threshold on pain-reducing effects of anesthetic interventions (14). In these cases, pain threshold (low vs. high) was used as a grouping factor. For all analyses, 
was 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUMMARYACKNOWLEDGMENTSREFERENCES |
|---|
Based on pain thresholds, the mean amplitude of test shocks was 158.75 V (range = 72–200 V). Overall, 12 participants with pain thresholds 
128 V received test shocks that were 1.5 times their pain thresholds, and 8 participants with pain thresholds 
140 V received 200 V shocks that were <1.5 times their pain thresholds.
Startle Eye-Blink Response
Shocks presented to the arm elicited eye-blink startle on 35% of the control (no prepulse) shock trials, averaged across trial blocks for the entire session. Eye-blink response probability was 27, 24, 18, and 22% on the 40 ms/1.0T, 40 ms/1.25T, 60 ms/1.0T, and 60 ms 1.25T trials, respectively. Therefore, prepulse seemed to inhibit blink probability, but this probability was very low even without prepulses. In contrast, shocks elicited pronounced contractions of the arm musculature, resulting in occasional large, flailing movements of the arm and in some cases more generalized skeletal motor responses.
Overall Prepulse Effects on Pain
ANOVA of relative VAS ratings for all shocks revealed a significant effect of stimulus condition (F = 3.10; df = 4,19; p < .025). Post hoc comparisons revealed a significant 17.5% reduction in pain with 60 ms/1.0T trials compared with control trials (F = 18.45; df = 1,19; p < .0005) as well as lesser but significant pain reduction with 40 ms/1.25T and 60 ms/1.25T prepulses (p < .05 and p < .005, respectively) ( Figure 1, inset). Similar outcomes were obtained when missing VAS values were replaced by participants’ average values for that trial condition (17.5% reduction with 60 ms/1.0T trials; effect of trial type: F = 3.11; df = 4,76; p < .025; control vs. 60 ms/1T trials: F = 18.49; df = 1,19; p < .0005).
|
A prepulse "impact" value for each stimulus condition was calculated by subtracting the relative VAS score for that stimulus condition from the relative VAS score for the no-prepulse condition. There was a significant negative correlation between the mean prepulse impact for each participant and that participant’s pain threshold (r (20) = -0.45, p < .05). Thus, the pain-reducing impact of prepulses for any participant was inversely related to their pain threshold: Participants who were most sensitive to pain benefited most from prepulses.
Because participants were told that they could terminate the session at any time, the sample population for this study included only individuals who were relatively insensitive to pain. The nine individuals who withdrew from the study before delivery of intense test shocks (1.5 times pain threshold) had lower pain thresholds (mean = 65.8 V, range = 36–158 V) than did the remaining 20 participants (mean = 120.9 V, range = 48–200 V) (F = 8.91; df = 1,27; p < .01). Although one of these nine participants was a relatively stoic "outlier" (pain threshold = 158 V), the pain thresholds of the remaining eight participants (mean = 54.3 V, range =36–90 V) were significantly lower than those of participants in the low pain threshold group who completed testing (F = 8.01; df = 1,16; p < .025).
Effect of Trial Block
The pain-reducing effects of prepulses were examined across the five trial blocks in the 19 participants who completed the test session ( Figure 2). Pain ratings for all stimulus conditions increased substantially across the session, with mean (SEM) relative VAS ratings on no-prepulse trials increasing from 0.45 (0.06) (block 1) to 0.83 (0.06) (block 5). In the low pain threshold group, these values approached the maximal level of 1.0 by block 5 (mean = 0.95, SEM = 0.03), indicating a "ceiling effect" that might diminish the sensitivity of detecting prepulse effects on pain perception. ANOVA with trial block as a within-participant factor and pain threshold as a between-groups factor revealed a significant effect of trial block (F = 15.33; df = 4,68; p < .0001) and stimulus condition (F = 2.73, df = 4,68; p < .05) and significant interactions of stimulus condition by pain threshold (F = 3.34; df = 4,68; p < .025) and stimulus condition by pain threshold by trial block (F = 1.88; df = 16,272; p < .025).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUMMARYACKNOWLEDGMENTSREFERENCES |
|---|
Mechanism of Pain Reduction by Prepulses
A reduction in pain on prepulse trials relative to control trials could be due to one or more of several processes, including prepulse inhibition, sensory assimilation, and refractoriness. Prepulse inhibition is the normal inhibition of a startle reflex that occurs when the startling stimulus is preceded by a weak prepulse (15). Prepulse inhibition stimulated by acoustic prepulses seems to be mediated by the effects of the acoustic prepulse on brain circuitry within the pontine tegmentum (16, 17), resulting in suppression of activity within the "primary" startle circuit at the level of the caudal pons (18, 19). If this same circuitry mediates the startle-inhibiting effects of electrical prepulses, the pontine circuitry activated by the prepulse may also project to other brainstem pain pathways, inhibiting the perception of painful stimuli.
