Pain, 41 (1990) 109-114 Elsevier
The effects of distraction on responses to cold pressor pain Robert
Eric W. Howland,
Pain Research Group, Dept. of Neurology, 648 WARE 610 Walnut St., Madison, WI 53705 (U.S.A.) (Received
1979, revision received
Subjective pain ratings and tolerance time were obtained during 2 cold pressor immersions for 3 groups of subjects. During the second immersion 1 group performed no task and the other 2 groups performed either an easy or difficult mental arithmetic task. The sensory-discriminative response to pain was measured by pain ratings. Pain ratings were collected every minute until subjects removed their arm from the cold pressor or until 4 min passed. Relative to a baseline cold pressor immersion, subjects in both the distraction conditions reduced their 1 min pain ratings more than control subjects. This effect was weaker at the 2 min pain rating and absent at the later ratings. The affective-reactive response to pain was measured by pain tolerance times. Tolerance time was defined as the time when subjects removed their arm from the cold pressor. Tolerance time was not altered by the distraction tasks. These findings suggest that affectively neutral distraction alters the sensory but not the reactive response to pain. Clinical implications are discussed.
Introduction Folk wisdom and modem behavioral psychology both appreciate the power of distraction in reducing the distress associated with painful stimulation. McCaul and Malott  recently reviewed laboratory studies on distraction and concluded that distraction procedures are superior to appropriate control interventions in producing analgesia. In their model, the analgesic property of distraction is attributed to the relative potency of the distractor as opposed to nociception in captur-
’ This research was supported by Grant No. NS22677 vided by the National Institutes of Health.
Correspondence to: C.S. Cleeland, Pain Research Group, Dept. of Neurology, 648 WARF, 610 Walnut St., Madison, WI 53705, U.S.A.
0 1990 Elsevier Science Publishers
ing the limited capacity of the brain’s perceptual processing mechanisms. This model can be tested by comparing the analgesic properties of complex distracters with the analgesia produced by control procedures. Barber and Cooper [l] used a pressure pain procedure and found that pain ratings were lower during story listening and serial addition by 7s as compared to the simpler task of counting by 1s. Horan and Dellinger  compared the analgesic effects of counting backwards by 1 (presumably a simple cognitive task) with relaxing imagery on cold pressor pain. As expected, imagery subjects tolerated cold pressor stimulation for twice as long as did subjects in the counting condition. These data support the potency of distracters in inhibiting pain response. However, 2 problems prevent a full endorsement of this conclusion. First, the above studies are flawed by their selec-
tion of distraction conditions that cannot be directly observed by the experimenters (e.g., pleasant imagery). This makes it difficult to confirm if the subjects were engaging in the experimental manipulation during the entire period of painful stimulation. Of greater significance, in the studies cited the affective valence of the distractor (i.e., relaxing imagery, story telling) was confounded with the attentional demands of the distractor. Does the potency of the distraction procedure come from its ability to capture the attention of the subject or from its facility in promoting a positive mood state? Some data support the latter interpretation. For example, Horan et al.  found that relaxing imagery led to lower discomfort ratings during a tooth stimulation procedure than did a neutral imagery task. Similarly, Stone et al. [ll] found that pleasant imagery increased cold pressor tolerance compared to a neutral imagery procedure. The distinction between the attentional demands and the affective impact of a distractor dovetails with contemporary multidimensional models of pain perception. Although various multidimensional models exist [2.8,10], they all recognize that pain stimuli signal more than just sensory information. Pain processing also involves cognitive evaluations of the painful stimulus as well as the indivual’s affective response. We prefer the term reactive to describe this second dimension of pain processing. The distinction between sensory and reactive pain processing helps define the boundary conditions for the efficacy of any distraction technique. Affectively neutral distracters. such as mental arithmetic, are assumed to work by providing sensory input that competes with nociception for access to a limited capacity sensory processing mechanism . In contrast, mood and cognitive manipulations modify pain by their influence on the emotional and evaluative processing of the painful stimulation. Affectively potent distracters, such as pleasant imagery, should attenuate responding in both neural pathways. In the present study. high and low difficulty distraction manipulations were compared with a no distraction condition on their effect on cold pressor induced pain. In both tasks, subjects were
required to make a behavioral response if a string of numbers met a predetermined criterion (e.g., an odd number followed by an even number). Errors of omission and commission could be monitored to determine if subjects continued to engage in the task. In addition, the tasks were chosen to be affectively neutral (both involved arithmetic tasks) but to vary on a dimension of complexity as determined by the number of mental operations required. Cold pressor stimulation was chosen because it allows for 2 response measures, i.e., ratings of pain intensity at fixed time intervals and the subject’s tolerance time. Tolerance time is known to be sensitive to both motivational  and affective variables . This response measure can be thought of as representing processing in the reactive pain system. Ratings of pain intensity are felt to measure the sensory dimension of pain. We predicted, therefore. that affectively neutral distraction techniques like mental arithmetic would reduce pain ratings but not pain tolerance to cold pressor stimulation. In contrast, the model proposed by McCaul and Malott  predicts that both pain tolerance and pain ratings would be diminished during distraction. Both models hold that the lowest pain ratings will be made by the subjects in the high difficulty distraction condition.
