DOES HUMOR MODERATE THE EFFECTS OF EXPERIMENTALLY-INDUCED S T R E S S ? t,2

Michelle Gayle Newman, Ph.D. and Arthur A. Stone, Ph.D. State University of New York at Stony Brook

ABSTRACT

(8-15). Fry (12-14) has suggested that mirthful laughter produces a physiological state comparable to that which follows vigorous exercise and has concluded that mirth and laughter are useful for the prevention of stress-related maladies such as heart disease, cancer, and depression. Other humor appreciation studies suggest that exposure to humorous stimuli reduces feelings of anger and aggressive behavior (9,16,17), subjective anxiety (15), learned helplessness (18), induced depression (19), and immunosuppression (20). Humor appreciation, however, may not moderate life stress (6,21). To moderate the effect of stress, some authors have argued an individual must be able to produce humor. This was concluded by Martin and Lefcourt (6) based on three retrospective studies and replicated by Nezu, Nezu, and Blisset (7) using prospective analyses. Support for the stress moderation of humor production has also been provided by studies showing that humor production buffers the effect o f stress on strains (22), positively affects immune functioning (23-25), is inversely related to incidence of upper respiratory infections in mothers and their newborn infants (26), and is linked to perceived health (27) and low concurrent distress following breast cancer surgery (28). In contrast, other studies have shown that sense of humor is inversely related to longevity (29) or have failed to show a relationship to longevity (30) and health (31-33). A possible reason for the equivocal nature of humor production research is that it is based on assumptions that have not been empirically substantiated. One of these assumptions is that a retrospective report as to how an individual usually copes is representative of their consistent response to stress. This is the premise of many studies that rely solely on retrospective selfreports to identify individuals who use humor to cope with stress (6,7,31-33). In addition to the multiple biases inherent in retrospective coping measures (34,35), coping studies show little consistency across types of stressors (36-38). To improve upon previous humor studies, we directly assessed the stressmoderating effects of experimentally-manipulated humor production during a uniform stressor. This situational assessment was conducted for retrospectively identified high- and low-trait humor subjects. In this way, we did not rely solely on retrospective reports of coping humor and stress. Another assumption is that the ability to make others laugh indicates the disposition to experience subjective amusement when faced with stress. This assumption is the basis for research on the mental and physical health of comedians (30,39), as well as the validation of humor measures using peer ratings of funniness and number of witty comments produced during a humorous monologue (6). It is not clear, however, that people who are able to make others laugh generally react with feelings of amusement when confronted with stressful situations. The present study had subjects rate their own level of amusement as they produced a humorous monologue during a stressful film.

This study attempted to determine whether humor production moderates mood and physiological responses to stress o f subjects high and low in trait humor. Forty subjects who were high and 40 subjects who were low in trait humor were selected. H a l f o f each group was randomly assigned to one o f two conditions. In one condition they generated a humorous monologue and in the other condition they generated a serious monologue to a silent stressful film. Heart rate, skin conductance level, and skin temperature were taken continuously f o r the fifteen minutes before, during, and fifteen minutes after the film. Pre- and post-stress mood and tension ratings were also recorded. Analyses o f covariance were conducted with baseline mood and tension as covariates. Compared to the production o f a serious narrative, humor production led to lower negative affect, lower tension, and reduced psychophysiological reactivity f o r both high and low trait-humor groups. These results suggest that humor production may be an effective coping strategy, even f o r individuals who do not typically use humor to cope with stress. (Ann Behav Med

1996, 18(2):101-109)

INTRODUCTION Numerous studies have supported the association between stressful life events and ensuing ill health, accidents, depressed mood, and anxiety (1-3). It is also apparent that there is a great deal of diversity in the impact of stress from individual to individual. Whereas most people exhibit negative consequences, some others do not (4). This has led researchers to search for variables which moderate the consequences of life stress. One variable that has been examined is humor. Several studies have shown that those who use humor as a coping strategy fare better than those who do not (5-7). A sense of humor can be thought of in two ways: as the response to humorous stimuli (humor appreciation) or the ability to create humor (coping humor/humor production). Most humor research has focused exclusively on humor appreciation Preparation of this manuscript was supported in part by a Sigma Xi grant in aid of research awarded to Michelle Newman. 2 The authors gratefully acknowledge Louis Castonguay, Ph.D. and C. Barr Taylor, M.D. for their editorial assistance. Reprint Address: M. G. Newman, Ph.D., Department of Psychology, 310 Moore Building, The Pennsylvania State University, University Park, PA 16802-3104. 9 1996 by The Society of Behavioral Medicine. 101

102

ANNALS

OF BEHAVIORAL

Newman and Stone

MEDICINE

In this way, we directly assessed feelings of amusement during a stressful situation. A final unsubstantiated assumption underlying previous research is that the ability to find humor in a stressful situation is a static personality variable rather then an acquired skill (29). This viewpoint is contradicted by authors who promote teaching humor coping to patients as a therapeutic intervention ( 4 0 43). Unfortunately, there is no empirical evidence for humor as a trait or as a skill. By preselecting individuals who scored high and low on a trait humor scale, this study tested whether it was possible for individuals who scored low on a trait humor measure to successfully appl3) the skill of coping humor. The present study was an attempt to measure the stressm o d e r a t i n g effect of humor production by using a research methodology that was not based on the above assumptions. Specifically, we predicted that humor production, as manipulated directly and measured in terms of self-amusement, would moderate the effects of stress for subjects identified as high and low in trait humor.

METHOD Subjects During a mass testing session conducted at State University of New York (SUNY) at Stony Brook, approximately 500 male subjects were administered both the Situational Humor Response Questionnaire (SHRQ) (44) and the Coping Humor Scale (CHS) (6). Eighty subjects were recruited: Forty of the subjects were selected based on high scores on the SHRQ (1 SD or more above the mean), and forty of the subjects were selected based on low scores (1 SD or more below the mean) on the SHRQ. The SHRQ was used as a preselection measure because it has been found to be a valid predictor of male behavior (44). Means and standard deviations were taken from the norms developed by Martin and Lefcourt (44) from a sample of 497 subjects (M = 59.6; SD = 9.06). This created a high-trait humor group (M = 74.4, SD = 5.39) and a low-trait humor group (M = 45.68, SD = 4.39). The high-trait humor group also scored significantly higher on the CHS (M = 22.61, SD = 2.89) than the low-trait humor group (M = 17.41, SD = 2.49) (F(1,73) = 69.55, p < .00005).

