The Psychological Record, 2003, 53, 243-252

EFFECTS OF COLD PRESSOR PAIN ON HUMAN SELF-CONTROL FOR POSITIVE REINFORCEMENT STEPHEN RAY FLORA and LINDY RAE WILKERSON Youngstown State University

DAVID B. FLORA University of North Carolina

Humans pressed 2 buttons for points that were exchangeable for money. Pressing 1 of the buttons, the impulsive choice, produced 2 points over 4 s. Pressing the other button, the self-controlled choice, produced 10 points over 4 s after a 16-s delay. Each point was exchangeable for 1 cent ($0.01). All subjects earned 1 cent ($0.01) for every 3 s they kept their right hand immersed in water. All subjects made 30 choices. In the experimental, pain, condition the water contained ice and was maintained at approximately 5 "C. In the control condition the water temperature averaged 33 "C. Subjects in the pain condition made fewer self-controlled responses, rated the water as more painful and more distracting, and kept their hands in the water for shorter durations than subjects in the control condition. The results have evolutionary and societal implications and suggest that variables affecting subjects' physiological state, or least their subjective physiological state, need to be taken into account for a complete understanding of human "self-controL" Impulsiveness may be an evolutionary adaptation to aversive stimulation.

Self-control is defined as the choice for a larger later reinforcer (LL) over the choice for a smaller sooner reinforcer (SS). Impulsiveness is defined as the choice for a SS reinforcer over a LL reinforcer. In experimental investigations of self-control and impulsiveness, adult humans typically respond with more self-control than impulsiveness (more LL choices than SS choices) (e.g., Flora, 1995; Logue, King, Chavarro, & Volpe, 1990). Exceptions occur with negative reinforcement: Even when delayed escape results in longer overall relief from aversive noise, humans usually chose immediate brief escape from aversive noise (SS) over longer but delayed escape (Navarick, 1982; Sol nick, Kannenberg, Eckerman, & Waller, 1980). Likewise, in an aversive context (loud noise), when adult humans make choices for points exchangeable for money-positive reinforcement-self-control is attenuated. This Address correspondence to Stephen Ray Flora, Department of Psychology, Youngstown State University, Youngstown OH, 44555. (E-mail: [email protected]). David B. Flora is now associated with Arizona State University.

244

FLORA ETAL.

increased impulsiveness occurs even though the resulting increase in impulsive behavior in no way functions to lessen the aversiveness of the context or remove the person from the context (e.g., Flora, Schieferecke, & Bremenkamp, 1992). Rats deprived of food for 22 hr (likely an "aversive physical state") respond more impulsively than rats deprived of food for only 2 to 4 hr (Eisneberger, Masterson, & Lowman 1982; but see Hastjarjo & Silberberg, 1992). Caffeine increases impulsive responding for food reinforcement in caffeine-naive rats (Flora & Dietze, 1993). The increased impulsiveness may have been caused in part by a possible aversive internal state produced by caffeine consumption in caffeine-naive rats. Taken together, these results suggest that both aversive environmental stimulation (loud noise) and aversive internal states (e.g., hunger or caffeine consumption in caffeine-naive organisms) increase impulsive behavior. If this is true, then physical pain should also result in impulsive behavior even when the impulsive behavior does nothing to escape or reduce the pain experienced. The present research was conducted to determine if the experience of physical pain (cold pressor) decreases self-control for positive reinforcement. Method Subjects Subjects were 34 general psychology students at Youngstown State University, 18-25 years old. Apparatus The apparatus used was a Commodore 128 computer interfaced with two push buttons protruding from a small platform positioned directly in front of the computer monitor. The front half of the platform was green and the rear half was blue. Each half had a center button. Each subject wore an automated blood pressure and pulse cuff on her or his right upper arm. A container of water was placed in a open desk drawer on the subject's right side. The computer monitor and platform sat on the desk directly in front of the subject. The keyboard was to the right of the monitor and not used by the subject. Procedure and Design Eighteen subjects (2 male, 16 female) served in the cold, pain experimental condition and 16 subjects (4 male, 12 female) served in the control, warm condition. Subjects were randomly assigned to conditions. The experiment was conducted in a well-lit 3-m x 4-m room, which was bare except for a desk, two chairs, a computer and automated blood pressure monitor on the desk. Responses were defined as closure of a microswitch, produced by the appropriate button press. Upon arriving in the experimental room, all subjects were allowed to read and sign an informed consent form and assigned to conditions. The subjects were then read the following instructions:

PAIN AND HUMAN SELF-CONTROL

245

I will be taking your blood pressure and pulse twice during the experiment, once now and once immediately after the computer program is complete. We will keep the blood pressure cuff on your upper right arm during the experiment. May I put it on now? When the program starts I will ask you to put your right hand, up to your wrist, in this water, I will keep track of how long you keep your hand in the water with a stop watch. You will be paid one cent for every three seconds you keep your hand in the water. You can earn up to two dollars for keeping your hand in the water. If it gets too uncomfortable you can take it out, but please put it back in when you can. Try to keep it in the water, that's how you make the most money. You also earn one cent for every point you score on the computer. I will be sitting behind you during the experiment, and I cannot answer any questions once the session starts. All of the instructions you will need will be given by the computer. If the computer says ''wait'' or "wait, computer working," then please remain in your seat and wait. Are you ready?

Once a subject indicated that he or she was ready, the experimenter pressed a key on the computer to start the experiment and sat behind the subject with a stop watch. The screen turned red for 2 s and "this experiment lasts 600 seconds" was printed on the screen. Next the screen turned yellow and the following instructions were printed on the screen: You earn one cent ($0.01) for every 3 seconds your entire hand is kept in the water. Now your task is to earn as many points as possible. You will given one cent ($0.01) for every point you earn. You can earn up to $5.12 total. Your actual earnings is up to you. You can earn $2.00 for keeping your hand in the water, and $3.12 total computer points are possible. You will be paid in cash after the experiment is over. If the computer reads 'wait' or if the computer reads 'wait computer working,' then please sit and wait, you may have to wait for a few minutes. When the computer reads 'push the green button,' push the button on the green half of the platform. When the computer reads 'push the blue button,' push the button on the blue half of the platform. When the computer reads 'choose a button and push it,' you should push the button of your choice. Please push a button when the computer asks you to do so. Now, tell the experimenter when you are ready to start.

When the subject told the experimenter that he or she was ready the experimenter said "put your right hand in the water up to the wrist" and pressed a key to start the experiment. The screen then immediately cleared, turned red and "This experiment lasts 600 seconds" was again printed on the screen for 2 s. Then the screen turned yellow and, and "Press the green button" was printed on the center of the screen prompting an impulsive forced choice. Directly below the choice prompt, "Your points=" followed by the points earned was printed on the screen and remained on the screen throughout the remainder of the experiment.

246

FLORA ET AL.

Once the green button was pressed, the screen immediately turned green, 1 point was added to the score and 2 slater 1 point again was added to the score, and the screen remained green for 2 more s. Thus, the impulsive choice produced an immediate reinforcement period in which 2 pOints were delivered over 4 s (1/2 point per s). After the forced choice, impulsive reinforcement period, the screen again turned yellow and "press the blue button" was printed on the screen. Once the blue button was pressed, the screen turned gray and remained gray for 16 s with "wait" printed above the score. After 16 s, the screen turned blue, 5 points were added to the score, and 2 slater 5 points were again added to the score. The screen remained blue for 2 more s. Thus, the self-control choice produced a delay of 16 s followed by 10 pOints delivered over 4 s. This combination produced 10 pOints delivered over 20 s, or a reinforcement ratio of 112 point per s, per trial. This ratio was equal to the impulsive choice reinforcement ratio of pOints per s, per trial. After the forced choice self-control reinforcement period, the screen again turned yellow and "chose a button and push if' was printed on the center of the screen prompting the first choice trial. Then, for 30 trials once a button was pressed the same programmed contingencies occurred as in the forced choice trials. Except for the periods following Trials 15 and 30 (see below), after the appropriate reinforcement period the screen always turned yellow and prompted a choice (See Figure 1). There was a wait period after Trial 15 that was increased by 16 s for each impulsive choice that was made during Trials 1 to 15. If no impulsive choice were made there was no wait period. An identical contingent wait period occurred after Trial 30 for any impulsive choices that were made during Trials 16 to 30. During the wait periods the screen turned gray and "wait, computer working" was printed on the screen. The wait periods were used to hold the length of the experimental sessions constant regardless of how many impulsive choices occurred. (See Flora, 1995, and Flora & Pavlik, 1992, for theoretical and empirical rational for wait periods). Although the wait periods prevented impulsive responding from redUCing the length of the experiment, the actual length of the experiment varied according to how long the individual subjects took to press a button once a choice prompt occurred (a value that would be somewhat over 600 s, or 10 min). The only difference between the experimental, pain condition and the control condition was the temperature of the water in which the subjects were asked to keep their hands immersed; 5° C with ice for the experimental, pain condition, or 33° C for control condition. Each subject's blood pressure and pulse were recorded before the subjects put their hand in the water and immediately after they completed the computer program. During debriefing, subjects were asked to rate on a scale of 1 to 10 how painful and how distracting the water was during the experiment. A multivariate analysis of variance (MAN OVA) was conducted on the resulting data with the experimental condition as the independent variable