Forty-five volunteers (16 males, 29 females) were recruited from the local and university community with posters announcing a study on the effects of cold on the perception of pain. Subjects were paid at a rate of $10.00/h. Subjects ranged in age from 18 to 35. Subjects with peripheral vascular disease (e.g., Raynaud’s syndrome) or high blood pressure, subjects who regularly used tobacco, and subjects taking medications with analgesic or vasoactive properties were excluded from the study. Subjects who were currently receiving psychiatric treatment or who scored high on the subscales of the SCL-90 were also excluded from the study.
Apparatus Subjects were seated in a comfortable recliner in a room that was adjacent to the experimenter room. Instructions were provided verbally by the experimenter using standard scripts. Cold pressor stimulation was provided by an apparatus adapted from Turk et al. . Briefly, a 25.25 gallon, thermally insulated cooler was modified into 2 chambers separated by a nylon mesh screen. The smaller chamber was filled with crushed ice and then the entire cooler was filled 3/4 full with water. An electrical fountain pump (Calvert Engineering Pump Model 875SF) circulated the water through the ice-filled chamber until the water reached a temperature between 0 and 1 o C. Water temperature was measured using an American Scientific Products laboratory mercury thermometer. Water cooling was done just prior to the start of the experiment. Because of the large mass of water cooled and the thermal insulation, subsequent circulation of water through the ice chamber was not necessary. The duration of arm immersion in the cold water was timed manually with the stopwatch. The distraction task was presented to subjects via a tape recording.
Procedure All subjects gave signed informed consent. Subjects received 2 trials of cold pressor stimulation. Each trial went until subjects withdrew their arm from the cold water or for a maximum of 240 sec. During each cold pressor trial, subjects rested their non-dominant arm in a cradle support and then immersed their arm to the elbow into the cold water. Subjects were asked to leave their arm in the water for as long as they could tolerate it (‘until you feel forced to take the hand out’) or until the end of the trial. They were instructed to spread their fingers and to avoid moving their hand during the cold pressor trial. Subjects were asked to make pain ratings at seconds 60, 120, 180, and 240 and at the termination of each trial. Ratings were made verbally using a O-10 numeric scale, with 0 representing ‘no pain’ and 10 representing ‘pain as bad as you can imagine.’ Brief pauses occurred in the distrac-
tion tape to allow subjects time to make these ratings. For all subjects, the experiment began with immersion of their non-dominant hand and forearm for 3 min into a tepid water bath (3OOC). This was done to moderate initial differences in arm temperature. Following this, all subjects received a baseline cold pressor as described above. After the completion of this cold pressor, they were again given the opportunity for a tepid water immersion until their arm again felt normal. Subjects were next read the instructions for their experimental condition. Previous pilot work led to the development of 2 arithmetic tasks that were known to vary on their degree of difficulty (as measured by error) and attentional demand (as measured by subjective ratings of involvement by subjects). Subjects in the high difficulty distraction group listened to a series of l-digit integers spoken at a rate of 1 integer each 2.4 sec. Subjects were required to determine if the sum of the last 2 numbers presented was greater than 12 or less than 6. If this condition was met, they were asked to tap on the table with a pen. If this condition was not met, no response was required. For example, if they heard the sequence of numbers 1, 4, 7, they would tap after 4 (since 1 + 4 = 5) but not tap after 7 (since 4 + 7 = 11). Subjects were given practice on this task with 5 written integer examples and then 13 spoken integers. Subjects in the low difficulty distraction group listened to the same series of integers but were asked to tap with their pen when the last 2 numbers heard fit the pattern ‘odd followed by even.’ For example, for the sequence 1, 4, 7, they wouid tap after 4, but they would not tap after 7. No response was required for any other sequence of integers. Again, practice was provided. The sequence of integers on the audiotape was constructed to equate the high and low difficulty task on the number of hits (i.e., trials on which subjects should have tapped with their pen). Subjects in the no distraction task did not listen to taped integers. All subjects then received a second cold pressor stimulation. Subjects in the appropriate conditions performed their distraction task simultaneous with
this second cold pressor stimulation. Following this, subjects received a final tepid water immersion and were debriefed and paid for their participation in the study.