Humor Measures Situational Humor Response Questionnaire: This scale was designed to measure sense of humor as measured by the frequency that individuals display mirth in a wide variety of situations (44). It also taps productive aspects of humor (the degree to which an individual tells funny stories and amuses other people) as demonstrated by high correlations between this scale and objective measures of humor production in a lab (e.g. peer ratings of humorousness, number of witty remarks produced during an impromptu comedy routine, rated humorousness of this routine, rated humorousness of a narrative produced while watching a stressful film). Subjects are asked to remember a time when they were in each of eighteen situations and then to indicate the degree to which they experienced mirth on a five-point Guttman-type scale which ranges from "I would not have been amused" to "I would have laughed heartily." In addition, three self-description items are included on the scale. Cronbach alphas of .70 to .79 have been obtained with this measure and test-retest reliability was .70 over a one-month period. It was uncorrelated with a social desirability index and positively correlated with frequency and duration of laughter

during an interview, peer ratings o f sense of humor, number of humorous remarks during an impromptu comedy routine, and humorousness of a narrative produced while watching a stressful film (44).

The Coping Humor Scale: This is a seven-item scale designed specifically by Martin and Lefcourt (6) to assess the degree to which people report intentionally using humor as a means of coping with stressful events (e.g. "I often lose my sense of humor when I ' m having problems"). Items are rated on a four-point scale where 1 -- strongly disagree, 2 = mildly disagree, 3 = mildly agree, and 4 = strongly agree. Internal consistency analyses have resulted in Cronbach alphas in the range of .60 to .70.

Self-Report Mood and Stress Measures Tension Index Scale: This four-item scale measures subjective feelings of arousal (45). Items on the scale are: "TenseNot at all tense," " A n g r y - N o t at all angry," "Irritable-Not at all irritable," and "Frustrated-Not at all frustrated." Subjects rate each item on a scale from 0 (not at all) to 100 (very) indicating how they are feeling at that moment. Items are summed to compute the tension index score. High scores were shown to be positively associated with physiological measures such as blood pressure and heart rate (45).

Positive and Negative Affect Schedule: This paper and pencil measure consists of two ten-item mood scales measuring both positive and negative affect (46). Subjects rated their moods on five-point scales from " v e r y slightly" or "not at all" to " v e r y much" indicating how they were feeling at that moment. The two scales have been found to be largely uncorrelated and stable over a two-month period. The scales are internally consistent and have excellent convergent and discriminant correlations with lengthier measures of underlying mood factors. They are also sensitive to fluctuations in mood (46).

Physiological Measures Continuous skin temperature, heart rate, and skin conductance were monitored and automatically averaged over oneminute periods using a personal computer (PC) with Davicon Medac System/3 software. For this program, the differential input impedance is greater than 1,000,000 Megohms and the common mode rejection is 150 Db. The bandwidth is 32-325 Hz. (60 Hz. notch) 18 Db/oct. with less than .05 Uv. noise. Sensor sets which consist of gold-plated bands mounted on Velcro strips with a log power detector are attached to the subjects' fingers. The Medac System/3 skin temperature measure has a calibrated temperature range of 6 0 - 1 0 0 degrees fahrenheit. Resolution is 1/400 degrees fahrenheit. Absolute accuracy is better than plus or minus 0.25 degrees fahrenheit over calibrated measurement range. It is better than plus or minus 0.1 degree fahrenheit including thermistor interchangeability. The thermal time constant is 320 mS. (standard sensor). The photoplethysmograph, which measures heart rate, has an emitter radiant output of 880 nm. Radiant power compensated for ambient illumination is 3 mW. maximum. The dynamic range is greater than 90 dB. Bandwidth is 0.2-35 Hz. Effective illuminated area is 2 cm 2. The skin conductance apparatus has a calibrated skin conductance range of .80-50 Us/cm 2. The sensor is a gold-plated brass, 2.5 cm 2 effect area, single-ended configuration. The mea-

Does Humor Moderate Stress surement technique is a constant current conductance A.C. log, 5 S. time constant, 3 uA/cm 2 constant current excitation with 12 Hz. low pass, and 18 Db/oct. (60 Hz. notch) filters. Two channels measure both tonic and phasic activity.

Stress Film (Industrial Accident Film) This 13-minute industrial safety film was designed to induce safety-mindedness in wood mill operators by dramatizing three accidents which result from careless safety practices by workmen. This film has induced substantial stress reactions in previous research (47,48).

Manipulation Check Because we were attempting to manipulate the appraisal o f amusement in response to a film that would normally be viewed as stressful, the most important manipulation check was an investigator-designed questionnaire measuring subjective amusement during the film narration. This was administered to determine if subjects were able to view the stressful film as either serious or amusing when so directed. The questionnaire included items such as: "Did you find the narrative that you made-up to be funny?" and " D i d you see this film in a humorous light?" Subjects were asked to rate their responses on a four-point scale from "not at all" to "extremely." Scores were summed to form a composite amusement measure. Also included in this questionnaire were items asking about stressfulness of the film, stressfulness of the directed task, and comfort with the directed task.

