| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Division of Cardiovascular Medicine, Department of Medicine (D.S., D.S.S), University of Florida, Gainesville, FL; and Division of Cardiology, Department of Medicine (P.L.B., H.O.), and Department of Endodontics (W.M.), University of North Carolina at Chapel Hill, Chapel Hill, NC.
Address reprint requests to: David Sheffield, PhD, Research Services - 151, Veteran Affairs Medical Center, 1601 SW Archer Road, Gainesville, FL 32608. Email: sheffdc{at}medicine.ufl.edu
| ABSTRACT |
|---|
|
|
|---|
METHODS: Pain perception was measured using a suprathreshold evaluation of pain intensity and pain unpleasantness to a series of thermal stimuli in 27 whites (14 men and 13 women) and 24 African Americans (12 men and 12 women). Blood pressure, depressive symptoms, anxiety state levels, and negative mood were assessed before pain testing to examine whether they might account for any sex or race differences in pain perception that emerged.
RESULTS: African Americans rated the stimuli as more unpleasant and showed a tendency to rate it as more intense than whites. Women showed a tendency to rate the stimuli as more unpleasant and more intense than men. In addition, systolic blood pressure was inversely related to pain intensity. After statistically adjusting for systolic blood pressure, sex differences in pain unpleasantness were reduced and sex differences in pain intensity were abolished; race differences were unaltered.
CONCLUSIONS: These differences in pain perception may be associated with different pain mechanisms: in the case of sex, differences in opioid activity and baroreceptor-regulated pain systems; in the case of race, unmeasured psychological characteristics are suggested by the larger differences in ratings of pain unpleasantness than pain intensity.
Key Words: pain intensity painunpleasantness blood pressure race sex
Abbreviations: ANOVA = analysis of variance; STAI = State-Trait AnxietyInventory; VAMS = Visual Analogue Mood Scale; VAS = VisualAnalogue Scale.
| INTRODUCTION |
|---|
|
|
|---|
One explanation for these sex differences concerns the observation that women have lower blood pressures than age-matched men. Several animal studies (2325) have shown an association between hypertension and diminished sensitivity to noxious stimuli. Decreased pain sensitivity has also been found in hypertensive humans (2630). Furthermore, the relationship between blood pressure and pain sensitivity has been reported in most (21, 3135), but not all (36), studies of normotensive individuals, suggesting a continuous, inverse relationship between blood pressure and pain sensitivity. The possibility that blood pressure differences between sexes might account for sex differences in pain sensitivity has recently been examined by Fillingim and Maixner (21). They found that sex differences in normalized thermal magnitude estimates were reduced after adjusting for resting blood pressure.
Another possibility is that psychological characteristics may explain sex differences in pain perception because many psychological characteristics differ by sex. Furthermore, pain perception and expression are significantly influenced by personality, emotional state, and socialization factors (37, 38). For example, depression has been related to pain perception: a number of researchers have found a relationship between depressed mood and the perception of anginal pain during exercise treadmill testing (3941). Similarly, anxiety state levels have been found to correlate with pain tolerance (4244). Accordingly, we assessed depressive symptoms, anxiety state levels, and negative moods in this study.
In contrast to the numerous studies examining sex differences in pain perception, few studies have examined race differences (for a review, see 5). To our knowledge only three experimental studies have examined differences in the perception of painful stimuli between whites and African Americans. Woodrow et al. (45) examined race differences in pain tolerance to mechanical pressure on the Achilles tendon in more than 40,000 subjects. Whites showed the highest pain tolerance, African Americans the second highest, and Asian Americans the lowest pain tolerance. In addition, Woodrow et al. noted that race differences were more marked in men than women. Similarly, Chapman and Jones (46) reported that North Americans of Northern European descent reported higher thermal pain threshold and tolerance than African Americans. Finally, Walsh et al. (47) examined pain tolerance to the cold pressor test in Anglo-Saxons, Hispanics, and African Americans. Anglo-Saxons had greater pain tolerance than nonAnglo-Saxons (Hispanics and African Americans), across sexes and at any given age. In addition, Sheffield and colleagues (48) found that African Americans reported anginal pain during exercise treadmill testing at nearly twice the rate of whites and had a significantly shorter time to angina, although no differences in disease severity were noted. Similarly, Christy et al. (49) noted that parents of African American infants reported more pain than parents of white infants after immunization, however differences in antigen levels were also found, which made interpretation difficult. Physiological, psychological, or ethnocultural differences may account for these differences in pain perception. Clinical studies have also suggested race differences in pain perception (5054), however differences in disease severity and treatment biases may account for these differences. For example, Stewart et al. (54) found that African American migraineurs reported higher levels of headache pain than white migraineurs in a random sample of telephone interviewees. However, migraine prevalence was higher in whites than African Americans. Even though standard International Headache Society criteria were used to define migraine, it is possible that differences in reporting style account for these data.