TWO measures of pain perception were available, i.e., tolerance (time to withdrawal of arm from the cold pressor) and numeric pain ratings. These data were analyzed with ANOVAs with 2 orthogonal comparisons (distraction groups vs. no distraction; high distraction vs. low distraction). The 3 groups did not differ on their initial cold pressor mean tolerance times (F (2, 42) = 0.42, n.s.). Cold pressor 2 - cold pressor 1 difference scores were calculated for each subject and these scores were analyzed with planned comparisons. Twenty-four of the subjects in this experiment maintained their arm in the initial cold pressor for the full 4 min. It was not possible for these subjects to increase their tolerance times during the distraction task. To eliminate this ceiling effect from the data, the analysis of tolerance difference scores was calculated for the subset of subjects who failed to reach tolerance on the first cold pressor. Using only these subjects, the increase in tolerance under distraction conditions was not statistically significant (t = 1.60, P < 0.14). The difference between subjects receiving high and low distraction was also not significant (t = 0.15, n.s.). A power analysis indicated that this effect size would not have reached statistical significance even if no subjects were lost by reaching tolerance on the first cold pressor. Pain rating change scores (cold pressor 2 - cold pressor 1) were calculated with separate planned comparisons at each rating interval (see Fig. 1). Two analyses were done. The first analysis involved missing data created as subjects terminated the cold pressor trial when pain tolerance was reached. This resulted in fewer subjects contributing to the analyses at later rating intervals (see Table I). Distraction groups differed from the control condition only at the first rating interval (t (35) = 2.47, P < 0.02) with a tendency toward a
% 5 -0.6 5
0 No Distraction
Low Distraction W High Distraction
Fig. 1. Changes in pain ratings for each minute of the cold cluded in the later ratings as their
(cold pressor 2 -cold pressor 1) pressor. Fewer subjects are insome subjects have withdrawn arms.
significant difference at the second rating interval (t (26) = 1.76, P < 0.09). The high and low difficulty distraction conditions produced comparable results at all rating intervals. In a second analysis, missing data for later time periods (i.e., after subjects withdrew their arms from the cold pressor) were estimated by the subject’s last pain rating at the time that they withdrew TABLE
MEAN PAIN RATINGS AT EACH RATING INTERVAL FOR SUBJECTS IN THE 3 DISTRACTION GROUPS Number of subjects making rating for the first immersion.
Rating min 1 Immersion 1 Immersion 2
5.63 (11) 5.00
5.33 (12) 5.23
5.71 (14) 6.07
Rating min 2 Immersion 1 Immersion 2
5.12 (8) 5.27
5.44 (9) 5.18
6.36 (11) 6.50
Rating min 3 Immersion 1 Immersion 2
5.15 (8) 6.00
6.12 (8) 5.64
5.71 (7) 5.44
Rating min 4 Immersion 1 Immersion 2
6.25 (8) 6.40
6.15 (8) 5.80
5.86 (7) 6.00
their arm. Difference scores between cold pressor 2 and cold pressor 1 were analyzed at each 60 set rating period with planned comparisons. As before, the reduction in pain ratings during the second cold pressor was greater for the distraction groups than for the no distraction group but only at rating interval 1 (t = 2.60, P < 0.01). As before, the 2 distraction groups did not differ from each other at any of the rating intervals. Performance data The sum of the errors of omission (failing to tap) and commission (tapping after an incorrect integer) were determined for each 60 set time block during the second cold pressor. Twenty subjects (high difficulty = 9; low difficulty = 11) contributed scores during each of the 4 time periods. These data were analyzed with a mixed model ANOVA with Difficulty (high, low) as a betweensubject variable and Period (l-4) as the withinsubject variable. The data are shown in Table I. As expected, subjects in the high difficulty condition made many more errors than did subjects in the low difficulty condition (Greenhouse-Geisser correction: F (1, 18) = 25.00, P < 0.0001). As can be seen in Table I, errors either decreased or stabilized across time (Greenhouse-Geisser correction: F (3, 54) = 5.50, P < 0.005) supporting the conclusion that subjects were involved in the distraction task during the entire task interval. Subjective ratings of involvement Subsequent to the second cold pressor, subjects in the distraction conditions were asked to make subjective ratings concerning their level of absorption during the distraction task. The higher rating of involvement by subjects in the high difficulty
MEAN ERRORS WITH STANDARD EACH MINUTE OF COLD PRESSOR
Low dist. High dist.
1.45 (1.91) 9.44 (4.27)
0.09 (0.30) 6.00 (5.57)
0.63 (1.02) 7.11 (5.21)
1.36 (1.63) 6.00 (3.04)
distraction condition approached but did not reach conventional levels of statistical significance (F (1, 28) = 3.67, P = 0.06).