Procedure After being selected for the study, half of the high- and low-humor subjects were randomly assigned to either the serious or funny narration conditions. Each condition lasted for one hour. During the experiment, subjects were seated in a comfortable chair. Equipment for the measurement of skin temperature, heart rate, and skin conductance was attached to the first and third fingertip of the non-dominant hand and to the first fingertip of the dominant hand. Subjects were then instructed to sit with their eyes closed and to attempt to relax for 15 minutes. The experimenter then left the subjects alone while continuous heart rate, skin conductance level (SCL), and skin temperature were recorded. When the experimenter returned, subjects were asked to open their eyes and to complete the Positive and Negative Affect Schedule (PANAS) and the Tension Index Scale. Next, subjects were given different instructions dependent upon the film narration condition to which they were assigned (funny or serious). For the funny narration condition, subjects were given the following instructions: The film which you are going to see is about three industrial accidents that are the result of carelessness in the workplace. When the film begins, I will turn on this tape recorder and will then leave the room. The film is a silent one, and while you are watching it, I would like you to try to make up a humorous narrative, describing what you are seeing in the film in as funny a way as you can. I am not interested in your ability to be a comedian, instead I would like you to try to narrate this movie in a way that you find amusing to yourself. Try to see this movie in an amusing light. I realize this is a difficult task, but I would like you to just try your best. Be sure to speak loudly enough so

VOLUME 18, N U M B E R 2, 1996

103

that what you say will be picked up by the tape recorder and please try not to move your hands. For the serious narration condition, subjects were given these instructions: The film which you are about to see is about three industrial accidents that are the result of carelessness in the workplace. When the film begins, I will turn on this tape recorder and will then leave the room. The film is a silent one, and while you are watching it, I would like you to describe very seriously what is taking place in this film in as much detail as possible. Be sure to speak loudly enough so that what you say will be picked up by the tape recorder, and please try not to move your hands. In both conditions, subjects were falsely told that their narratives were being tape-recorded to help ensure cooperation. After the instructions, the experimenter turned on the videotape of the fihn (with the sound turned off) and then left each subject alone in the room for the duration of the film (about 13 minutes). During this time, continuous physiological measures were recorded. As subjects narrated, the experimenter was outside the experimental room and could hear what was being said. All subjects produced a narrative and seemed to be attempting the directed task. When the film was completed, the experimenter returned to the room and administered the PANAS and Tension Index Scale. Following this, subjects were again asked to relax for 15 minutes, and the experimenter left the room. Physiological measures continued to be recorded until the end of the relaxation period. This was done because it has been suggested by some researchers (49) that it is not the amount of physiological reactivity to an event that is significant, but the amount of time necessary to return to baseline that is important to the development of illness. At the end of the 15-minute period, the experimenter returned to the room, shut off the physiological monitors, and administered the PANAS, the Tension Index Scale, and the manipulation check. Upon completing the study, subjects received a brief description of the theoretical background of the study, and the experimenter answered their questions. RESULTS

Data Analyses Measures administered only once (e.g. manipulation check, film stress, stress of task, comfort with task) were initially analyzed using a 2(trait humor: high versus low) X 2(film narration: serious versus funny) Multivariate Analysis of Variance (MANOVA). Statistically significant findings were subsequently analyzed with univariate Analysis of Variance (ANOVAS). For measures taken more than once (i.e. mood, tension appraisal, and physiological reactivity), separate 2(trait humor: high versus low) x 2(film narration: serious versus funny) repeated measures analyses of covariance were conducted with baseline scores as the covariate. Mean heart rate, temperature, and skin conductance were averaged into five summary scores representing five time periods: the last five minutes of baseline, the thirteen-minute stress period, the first five minutes poststress, the second five minutes poststress, and the third five minutes poststress. The preferred analysis for the repeated measures would have been a MANOVA, examining the effects of time, humor level, and film narration; however, this analysis would not have allowed us to control for the separate baselines of

104

ANNALS OF BEHAVIORAL

Newman

MEDICINE

and Stone

TABLE 1 110

Mean Appraisal Scores by Film Narration

@

Measure Manipulation Check Film Stress Task Stress Task Comfort

Funny (N = 40)

Serious (N = 40)

14.00 (4.34) 1.88 (0.76) 2.30 (0.79) ~3.75 (1.39)

9.78 (3.33) 2.15 (0.77) 1.95 (0.85) 4.38 (1.03)

HighHumor Humor

F

100 ~k0

31.81 *** I~

90'

~

80'

2.59 3.659 5.81"

Note: Numbers in the table are means with standard deviations in parentheses. F score refers to main effect. * p < .06, * p < .05, **p < .01, * * * p < .0001.

60

50

between-subjects measures. Therefore, for measures taken more than once, multivariate tests of repeated measures only examined the effect of time.

Multivariate Analysis Results of the multivariate analysis of variance showed a significant effect of humor level F(4,73) = 7.74, p < .001 and film narration F(4,73) = 14.97, p < .001 on ratings of amusement (manipulation check), film stress, stress of task, and comfort with task. There was no observed interaction between humor level and film narration. There was also a significant effect of time on mood measures (F(6,310) = 13.13, p < .0001) and on physiological arousal (F(12,936) = 33.23, p < .0001). Manipulation ,Checks There was a main effect of film narration on amusement F(1,78) = 31.81, p < .0001, indicating that subjects were significantly more amused during the funny film narration than the serious film narration (see Table 1). This result provides support for the success of the manipulation of humor production. To further explore multivariate findings of change over time in subjects' mood and arousal, univariate comparisons were conducted to determine whether reactivity associated with the stressor differed significantly from baseline. Results demonstrated stress reactions for heart rate F(1,79) = 55.21, p < .0001, SCL F(1,79) = 122.32, p < .0001, and skin temperature F(1,79) = 15.12, p < .0001 during the stress manipulation. Stress reactions were also demonstrated in terms of negative affect F(1,79) = 15.31, p < .001 and tension index F(1,79) = 22.59, p < .0001 immediately poststress. These results indicate a significant stress reaction across both film narration conditions. Trait H u m o r Trait humor showed a main effect for amusement F(1,76) = 26.76, p < .0001, indicating that high-humor subjects were significantly more amused during their narration of the film than low-humor subjects (M = 13.83 versus 9.95). There was also a marginal main effect of trait humor on ratings of film stress (F(1,76) = 3.73, p < .06), indicating that low-humor subjects found the film to be more stressful than high-humor subjects (M = 2.18 versus 1.85). Moreover, there was a main effect of trait humor F(1,76) = 6.77, p = .01 on task comfort, showing that high-humor subjects were more comfortable with the film narration task than low-humor subjects (M = 4.40 versus 3.73).