247

PAIN AND HUMAN SELF-CONTROL

20 5

SELF-CONTROL

5 S 15 E C

0 N 10 D·

S 5

- -I

IMPULSIVE I

0

I I

I

~)

Button Choice

- _.

Figure 1. Flow chart of delay and amount of reinforcement for an impulsive or self-control choice on an individual choice trial. Time proceeds from the bottom of the figure. Dashed lines indicate an immediate return to choice point. Number of points delivered is indicated in the boxes. For each implusive choice made, 16 s were added to the wait periods following Trials 15 or 30 (see text).

(warm or cold water) and the behavioral variables (number of self-control choices and number of seconds in water) and psychological variables (pain ratings and distraction ratings) as the dependent variables. Blood pressure and pulse rate measures were evaluated separately. Results The multivariate analysis of variance omnibus null hypothesis for the collection of dependent variables (number of self-control choices, seconds in water, pain rating, and distraction rating) was rejected, F(4, 29) = 22.45, P < .001, Wilks' Lambda = .244. Approximately 76% (i.e., 1

248

FLORAETAL.

- Wilks' Lambda) of the variability in the set of outcomes is explained by the experimental manipulation (pain condition vs. control condition). Univariate analyses were conducted to evaluate the effects of the experimental manipulation in each dependent variable separately. To control for accumulated Type I error probabilities due to multiple comparisons, these univariate tests were evaluated based on a Bonferroni-corrected alpha level of .0125 rather than the conventional alpha level of .05. Using this Bonferroni criterion, the difference in the mean number of self-control choices made by subjects in the pain condition versus subjects on the control condition was only marginally significant, F(1, 32) = 6.74, P = .014. Subjects in the pain condition made fewer self-control choices (10.5) than subjects on the control condition (16.31). Inspection of side-by-side box plots (Figure 2) of the distribution of the number of self-control choices as a function of the experimental manipulation revealed an outlier in the pain condition (this subject made

0

30

0 en (]) u

'0 20 .I:: U

ei::

I

0

U ..L Q5

10

(f) "-

0

~

Q)

.0

E

:::J

Z

0

I

Warm

Cold

0

Figure 2. Side-by-side box plots of number of self-control choices by pain manipulation. In a box plot the ends of the box are the quartiles (i.e., the 25th and 75th percentiles), so that the length of the box is the interquartile range (lOR). lOR equals the difference between the 75th and the 25th percentile. The line within the box marks the median. The two lines ("whiskers") outside the box exetend to the smallest and largest observations, but only if these values are less than 1.5 x lOR beyond the quartiles. Otherwise, the whiskers extend to the most ex1reme observations still within 1.5 x lOR of the quartiles and the remaining cases (i.e., outliers) are plotted individually as circles (Moore & McCabe, 1993, pp. 42-43).