The pain measurement technique used in various studies may influence the findings (9, 21). In this study we used the magnitude estimation technique, a suprathreshold evaluation of both pain intensity and unpleasantness. Suprathreshold evaluation of both intensity and affective components of pain may provide more information about pain perception than threshold measurement alone. For example, Gracely (5557) showed that chronic pain patients perceive thermal stimuli normally but find these stimuli less unpleasant.
In this study, we examined sex and race differences in cutaneous pain perception and blood pressure. Specifically, the objectives of this study were to determine a) the differences in cutaneous pain perception between men and women, and whites and African Americans; b) the influence of resting blood pressure and psychological characteristics on cutaneous pain perception in white and African American men and women; c) whether blood pressure or psychological characteristics might explain race or sex differences in pain perception.
| METHODS |
|---|
|
|
|---|
Experimental Protocol
The entire laboratory session was performed between 8 and 10 AM to control for circadian variations in pain perception. The study was reviewed, and the appropriate instructions were delivered by two white investigators, a man and a woman, who conducted the study. After obtaining informed consent, subjects completed three psychological questionnaires. Subjects were then instructed that they would be asked to rate their impression of both the pain intensity and the pain unpleasantness of a series of thermal stimuli. The difference between the intensity and the unpleasantness of a painful stimulus was explained using previously described instructions (58). After a 30-minute rest period, the investigator carefully measured resting blood pressure by cuff sphygmomanometer. Subjects underwent one practice session in rating thermal stimuli. After randomizing subjects to receive either the intensity or unpleasantness first, a series of thermal stimuli (45, 46, 47, 48, and 49°C) were applied to the right volar forearm from a base temperature of 37°C (rate = 19°C/second), and subjects rated their impression of the intensity and the unpleasantness by marking on visual analogue scales. The thermal stimuli were delivered by a probe (9.3-mm square) manufactured by the Medical Instrumentation Department at Yale University to use in the Psychophysiological Investigations of Myocardial Ischemia (PIMI) study (59). Each thermal trial included a 5-second period at base temperature followed by a 5-second period at each trial intensity with 20-second pauses between stimuli. Thermal stimuli were delivered by the investigator with a hand held thermal probe to five different spots marked on the forearm in an orderly fashion. The intensities of thermal stimuli were randomly presented under computer control with a total of 20 trials (four trials at each stimulus intensity). After a 3-minute rest period, the thermal procedure was repeated using the other pain dimension.
Visual Analogue Scales
The VASs used in this study were modified from those described in previous studies (58). The intensity and the unpleasantness VAS consisted of a 150-mm vertical line with the phrase "no pain at all" at the lower end and "the worst pain imaginable" at the upper end. Subjects rated their pain intensity or unpleasantness by marking on the line between the two endpoints. Values were obtained by measuring the distance from the zero to the mark that the subject made on the line. This method has been shown to demonstrate good correlation between repeated rating of a recall pain distant in time and to yield reproducible values for a constant pain stimulus (58, 6062).
Psychological Instruments
The 20-item state component of the STAI (63) was used to measure state anxiety, a transitory emotional condition characterized by tension and apprehensive feelings. The STAI has well established validity and reliability. The Carroll Rating Scale, a 52-item questionnaire with established validity and reliability, was used to evaluate depressive symptoms (64). Finally, the VAMS was used to measure negative mood (65). This instrument consists of seven analogue graphic rating scales and has well-established validity and reliability.
Statistical Analysis
Continuous variables were described with mean ± SE, and categorical variables were described with frequencies and percentages. The mean values of visual analogue scales were calculated for the perceived pain intensity and unpleasantness associated with each of the five stimulus levels for each subject. ANOVA was used to analyze within-subjects effects of temperature and between-subjects effects of race and sex on pain intensity and pain unpleasantness. ANOVA was also used to determine whether blood pressure, age, or psychological characteristics differed by race or sex. General Linear Models were constructed to examine the relationship between hemodynamic and psychological indices and pain perception across the range of temperatures. Finally, analysis of covariance was used to reexamine the relationships between sex, race, and pain ratings, controlling for blood pressure and other factors (eg, age and depressive symptoms) that might influence the relationship. A p value <0.05 was considered significant. All analyses were performed using Systat (66).