i

i

Post-stress

Follow-up

Time

F I G U R E 1: Tension Index Rating: Time x Trait H u m o r Interaction. Poststress = immediately following film narration; Follow-up = 15 minutes after film narration. Figure depicts means adjusted for baseline differences. Vertical lines depict standard errors of the means. There was also a main effect of trait humor on baseline negative affect F(1,77) = 4.46, p < .05. At baseline, highhumor subjects had lower negative affect than low-humor subjects (M = 13.9 versus 15.9). However, there were no trait humor mood differences with respect to the experimental manipulation. There was a time by trait humor interaction for tension index F(1,75) = 4.01, p < .05. Fisher tests of the means of this interaction showed that whereas high-humor subjects rated themselves as less tense than low-humor subjects immediately following stress (p < .05), fifteen minutes poststress there were no significant differences between the two groups (Figure 1). There were no observed trait humor effects associated with physiological measures. In sum, various effects were associated with trait humor. High-humor subjects appraised the film as less stressful, found more humor in the film, and were more comfortable with the task of being humorous than low-humor subjects. Moreover, high-humor subjects had higher positive affect at baseline and rated themselves as less tense than low-humor subjects immediately following stress.

Film Narration Analyses of effects of film narration showed a main effect of film narration F(1,78) = 5.81, p < .05 on task comfort indicating that subjects were less comfortable with the humorous narration than with the serious narration (see Table 1). There was also a marginal effect of film narration on task stress F(1,78) = 3.65, p < .06 indicating that subjects rated the funny film narration task as marginally more stressful than the serious film narration task (Table 1). There was no observed effect of film narration on ratings of film stress. Negative affect showed a time by film narration interaction F(1,75) = 4.96, p < .05. Immediately post-stress, subjects producing the serious film narration had higher negative affect than those producing the funny film narration. Fifteen minutes poststress, there were no significant differences between the two film narration conditions. There was also a marginal time x

Does Humor

Moderate

Stress

VOLUME

TABLE 2 Mean Mood and Tension Scores by Film Narration

Film Narration

PA

Serious~ M

Funny~

LSM M

30.78 (6.41) 29.83 (7.85)

LSM NA

Serious~ M

Funny~

LSM M LSM

TI

Serious~ M

Funny~

LSM M LSM

14.80 (4.33) 15.03 (4.43)

29.13 (7.81) 28.72 29.15 (8.60) 29.55

26.80 (6.60) 26.36 1.39 27.45 (9.09) 27.89

18.25 13.23 ( 6 . 6 4 ) (3.05) 9 13.30 16.73 14.48 ( 6 . 9 8 ) (6~23) 16.63 14.40

69.83 1 0 1 . 3 8 59.80 (59.47) (80.46) (52.51) 104.70 62.57 76.05 9t.85 69.03 (58.83) (75.55) (82.07) 88.52 66.26

(p I.

T

105

Serious Funny

12.5 11.5

.42

10.5 9.5

o

8.5

I

I

I

I

Stress

5 Min Post

10 Min Post

15 Min Post

Time

4.96*

100.52

~

13.5

.= .11

2, 1 9 9 6

14.5

Immedi15 F ately Minutes F (Time PostPost- (Film x Film Baseline s t r e s s s t r e s s Narr.) Narr.)

Measure

18, N U M B E R

3.72t

Note: Numbers in the table are means with standard deviations in

T

-'--O"----

Serious Funny

gl,

90'

ae

,s

parentheses. F scores refer to ANCOVA; LSM = least square means; Film Narr. = film narration; PA = positive affect; NA = negative affect; TI = tension index, t P < .06, *p < .05, 9N = 40. 80 film narration interaction for tension index F(1,75) = 3.72, p < .06. Fisher tests indicated that subjects producing the serious film narration appraised themselves as significantly more tense immediately post-stress than those producing the funny film narration (p < .05). Fifteen minutes post-stress, there were no significant differences between the two film narration conditions (Table 2). Taken together, these results suggest that regardless of trait humor, there were short-term effects of humor production on self-reported negative affect and tension scores (Table 2). Physiological measures also indicated that all subjects were more relaxed following the funny film narration than following the serious film narration (Figure 2). Findings that support this pattern include main effects of film narration on skin temperature F(1,75) = 6.87, p < .01 and skin conductance level F(1,75) = 3.8, p < .05, as well as a marginal effect of film narration on heart rate F(1,75) = 3.68, p < .06. Inspection of the means of these analyses indicated that subjects demonstrated greater reactivity (i.e. higher heart rate and skin conductance level and lower temperature) during the serious film narration than during the funny film narration (Table 3). Moreover, those producing the serious film narration recovered more slowly from stress than those producing the funny film narration (Figure 2). There was also a time x trait humor • film narration interaction for skin temperature F(3,225) : 5.00, p < .001. Fisher tests indicated that low-humor subjects recovered more quickly from stress when they narrated the film humorously than when they narrated the film seriously (p < .0l), whereas high-humor individuals recovered equally quickly following both film narrations (Figure 3).

?

I

i

i

I

I

Stress

5 Min Post

10 Min Post

15 Min Post

Time

93 92"

i.

[-

91 90 89 88 87

I

I

I

I

Stress

5 Min Post

10 Min Post

15 Min Post

Time F I G U R E 2: Effects o f F i l m Narration on Physiological Arousal. Serious = serious film narration; F u n n y = f u n n y film narration. Stress = average o f thirteen-minute stress period; 5 M i n Post = average o f first five m i n u t e s poststress; 10 M i n Post = average o f second five minutes poststress; 15 M i n Post = average o f third five minutes poststress. Figure depicts m e a n s adjusted for baseline differences. Vertical lines depict standard errors o f the means.