I

..I

249

PAIN AND HUMAN SELF-CONTROL

800~------------------------------------------'

o T 600

400

200 til

"0 C

0

(J Q)

en

0 Warm

Cold

Figure 3. Side-by-side box plots of the number of seconds subjects kept their hands immersed in water in the control, warm water, condition and the experimental, cold pressor pain, condition. 10

8

]'

iO

7

OPain

6

S Distractiveness

II:

S

:i

4

..

c

3 2

Cold

Warm

Experimental Condition Figure 4. Mean subjective ratings of pain (open bars) and distraction (hatched bars) by subjects in the experimental, cold pressor pain, condition (left set of bars) and subjects in the control, warm water, condition (right set of bars).

250

FLORA ET AL.

30 self-control choices). When this outlying observation was removed from the data, the difference in the mean number of self-control choices made by subjects in the pain condition versus subjects in the control condition became significant, F(1, 31) = 12.96, P = .001. Although highly variable (see Figure 3), subjects in the pain condition kept their hands in the water for shorter durations than subjects in the control condition, F(1,32) = 12.16, P = .001. Subjects in the pain (cold) condition rated the water as more painful than subjects in the control condition, F(1, 32) = 71.48, P < .001 (Figure 4). Subjects in the pain condition rated the water as more distracting than subjects in the control condition, F(1, 32) = 11.71, P = .002 (Figure 4). The pulse rate of subjects in the pain condition significantly decreased (77.7 to 70.4 beats per min), t(15) = 3.24, P = .006. Otherwise there were no changes in blood pressure or pulse rate for subjects in either condition. Discussion

Pain. Subjects rated the cold pressor as more painful-a subjective indication of pain-and subjects in the pain condition kept their hands in the water for shorter durations-an objective, behavioral indication of pain. However, there were no differences in blood pressure-a gross phYSiological indication of pain. These results raise the debate as to whether pain is a physiological phenomenon, or a psychological, operant phenomenon (e.g., Rachlin, 1985). The cold pressor is a standard experimental pain induction procedure that is often associated with cardiovascular reactivity (Andreassi, 2000; Edens & Gil, 1995). However, in the present experiment the only physiological measurement associated with the cold pressor was decreased pulse rate. This may be expected with hypothermia but not physical pain. Despite a lack of physiological evidence of pain, subjects in the pain condition still rated their pain higher than those subjects in the control condition and kept their hands in the water for a significantly shorter time than subjects in the control condition. In short, the cold pressor task did not produce the expected physiological changes but did produce both subjective and behavior measurements associated with pain. This finding is consistent with the large body of medical research indicating that much of the pain reported by pain patients is not physical but social and psychological in nature (e.g., Bebbington & Delemos, 1996; Whitehead, Crowell, Heller, Robinson, Schuster, & Horn, 1994). Self-control. The present results suggest that variables affecting subjects' physiological states, or least their subjective phYSiological states, need to be taken into aocount for a complete understanding of human "self-control." Kirk and Logue (1997) found that food-deprived humans showed more impulsiveness than non-food-deprived humans. Kirk and Logue argued that "from an evolutionary perspective, it may actually be adaptive for food-deprived organisms to choose the smaller amount of food available sooner in order to