Discussion Simultaneous performance of a mental arithmetic task significantly reduced pain ratings but not pain tolerance to cold pressor stimulation. The reduction in pain ratings was greater than the change that was seen in the no distraction or habituation control group. These data contrast with previous studies where both pain ratings and pain tolerance were affected by distraction. These earlier studies employed distracters, such as relaxing imagery, that both competed with the sensory processing of the painful stimulus and that altered the subject’s affective state. In the present study, the distractor was chosen to be affectively neutral. As predicted, affectively neutral distraction only affected pain ratings and by implication, only altered processing in sensory pain systems. It is notable that distraction did not alter pain ratings for minutes 2-4 of the cold pressor. Two explanations arise. First, this pattern is consistent with models that argue that distraction is effective for only low intensity  or for brief  pain stimulation. Alternatively, the latter minutes of the cold pressor are over-represented by cold tolerant subjects. Distraction may have differential impact on pain sensitive and pain tolerant individuals. Future studies need to sort out these possibilities. Contrary to our hypothesis, the high and low difficulty distraction conditions did not differ on either pain ratings or tolerance time. It was assumed that error rates provided indirect information about the attentional demands of a task. As expected, the high difficulty condition was associated with a greater number of performance errors throughout the duration of the task. However, differences in error rates may not accurately reflect differences in the allocation of attentional resources. As an extreme example, subjects who give up when performing a second task will have high error rates but will not be allocating attention to that task. Other methods for measuring atten-
tional allocation, such as dual task paradigms, may help clarify these findings. Taken as a whole, the data do provide consistent support for the hypotheses that (1) distraction has varying impact on different aspects of pain responding, and (2) affectively neutral distraction during tonic pain stimulation reduces sensory pain ratings but not pain tolerance. These results may have useful implications for the treatment of clinical pain. Specifically, the data suggest that affectively neutral distraction will be most beneficial when used to manage a patient’s reaction to relatively brief painful stimulation. Examples would include episodes of pain associated with procedures such as lumbar puncture or venipuncture. In contrast, affectively neutral distraction will be relatively ineffective for making chronic pain more tolerable. For persistent pain, the choice of any distraction technique should not be guided by the distractor’s attentional demands but by the therapist’s attempt to either modify the patient’s affective state, e.g.. the use of relaxing imagery to promote a positive mood state, or by an attempt to modify the patients’ appraisal or evaluation of their chronic pain. As an example of the latter, distraction training could include a component designed to improve the patients’ selfefficacy over their pain.
References 1 Barber, T.X. and Cooper, B.J., Effects on pain of experimentally induced and spontaneous distraction, Psychol. Rep., 31 (1972) 6477651. 2 Beecher, H.K., The Measurement of SubJective Responses. Oxford University Press, New York. 1959. 3 Blitz, B. and Dinnerstein, A.J.. Effects of different types of instructions on pain parameters, J. Abnorm. Psychol.. 73 (1968) 2766280. 4 Chen, A.C.N., Dworkin, SF.. Haug, J. and Gehrig, C; , Human pain responsivity in a tonic pain model: psychological determinants, Pain. 37 (1989) 143-160. 5 Horan, J.J. and Dellinger, J.K.. ‘In viva’ emotive imagery: a preliminary test. Percept. Motor Skills, 39 (1974) 3599362. 6 Horan, J.J.. Layng, F.C. and Pursell, C.H.. Preliminary study of effects of ‘in viva’ emotive imagery on dental discomfort, Percept. Motor Skills, 42 (1976) 105- 106. 7 Kahneman. D., Attention and Effort, Prentice Hall, Englewood Cliffs, NJ, 1973. 8 Leventhal, H. and Everhart, D.. Emotion, pain, and physical illness. In: C.E. lzard (Ed.), Emotions in Personality and Psychopathology. Plenum Press, New York. 1979. pp. 263-299. 9 McCaul, K.D. and Malott, J.M.. Distraction and coping with pain. Psychol. Bull., 95 (1984) 516-533. and 10 Melzack, R. and Casey, K.L.. Sensory, motivational, central determinants of pain. In: D.R. Kenshalo (Ed.), The Skin Senses, Thomas, Springfield, IL, 1968, pp. 4233443. II Stone. C.I., Demchik-Stone, D.A. and Horan, J.J., Coping with pain: a component analysis of Lamaze and cognitivebehavioral procedures, J. Psychosom. Res.. 21 (1977) 452 456. D. and Genest, M., Pain and 12 Turk. D.C., Meichenbaum, Behavioral Medicine: a Cognitive-Behavioral Perspective. Guilford Press. New York, 1983.