On the whole, these findings indicate that regardless of trait humor, there were multiple effects of the film narration. Immediately following the funny film narration, subjects reported lower negative affect and tension index scores than following the serious film narration. Moreover, for all three of the

ANNALS OF BEHAVIORAL

106

MEDICINE

Newman

and Stone

TABLE 3 Mean Skin Temperature, Heart Rate, and SCL by Film Narration Measure

Film Narration

Temp.

Set?

M

Funnya

LSM M

Baseline

During Stress

5 Minutes Post-stress

10 Minutes Post-stress

15 Minutes Post-stress

91.08 (2.27)

89.22 (3.27) 89.04 89.9 (3.58) 90.10

88.65 (4.00) 88.45 89.84 (4.34) 90.04

89.66 (3.94) 89.49 91.26 (3.52) 91.43

90.70 (3.82) 90.57 92.23 (2.93) 92.36

99.09 (15.64) 97.91 91.29 (16.76) 92.48

94.43 (15.43) 93.45 87.68 (12.91) 88.65

87.12 (12.37) 85.83 80.78 (12.60) 82.06

86.36 (12.75) 85.17 80.97 (14.23) 82.16

11.51 (6.09) 13.40 13.66 (12.49) l 1.77

l 1.13 (5.81) 13.26 13.61 (14.06) 11.48

9.32 (5.13) 11.60 12.46 (15.07) 10.18

8.37 (4.88) 10.77 11.91 (15.97) 9.51

90.55 (4.32)

LSM

HR

Set?

M

84.56 (9.52)

LSM

Funnya

M

80.95 (12.75)

LSM

SCL

Ser?

M

5.03 (2.44)

LSM

Funnya

M

7.45 (7.81)

LSM

F

6.87**

3.68]

3.80*

Note: Numbers in the tables are means with standard deviations in parentheses.

F scores refer to main effects of condition when baseline is a covariate. LSM = least square means; Film Narr. = film narration; Ser. = serious film narration; Temp. = skin temperature; HR = heart rate; SCL = skin conductance level. aN = 40, I P < .06, *p < .05, **p < .01.

physiological variables, both high- and low-humor subjects demonstrated lower reactivity (i.e. lower heart rate and skin conductance level and higher temperature) to the funny film narration than to the serious film narration. Overall, these re-

95 --'-O---

High Humor Serious Narration

94 High Humor Funny Narration

93 I.

~

Low Humor Serious Narration -'l--"

Low Humor Funny Narration

92

I,J

em

~

9o

~ 2~

89

88

87

| Stress

i 5 Min Post

i 10 Min Post

i 15 Min Post

Time

F I G U R E 3: Skin Temperature: T i m e • Trait H u m o r x F i l m Narration Interaction. H i g h H = high h u m o r ; L o w H = low h u m o r ; Ser N = serious film narration; F u n N = f u n n y film narration. Stress = average o f thirteen-minute stress period; 5 M i n Post = average of first five m i n u t e s post-stress; 10 Min Post = average of second five m i n u t e s post-stress; 15 Min Post = average o f third five m i n u t e s post-stress. Figure depicts m e a n s adjusted for baseline differences. Vertical lines depict standard errors o f the means.

suits suggest that the production of humor may be an effective stress-reducer for both high- and low-humor subjects. DISCUSSION This study attempted to determine if humor production moderated the effects of experimentally-induced stress and whether it moderated the impact of such stress for retrospectively identified high- and low-humor subjects. Findings supported the stress-moderating role of humor, both in terms of physiological and mood measures. Subjects who produced a humorous narrative had lower negative affect and rated themselves as marginally less tense immediately after the stressful film than those who produced a serious narrative. Moreover, for all three physiological variables, high- and low-humor subjects who produced a funny film narration demonstrated lower reactivity during the stressful film than those producing the serious narration. In addition, those in the serious condition recovered from stress more slowly than those in the funny narration condition. Despite ratings of greater discomfort with the task of producing a humor narrative, low-humor subjects demonstrated a buffering effect of humor. Moreover, despite marginal ratings of greater stressfulness associated with the task of being humorous, all subjects demonstrated a buffering effect of humor. Thus, as predicted, humor production moderated the effects of stress for subjects measured retrospectively as both high and low in trait humor. A possible mechanism underlying the effects of humor production is that the ability to find humor in a stressful situation depends primarily on cognitive reappraisal. This may serve to "distance" an individual from the stressfulness of a situation, thus reducing its threatening nature (50). Previous research (51) has demonstrated that if the sound track of a stressful film is changed in such a way that it is more intellectual,

Does Humor Moderate Stress thereby distancing the person from the emotional content of the film, reactivity is reduced. Humor may serve the same distancing function. However, if reappraisal of the stressfulness of the film led to distancing from the emotional content, it would follow that subjects who narrated the film humorously would rate it as less stressful than those who narrated the film seriously. This was not the case, as there was no observed effect of film narration on ratings of film stressfulness. Another possibility is that rather than lead to distancing from emotional content, reappraisal changes the valence of the emotional response. This hypothesis is supported by research showing that high-humor subjects tend to appraise an exam as more of a positive challenge than low-humor subjects (52). A second potential mechanism leading to the effect of humor production may be, as shown by Fry (12-14), that mirth decreases sympathetic activity, creating a state of relaxation and lowering negative mood. This explanation is supported by the results of the present study, as the production of humor led to lowered negative affect and lower physiological reactivity than production of a serious narration. The moderation of physiological reactivity to stress by using humor may have important health implications. Although it has not been unequivocally demonstrated, converging evidence from a number of experiments suggests that individuals who are hyperreactive to stress may be at greatest risk for both the development of hypertension (53-56) and the development of coronary heart disease (57-60). The stress reduction from humor demonstrated in this study may well be protective or prophylactic for stress-related conditions. Contrary to previous studies, the present investigation allowed for the comparison of trait humor (measured retrospectively) and situational humor (experimentally-manipulated). An interesting finding was that whereas the latter was associated with stress moderation, the former was not. The lack of a differential effect for trait humor cannot be attributed to an absence of differences between the two trait humor groups. Consistent with previous research (7), we found that high-humor subjects had lower negative mood than low-humor subjects at baseline, that they rated their tension lower, appraised the film as less stressful, viewed the film as more amusing, and were more comfortable narrating the film humorously than lowhumor subjects. Our results also cannot be explained by an imbalance between the two film narration conditions. Subjects rated the task of producing the humorous monologue to be marginally more stressful than the task of producing the serious monologue. In addition, subjects reported being less comfortable with the task of humor production than with the serious narration. Therefore, any imbalance between the tasks appears to be in the opposite direction of our results (i.e. the more stressful humorous narrative task led to lower physiological arousal). Our results were also not due to failures of the manipulations. All subjects produced a narrative, and findings indicated that high- and low-humor subjects rated themselves as being more subjectively amused during the funny film narration than during the serious narration. Moreover, stressfulness of the film was demonstrated by a significant increase in physiological arousal from baseline. Stress effects exhibited in this study (e.g. 13 heartbeats per minute and 6.34 micrornhos in SCL) were comparable to those demonstrated by Lazarus at the time his research was published (11 heartbeats per minute, 6.5 micromhos) (61), (12 heartbeats per minute, 6 micromhos) (62), as well as change found in other studies that have imposed stress in the