A less likely mechanism by which a prepulse might decrease the pain of a shock to the arm might involve the process of perceptual averaging, referred to in the auditory system as "loudness assimilation," wherein the perception of an intense stimulus can be modulated by a weaker stimulus presented just before the intense stimulus (20). Loudness assimilation involves a shift in the perceptual attributes of both stimuli toward each other, so that the weak prepulse sounds louder and the intense pulse sounds weaker. If an analogous process regulates the perception of electrical stimuli in the present situation, the perception of the electrical prepulse would lead to the subsequent intense stimulus being perceived as less intense and therefore less painful. Such a process could be relevant to the present findings only if perceptual averaging does not require a conscious awareness of the prepulse stimulus, because the prepulse intervals in this study (40–60 ms) do not typically permit the detection of two distinct stimuli. With acoustic stimuli these prepulse intervals are too short to detect the effects of attentional allocation on prepulse inhibition (21), suggesting that separate attentional mechanisms are not engaged by lead stimuli with this stimulus configuration.
Refractoriness of pain sensation caused by the prepulse is an unlikely explanation for the present findings, because prepulse amplitude was well below pain threshold.
Ancillary Findings
We note two additional results of this study. The first is that pain ratings increased temporally across the session for all stimulus conditions. This sensitization to painful stimuli was also found by Duker et al. (22) and may be due to depletion of endorphins with repeated presentation of painful stimuli. The fact that participants attended to the painful stimulus (to provide VAS scores) also probably contributed to the increased painfulness of those stimuli across trials (23).
A second ancillary finding is the low probability of eye-blink responses to the intense shocks presented to the arm, although prepulses seemed to inhibit response probability slightly. If shocks of considerably lower intensity had been presented to the forehead, eye blinks would have occurred on most trials (11). Given the fact that ascending sensory projections arising in the arm eventually pass through the pons, these data suggest that most, but not all, of these projections bypass the pontine startle center or the nucleus of the facial nerve, which drives the orbicularis oculi and causes the eye-blink response (24, 25). Sarno et al. (26) have suggested that the electrically elicited blink response may not be an example of a startle response, and the present study shows that the location of the electrical stimulus is one factor that determines the degree to which an eye-blink response will be activated.
Potential Clinical Application
Prepulses may reduce the pain of intense therapeutic electrical stimuli such as those used to treat cardiac arrhythmias in conscious patients. Although ICDs have become first-line therapy for some patients with life-threatening ventricular tachycardia or fibrillation (1–3), shock-induced pain is an important problem and may cause psychiatric morbidity (4–8). Recently ICDs have been used to treat atrial fibrillation (27). Shock-induced pain severely limits patient acceptance of ICD treatment of atrial fibrillation, which is not life-threatening (27, 28). Furthermore, startle is an important component of shock-related psychiatric morbidity (7). In the present study, we did not strongly confirm inhibition of eye-blink startle responses by weak electrical prepulses because of the relative insensitivity of the eye-blink measures to the startling effects of arm shocks. However, a substantial amount of literature supports the notion that a startle reflex elicited by electrical, acoustic, or tactile stimuli can be reliably suppressed by appropriate prestimuli (11, 15, 19, 29). Thus, prepulses delivered before ICD shocks may reduce both pain and startle.
The pain of electrical pulses also limits patient acceptance of transcutaneous (30, 31) and esophageal (32) cardiac pacing. An important difference between ICD shocks and cardiac pacing is that ICD shocks usually are single, but pacing consists of a series of stimuli. The present findings suggest that optimal prepulse parameters for reducing pain from a single shock may not be optimal for reducing pain elicited by a series of electrical pulses. Thus, the optimal prepulse interval for pain reduction may differ for ICD shocks and transcutaneous or esophageal pacing.
There are important differences between the present study and actual use of therapeutic electrical shocks in patients: 1) Subjects in this study were healthy college students; most ICD recipients are older and have serious cardiac disease. 2) Shocks in this study were delivered to the skin of the arm through small electrodes; ICD shocks are delivered through intrathoracic electrodes with 25 to 50 times more surface area. 3) Subjects in this study received more shocks than ICD recipients receive. 4) The characteristics of the shock waveforms differ. 5) Before receiving the initial test shock, subjects received multiple high-intensity shocks to measure pain threshold. This contrasts with the usual clinical situation, in which patients receive single ICD shocks. Given these differences, the possible utility of prepulses in reducing pain associated with therapeutic electrical pulses should, in future studies, be tested under conditions that more accurately model clinical use of these therapies.
Limitations
Participants in this study were self-selected for relatively high pain thresholds. Because the pain-reducing impact of prepulses was inversely related to pain threshold, the present findings may underestimate the pain-reducing effect of prepulses in an unselected population. Also, the present study was not designed to assess the wide range of stimulus parameters that might influence prepulse inhibition of perceived pain. Based on the fact that prepulse inhibition of startle is highly sensitive to variations in stimulus conditions (15), prepulse effects on pain perception may be similarly sensitive to stimulus conditions. Careful parametric studies are needed to identify the experimental conditions that will yield maximal pain reduction by prepulses.
|
SUMMARY |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUMMARYACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUMMARYACKNOWLEDGMENTSREFERENCES |
|---|
Received for publication March 10, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUMMARYACKNOWLEDGMENTSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