| RESULTS |
|---|
|
|
|---|
Pain Perception, Sex, and Race
Four-way (race, sex, temperature, and pain dimension) ANOVAs were applied to the pain ratings. ANOVA revealed a main effect for temperature (F(4,184) = 51.20, p < .0001,
= 0.63), a temperature by pain dimension interaction effect (F(4,184) = 54.19, p < .0001,
= 0.53), a main effect for race (F(1,47) = 4.82, p = .03), and the indication of a main effect for sex (F(1,47) = 3.60, p = .06). There were no other main or interaction effects (p > .10). As inspection of Figure 1 reveals, whites rated the thermal stimuli as less painful (unpleasant and intense) than African Americans. In addition, there was a tendency for men to rate the thermal stimuli as less painful (unpleasant and intense) than women (Figure 2). Finally, there was a temperature by pain dimension interaction: at each temperature, except 47°C, unpleasantness ratings were higher than intensity ratings.
|
|
|
Pain Perception, Sex, and Race After Adjustment
Finally, analyses of covariance were computed to reexamine the relationship between sex, race, and pain perception after adjusting for systolic blood pressure. For pain ratings, race differences in pain perception remained unchanged after controlling for systolic blood pressure (F(1,46) = 5.33, p = .03), whereas sex differences were somewhat reduced (F(1,46) = 1.58, p = .21).
| DISCUSSION |
|---|
|
|
|---|
Despite much interest in the relationship between clinical symptoms, disease severity, and ethnicity (5054), few laboratory studies have examined race differences in pain perception. The race differences in pain perception we observed accord well with those found by Woodrow and colleagues (45) and Chapman and Jones (46): whites were less sensitive to the pain stimuli than African Americans. In addition, Woodrow et al. (45) and Walsh et al. (47) noted that race differences were more marked in men than women. There was no indication of a sex by race interaction in our data, however, our sample size was many times smaller than that of Woodrow et al. (45) or Walsh et al. (47), who did find larger racial differences in men. As has previously been observed (48), resting blood pressure did not account for the race differences we found. It is possible that it is not blood pressure per se, but risk for hypertension, that is related to pain perception (67, 68). A number of investigators have found that family history of hypertension (67, 68) or exaggerated cardiovascular responses to stress (34, 36) instead of or in addition to resting blood pressure are related to pain perception. Because we did not assess family history or cardiovascular reactivity, it was not possible to evaluate whether risk for hypertension explained these race differences. Furthermore, none of psychological (depressive symptoms, state anxiety, and mood) measures we examined explained the relationship between pain sensitivity and race. However, it is worth noting that African Americans had somewhat higher reports of depressive symptoms than whites; the lack of variance with our measure may account for the lack of statistically significant effects. Other psychological and sociocultural factors likely play a role (69). For example, Greenwald (70) found differences in affectivity ratings of cancer pain among ethnic groups despite similarities in ratings of sensitivity. Clearly, the mechanisms underlying the ethnic differences we observed require study.
We also demonstrated that high resting blood pressure levels were associated with decreased pain perception. The mechanisms of hypertension-related hypoalgesia are not completely understood (30) but may be due to differences in opioid activities and baroreceptor-related pain regulatory systems (71, 72). For example, it has been shown that hypertensive rats have increased opioid activity in brain tissue (24) and that diminished pain responses in hypertensive rats, induced experimentally or genetically, can be reversed by opioid receptor antagonists (24, 25, 73). Sheps et al. (28) showed higher levels of circulating ß-endorphins in hypertensive humans relative to normotensive subjects; this difference explained, at least in part, differences in thermal pain threshold and tolerance between hypertensive and normotensive subjects. These studies suggest that decreased pain perception in individuals with elevated blood pressure may be related to opioid activity. Alternatively, the mechanism by which individuals with elevated blood pressure have decreased pain sensitivity may be through activation of baroreceptor systems that engage in central nervous system pain regulation (73). Many studies have demonstrated that activation of carotid sinus baroreceptors, mechanically or pharmacologically, decreased pain responses in animals and pain sensitivity in humans (71, 72, 74, 75).