PAIN AND HUMAN SELF-CONTROL

251

avoid weakness and possibly death due to the lack of the food reinforcer" (p. 225). Likewise, impulsiveness may have been evolutionarily advantageous in times of physical pain, because depending on the source of the pain, a response that produced immediate rather than delayed consequences might prolong survival allowing the organism to live to possibly obtain larger later reinforcers. For example during times of cold instead of using wood to build shelter (self-control), buming the wood for immediate warmth would be impulsive and eliminate the possibility of its later use for shelter. However, this "impulsive" behavior would be adaptive if the individual would have frozen to death before the shelter could have been built. Furthermore, during the respite from the cold provided by the impulsive act, it is possible that other means of warmth would be found such as wearing animal fur. In terms of evolution, in most, if not all, aversive situations behaving in ways that resulted in immediate consequences may have been a relatively more adaptive behavioral strategy than behaving in ways that produced delayed consequences even if the delayed consequences were ultimately greater. Although an impulsive response may not always reduce or permanently terminate the aversive, possibly life threatening, stimulation, self-control by definition (choice for a larger-later reinforcer) certainly would not impact the immediate aversiveness being experienced. A response that produced immediate consequences might prolong survival (at least to the next procreation opportunity), whereas death may occur before any delayed reinforcers were obtained from "self-controlled" behavior. Thus, over successive generations the tendency to respond impulsively in aversive situations may have been selected. In sum, the findings that food deprivation (Kirk & Logue, 1997); aversive noise (e.g., Flora et ai., 1992); and cold pressor pain (present results) all increase impulsive responding suggest that impulsive behavior in times of aversive stimulation (environmental or physical) may be an evolved generalized behavioral tendency. Of course, for most humans in much of the "developed" world, while the problems of pain, aversive feelings, and aversive stimulation have not been solved, the problem of immediate survival has been solved. Thus, although impulsiveness may have been adaptive evolutionarily, in situations many humans now find themselves, impulsiveness is often maladaptive rather than adaptive. But regardless of the validity of the above evolutionary hypothesis, if the current finding that pain increases impulsiveness has generalization, then individuals living with physical pain, or even purely subjective pain, are at increased risk for making impulsive choices, such as drug abuse, binge eating, or work absenteeism; and that while producing immediate reinforcement, these choices may ultimately result in greater pain.

References ANDREASSI, J. L. (2000). Psychophysiology: Human behavior and physiological response (4th ed.). Mahwah, NJ: Lawrence Erlbaum.

252

FLORA ET AL.

BEBBINGTON, P., & DELEMOS, I. (1996). Pain in the family. Journal of Psychosomatic Research, 40, 451-456. EDENS, J. L. & GIL, K. M. (1995). Experimental induction of pain: Utility in the study of clinical pain. Behavior Therapy, 26, 197-216. EISENBERGER, R., MASTERSON, F. A., & LOWMAN, K. (1982). Effects of previous delay of reward, generalized effort, and deprivation on impulsiveness. Learning and Motivation, 13, 378-389. FLORA, S. R. (1995). Molar and molecular contingencies and effects of punishment in a human self-control paradigm. The Psychological Record, 45, 261-281. FLORA, S. R., & DIETZE, M. A. (1993). Caffeine and implusiveness in rats. Bulletin ofthe Psychonomic Society, 31, 39-41. FLORA, S. R., & PAVLIK, w. B. (1992). Human self-control and the density of reinforcement. Journal of the Experimental Analysis of Behavior, 57, 201-208. FLORA, S. R., SCHIEFERECKE, T. R., & BREMENKAMP, H. G. (1992). Effects of aversive noise on human self-control for positive reinforcement. The Psychological Record, 42, 505-517. HASTJARJO, T., & SILBERBERG, A. (1992). Effects of reinforcer delays on choice as a function of income level. Journal of the Experimental Analysis of Behavior, 57, 119-125. KIRK, J. M., & LOGUE, A. W. (1997). Effects of deprivation level on humans' selfcontrol for food reinforcers. Appetite, 28, 215-226. LOGUE, A. W., KING, G. R., CHAVARRO, A., & VOLPE, J. S. (1990). Matching and maximizing in a self-control paradigm using human subjects. Learning and Motivation, 21, 340-368. MOORE, D. S., & MCCABE, G. P. (1993). Introduction to the practice of statistics (2nd ed.). New York: W. H. Freeman and Company. NAVARICK, D. J. (1982). Negative reinforcement and choice in humans. Learning and Motivation, 13, 361-377. RACHLIN, H. (1985). Pain and behavior. The Behavioral and Brain Sciences, 8, 43-83. SOLNICK, J. v., KANNENBERG, C. H., ECKERMAN, D. A., & WALLER, M. B. (1980). An experimental analysis of impulsivity and impulse control in humans. Learning and Motivation, 11, 61-77. WHITEHEAD, W. E., CROWELL, M. D., HELLER, B. R., ROBINSON, J. C., SCHUSTER, M. M., & HORN, S. (1994). Modeling and reinforcement of the sick role during childhood predicts adult illness behavior. Psychosomatic Medicine, 56, 541-550.