VOLUME 18, N U M B E R 2, 1996

107

forms of threat of electric shock (10 heartbeats per minute) (63), Subincision film (2.66 heartbeats per minute, 5 micromhos) (64), and Stroop test (18 heartbeats per minute, 3.3 micromhos SCL) (65). A potential explanation for the lack of trait humor differences found in this and other studies (29-33) may be that trait humor measures are based on incorrect assumptions. First, contrary to assumptions of previous research (6,7,31-33), a retrospective report regarding usual coping reactions may not be representative of a consistent response to stress. Research shows that people do not cope consistently across types of stressors (36-38) and that the context or situation must be taken into account when examining coping efficacy (61,62). Therefore, ratings on the SHRQ of usual coping strategies may not adequately reflect day-to-day coping. It should be noted, however, that it is not possible to determine from this study whether trait humor measures are adequate in predicting who will use humor in the real word. The forced nature of the experimental humor task obscured any real-world variability in use of humor production that might have been predicted by the SHRQ. A second assumption is that humor production should be measured as the ability to make others laugh. As a result, humor measures such as those by Martin and Lefcourt (44) have been validated using ratings of objective funniness (e.g. peer ratings of sense of humor, number of humorous remarks during an impromptu comedy routine, and humorousness of a narrative produced while watching a stressful film). As stated earlier, there is no evidence that ability to make others laugh is the same thing as self-amusement in response to a stressful stimulus) A third assumption, and one directly challenged by results of this study, is that the ability to respond to stress with humor is an unchangeable trait rather than an acquired skill. Our findings point to a stress-moderating effect of humor production for both high- and low-humor subjects. Such results suggest that if they make the effort, it may be feasible for individuals low in trait humor to successfully apply the skill of coping humor. This viewpoint is also supported by those who promote teaching humor coping to patients as a therapeutic intervention (4043,63). Coping humor might be implemented in the same way as other cognitive restructuring techniques. Although this research contains some similarities to research conducted in the 1960s by Lazarus (e.g. 51), it is different in significant ways. Our research asked subjects to narrate the film themselves, whereas Lazarus changed the narrative for his subjects. The difference between these two manipulations can be viewed as similar to the difference between humor appreciation and production. Moreover, all of the narratives used by Lazarus (e.g. 51) were serious and this research never examined the effect of a humorous narrative on stress reactions. In addition, Lazarus did not examine differences between highand low-trait humor subjects. Limitations to this study should be noted. First, this was a contrived analogue study, and interpretations about the generAlthough Martin and Lefcourt (44) also validated their humor scale using frequency and duration of laughter during an interview, laughter in this situation may be indicative of nervousness. Moreover, there is some evidence that humor results in improved immunity even in the absence of overt laughter (66). Therefore, signs of overt laughter may not be the active component of humor production. Instead, as indicated by the results of the present study, subjective amusement may be the active variable.

108

ANNALS OF BEHAVIORAL MEDICINE

alizability of these results are tentative. Second, any generalizability of these results may be limited to relatively low-level stressors. In a previous study (70), the Industrial Accident film was rated as "slightly stressful" whereas the Subincision film (71) was rated as "moderately stressful." Moreover, high- and low-humor subjects had difficulty producing humor to the more stressful Subincision film and showed greater arousal during the humorous narrative than during the serious narrative (70). Therefore, humor production may only moderate the effects of low-level stressors. Third, it is possible that subjects exaggerated their reports of amusement during the humorous film narration due to demand characteristics. This may have biased the manipulation check. However, in prior research (70) using the same design but two different stress films, subjects reported being unable to create an amusing narrative to one of the films, and their lack of amusement during this condition was associated with higher arousal. This provides some evidence that demand characteristics may not be influencing amusement ratings. Fourth, this experiment was conducted on an all male subject sample. Since there is gender variability in coping (67,68), it is not clear whether the present results will generalize to a female population. Finally, our failure to include a neutral condition makes some of our interpretations of the data tentative. Without such a condition, we cannot definitively determine whether differences between funny and serious narrative groups were due to a stress-buffering effect of humor production or due to a stress-inducing effect of seriousness production. Future research should attempt to replicate and extend the present results using females, subjects who are older and younger, and various types of stressors. Moreover, longitudinal research should be conducted to examine the long-term effects of training subjects to use humor as a coping strategy.

REFERENCES (1) Holmes TH, Masuda M: Life changes and illness susceptibility. In Dohrenwend BS, Dohrenwend BP (eds), Stressful Life Events: Their Nature and Effects. New York: Wdey, 1974. (2) Paykel ES: Life stress and psychiatric disorder: Applications of the clinical approach. In Dohrenwend BS, Dohrenwend BP (eds), Stressful Life Events: Their Nature and Effects. New York: Wiley, 1974. (3) Rabkin JC, Streuning EL: Life events, stress, and illness. Science. 1976, 194:1013-1020. (4) Millon T, Green C, Meagher R (eds), Handbook of Clinical Health Psychology. New York: Plenum Press, 1982, 377-399. (5) Deaner SL, McConatha JT: The relationship of humor to depression and personality. Psychological Reports. 1993, 72(3, Pt. 1): 755-763. (6) Martin RA, Lefcourt HM: Sense of humor as a moderator of the relation between stressors and moods. Journal of Personality and Social Psychology. 1983, 45:1313-1324. (7) Nezu AM, Nezu CM, Blisset SE: Sense of humor as a moderator of the relation between stressful events and psychological distress: A prospective analysis. Journal of Personality and Social Psychology. 1988, 54:520-525. (8) Baron RA: Aggression-inhibiting influence of sexual humor. Journal of 'Personality and Social Psychology. 1978, 36:189-197. (9) Berkowitz L: Aggressive humor as a stimulus to aggressive responses. Journal of Personality and Social Psychology. 1970, 16: 710-717. (10) Carroll JL: The relationship between humor appreciation and perceived physical health. Psychology--A Quarterly Journal of Human Behavior. 1990, 27:34-37.

Newman and Stone (11) Goldstein JH, Mantell M, Pope B, Derks P: Humor and the coronary-prone behavior pattern. Current Psychology: Research and Reviews. 1988, 7:115-121. (12) Fry WF, Rader C: The respiratory components of mirthful laughter. Journal of Biological Psychology. 1977, 19:39-50. (13) Fry WF: Humor and the human cardiovascular system. In Mindess H, Turek J (eds), The Study of Humor. Los Angeles: Antioch University, 1979. (14) Fry WF, Savin M: Mirthful laughter and blood pressure. Third International Conference on Humor. Washington, DC: 1982. (15) Ribordy SC, Holmes DS, Buchsbaum HK: Effects of affective and cognitive distractions on anxiety reduction. Journal of Social Psychology. 1980, 112:121-127. (16) Mueller C, Donnerstein E: The effects of humor-induced arousal upon aggressive behavior. Journal of Research in Personality. 1977, 11:73-82. (17) Singer DL: Aggression arousal, hostile humor, catharsis. Journal of Personality and Social Psychology. 1968, 8(1, Pt. 2): 1-14. (18) Trice AD: Alleviation of helpless responding by a humorous experience. Psychological Reports. 1985, 57:474. (19) Danzer AD, Alexander J, Klions HL: Effect of exposure to humorous stimuli on induced depression. Psychological Reports. 1990, 66(3, Pt. 1):1027-1036. (20) Berk LS, Tan SA, Fry WE et al: Neuroendocrine and stress hormone changes during mirthful laughter. American Journal of The Medical Sciences. 1989, 296:390-396. (21) Safranek R, Schill T: Coping with stress: Does humor help? Psychological Reports. 1982, 51:222. (22) DesCamp KD, Thomas CC: Buffering nursing stress through play at work, Western Journal of Nursing Research. 1993, 15:619-627. (23) Dillon KM, Minchoff B, Baker KH: Positive emotional states and enhancement of the immune system. International Journal of Psychiatry in Medicine. 1985-1986, 15:13-17. (24) Martin RA, Dobbin JP: Sense of humor, hassles, and immunoglobulin A: Evidence for a stress-moderating effect of humor. International Journal of Psychiatry in Medicine. 1988, 18:93-105. (25) Lefcourt HM, Davidson-Katz K, Kueneman K: Humor and immune system functioning. Humor: International Journal of Humor Research. 1990, 3:305-321. (26) Dillon KM, Totten MC: Psychological factors, immunocompetence, and health of breast-feeding mothers and their infants. Journal of Genetic Psychology. 1989, 150:155-162. (27) Carroll JL, Shmidt JL: Correlation between humorous coping style and health. Psychological Reports. 1992, 70:402. (28) Carver CS, Pozo C, Harris SD, et al: How coping mediates the effect of optimism in distress: A study of women with early-stage breast cancer. Journal of Personality and Social Psychology. 1993, 65:375-390. (29) Freidman HS, Tucker JS, Tomlinson-Keasey C, et al: Does childhood personality predict longevity? Journal of Personality and Social Psychology. 1993, 65:t76-185. (30) Rotton J: Trait humor and longevity: Do comics have the last laugh? Health Psychology. 1992, 11:262-266. (31) Anderson CA, Arnoult LH: An examination of perceived control, humor, irrational beliefs, and positive stress as moderators of the relation between negative stress and health. Basic and Applied Social Psychology. 1989, 10:101-117. (32) Labott SM, Martin RB: The stress-moderating effects of weeping and humor. Journal of Human Stress. 1987, 13:159-164. (33) Porterfield AL: Does sense of humor moderate the impact of life stress on psychological well-being? Journal of Research in Personality. 1987, 21:306-317. (34) Stone AA, Kennedy-Moore E, Newman MG, Greenberg MA, Neale JM: Conceptual and methodological issues in current coping assessments. In Carpenter BN (ed), Personal Coping: Theory, Research, and Application. Westport, CT: Praeger Publishers, 1992, 15-29,

Does Humor

Moderate

Stress

(35) Stone AA, Greenberg MA, Kennedy-Moore E, Newman MG: Self-report situation-specific coping questionnaires: What are they measuring? Journal of Personality and Social Psychology. 1991, 61:648-658. (36) Compas BE, Forsythe CJ, Wagner BM: Consistency and variability in causal attributions and coping with stress. Cognitive Therapy and Research. 1988, 12:305-320. (37) Patterson TL, Smith LW, Grant I, et al: Internal vs. external determinants of coping responses to stressful life-events in the elderly. British Journal of Medical Psychology. 1990, 63:149-160. (38) Wong M, Kaloupek DG: Coping with dental treatment: The potential impact of situational demands. Journal of Behavioral Medicine. 1986, 9:579-597. (39) Fisher S, Fisher RL: Personality and psychopathology in the comic. In McGhee PE, Goldstein JH (eds), Handbook of Humor Research. New York: Springer Verlag, 1983, 41-49. (40) Saper B: Humor in psychiatric healing. Psychiatric Quarterly. 1988, 59:306-319. (41) Prerost FJ: Use of humor and guided imagery in therapy to alleviate stress. Journal of Mental Health Stress. 1988, 10:16--22. (42) Prerost FJ: Intervening during crises of life transitions: Promoting a sense of humor as a stress moderator. Counseling Psychology Quarterly. 1989, 2:475-480. (43) Prerost FJ: A strategy to enhance humor production among elderly persons: Assisting in the management of stress. Activities, Adaptation and Aging. 1993, 17:17-24. (44) Martin RA, Lefcourt HM: Situational humor response questionnaire: Quantitative measure of sense of humor. Journal of Personality and Social Psychology. 1984, 47.'145-155. (45) Glass DC, Lake CR, Contrada RJ, Kehoe K, Edanger LR: Stability of individual differences in psychological responses to stress. Health Psychology. 1983, 2:317-341. (46) Watson D, Clark LA, Tellegin A: Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology. 1988, 54:10631070. (47) Alfert E: Reactions to a vicariously experienced and direct threat. Unpublished Doctoral Dissertation. University of California, Berkeley, 1964. (48) Birnbaum R: Autonomic reaction to threat and confrontation conditions of psychological stress. Unpublished Doctoral Dissertation. University of California, Berkeley, 1964. (49) Gal R, Lazarus RS: The role of activity in anticipating and confronting stressful situations. Journal of Human Stress. 1975, 1:420. (50) Dixon NF: Humor: A cognitive alternative to stress? In Sarason IG, Spielberger CD (eds), Stress and Anxiety (Vol. 7). Washington, DC: Hemisphere, 1980, 281-289. (51) Lazarus RS: A laboratory approach to the dynamics of psychological stress. American Psychologist. 1964, 19:400-411. (52) Kuiper NA, Martin RA, Olinger LJ: Coping humour, stress, and cognitive appraisals. Canadian Journal of Behavioural Science. 1993, 25:81-96. (53) Ditto WB: Parental history of essential hypertension and cardiovascular response to psychological stressors. Doctoral Dissertation, Indiana University, 1983. (54) Fredrikson M, Dimberg U, Frisk-Hamberg M, Strom G: Hemodynamic and electrodermal correlates of psychogenic stimuli in

VOLUME

(55)

(56)

(57)

(58)

(59)

(60)

(61)

(62)

(63)

(64)

(65)

(66)

(67)

(68)

(69)

(70)

(71)

18, N U M B E R

2, 1 9 9 6

109

normotensive and hypertensive subjects. Biological Psychology. 1982, 15:63-73. Henry JP, Cassel JC: Psychosocial factors in essential hypertension: Recent epidemiological and animal experimental evidence. American Journal of Epidemiology. 1969, 90:1971. Schneiderman N: Behavior, autonomic function, and animal models of cardiovascular pathology. In Dembroski TM, Schmidt TH, Blumchen G (eds), Biobehavioral Bases of Coronary Heart Disease. Basel, Switzerland: Karger, 1983, 304-364. Devereaux RB, Picketing TG, Harshfield GA, et al: Left ventricular hypertrophy in patients with hypertension: Importance of blood pressure response to regularly occurring stress. Circulation. 1983, 68:470-476. Keys A, Taylor HL, Blackburn H, et al: Mortality and coronary heart disease among men studied for 23 years. Archives oflnternal Medicine. 1971, 128:201-214. Sokolow N, Werdegar D, Perloff D, Cowan R, Brenenstuhl H: Preliminary studies relating recorded blood pressures to daily life events in patients with essential hypertension. In Koster M, Musaph H, Visser P (eds), Psychosomatics in Essential Hypertension. Basel, Switzerland: Karger, 1970. Thomas CB: Genetic patterns of hypertension in man. In Onesti G, Kim KE, Moyer JH (eds), Hypertension: Mechanisms and Management. New York: Grune Stratton, 1973, 67-73. Folkins CH, Lawson KD, Opton Jr. EM, Lazarus RS: Desensitization and the experimental reduction of threat. Journal of Abnormal Psychology. 1968, 73:100-113. Lazarus RS, Opton Jr. EM, Nomikos MS, Rankin NO: The principle of short-circuiting of threat: Further evidence. Journal of Personality. 1965, 33:622-635. Breznitz S: Incubation of threat: Duration of anticipation and false alarm as determinants of fear reaction to an unavoidable frightening event. Journal of Experimental Research in Personality. 1967, 2:173-179. Lazarus RS, Alfert E: The short-circuiting of threat by experimentally altering cognitive appraisal. Journal of Abnormal and Social Psychology. 1964, 69:195-205. Stone AA, Valdimarsdottir HB, Katkin ES, et al: Effects of mental stressors on mitogin-induced lymphocyte responses in the laboratory. Psychology and Health. 1993, 8:269-284. Labott SM, Ahleman SW, Mark E, Martin RB: The physiological and psychological effects of the expression and inhibition of emotion. Behavioral Medicine. 1990, 16:182-189. Folkman S, Lazarus RS: An analysis of coping in a middle-aged community sample. Journal of Health and Social Behavior. 1980, 21:219-239. Walbott HG, Scherer KR: Stress specificity: Differential effects of coping style, gender, and type of stressor on autonomic arousal, facial expression, and subjective feeling. Journal of Personality and Social Psychology. 1991, 61:147-156. Saper B: Humor in psychotherapy: Is it good or bad for the client? Professional Psychology--Research and Practice. 1987, 18:360367. Newman MG, Stone AA: Humor as a moderator of physiological responsiveness to stress. American Psychological Association. Boston, MA: August 1990. Lazarus RS: Psychological Stress and the Coping Process. New York: McGraw-Hill, 1966.