Finally, we noted a tendency for women to rate the thermal stimuli as more unpleasant and more intense than men. Many previous studies have also found that women are more sensitive to painful stimuli than men (5, 6, 1317). However, a few studies have noted sex differences in thermal pain sensitivity (8, 9, 1822) and have used cross-modality magnitude matching or, in the case of our study, magnitude estimation, more sensitive techniques than threshold and tolerance testing used in other studies. In accordance with past data (21), we found that these sex differences were explained, to some extent but not completely, by differences in resting systolic blood pressure. This implies that physiologic mechanisms, including baroreceptor activation and opioid activity, may underlie sex differences in pain perception. However, the difference between these sex differences when blood pressure was added as a covariate were not different from when they were not. Other studies have found that other physiologic and psychological mechanism are involved (6, 19); the relative contribution of each has yet to be determined.
Despite the potential importance of these findings, certain weaknesses and limitations should be highlighted. First, only one type of pain test was administered. Although other researchers have reported similar findings, it is unclear that these data would generalize across tasks (5, 21). Second, both experimenters were white; it is unclear whether the race of the experimenter influences ratings of pain perception differently by race of the subject (53, 76, 77). Third, we made no measure of socioeconomic status, education or life experiences that might be confounds of the race differences we observed (76, 78). Finally, we did not specify different ethnic subgroups or cultures within the "white" and "African-American" race categories (79); small sample size prevented any further meaningful subgroup analyses. However, Chapman and Jones (46) and Sternbach and Tursky (80) noted pain differences in subgroups of whites and anecdotal reports suggest the same is true for individuals of African decent (81). Similarly, few researchers, ourselves included, have focused on whether observed sex differences in pain perception are determined biologically, or whether they are, in part, cultural or social in nature, and are therefore sex-related differences.
In conclusion, we demonstrated race and, to a lesser extent, sex differences in ratings of thermal pain perception: African Americans were more sensitive than whites, and women were more sensitive than men. Furthermore, we noted that pain perception was related to resting systolic blood pressure levels and that systolic blood pressure explained sex, but not race, differences in pain perception. These data suggest that different pain mechanisms underlie sex and race differences in pain perception. However, to date, few studies have examined the nature and strength of race differences or the underlying mechanisms for sex and race differences. Research suggests that acculturation and assimilation are important predictors of clinical pain (78). Twin studies might also help elucidate whether genetic factors contribute to the race differences we observed. These and other mechanisms should be explored in future research.
| ACKNOWLEDGMENTS |
|---|
|
|
|---|
Received for publication April 19, 1999.
Revision received January 11, 2000.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
D. P. Wallace, K. D. Allen, A. E. Lacroix, and S. L. Pitner The "Cough Trick:" A Brief Strategy to Manage Pediatric Pain From Immunization Injections Pediatrics, February 1, 2010; 125(2): e367 - e373. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Hernandez and N. Sachs-Ericsson Ethnic Differences in Pain Reports and the Moderating Role of Depression in a Community Sample of Hispanic and Caucasian Participants With Serious Health Problems Psychosom Med, January 1, 2006; 68(1): 121 - 128. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. B. Mechlin, W. Maixner, K. C. Light, J. M. Fisher, and S. S. Girdler African Americans Show Alterations in Endogenous Pain Regulatory Mechanisms and Reduced Pain Tolerance to Experimental Pain Procedures Psychosom Med, November 1, 2005; 67(6): 948 - 956. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. S. Sheps, F. Creed, and R. E. Clouse Chest Pain in Patients With Cardiac and Noncardiac Disease Psychosom Med, November 1, 2004; 66(6): 861 - 867. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. T. Whelan, L. Jin, and D. Meltzer Pain and Satisfaction With Pain Control in Hospitalized Medical Patients: No Such Thing as Low Risk Arch Intern Med, January 26, 2004; 164(2): 175 - 180. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. C. Callister Cultural Influences on Pain Perceptions and Behaviors Home Health Care Management Practice, April 1, 2003; 15(3): 207 - 211. [Abstract] [PDF] |
||||
![]() |
K Barakat, Z Wells, S Ramdhany, P G Mills, and A D Timmis Bangladeshi patients present with non-classic features of acute myocardial infarction and are treated less aggressively in east London, UK Heart, March 1, 2003; 89(3): 276 - 279. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. D. Myers, M. E. Robinson, J. L. Riley III, and D. Sheffield Sex, Gender, and Blood Pressure: Contributions to Experimental Pain Report Psychosom Med, July 1, 2001; 63(4): 545 - 550. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. R. Edwards, D. M. Doleys, R. B. Fillingim, and D. Lowery Ethnic Differences in Pain Tolerance: Clinical Implications in a Chronic Pain Population Psychosom Med, March 1, 2001; 63(2): 316 - 323. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |