Sensory, Affective, and Catastrophizing Reactions to Multiple Stimulus Modalities: Results from the Oklahoma Study of Native American Pain Risk (OK-SNAP)

Abstract

Native Americans (NAs) have a higher prevalence of chronic pain than any other U.S. racial/ethnic group; however, little is known about the mechanisms for this pain disparity. This study used quantitative sensory testing (QST) to assess pain experience in healthy, pain-free adults (N=137 NAs [87 female], N=145 non-Hispanic whites [NHW; 68 female]) following painful electric, heat, cold, ischemic, and pressure stimuli. Following each stimulus, ratings of pain intensity, sensory pain, affective pain, pain-related anxiety, and situation-specific pain catastrophizing were assessed. Results suggested that NAs reported greater sensory pain in response to suprathreshold electric and heat stimuli, greater pain-related anxiety to heat and ischemic stimuli, and more catastrophic thoughts in response to electric and heat stimuli. Sex differences were also noted; however, with the exception of catastrophic thoughts to cold, these were not moderated by race/ethnicity. Together, findings suggest NAs experience heightened sensory, anxiety, and catastrophizing reactions to painful stimuli. This could place NAs at risk for future chronic pain and could ultimately lead to a vicious cycle that maintains pain (e.g., pain➔anxiety/catastrophizing➔pain).

Perspective:

Native Americans experienced heightened sensory, anxiety, and catastrophizing reactions in response to multiple pain stimuli. Given the potential for anxiety and catastrophic thoughts to amplify pain, this may place them at risk for pain disorders and could lead to a vicious cycle that maintains pain.

Introduction

Native Americans (NAs) have a higher prevalence of chronic pain than any other U.S. racial/ethnic group^4,35,83, yet little has been done to understand what contributes to this pain disparity. Access to health care^1,46, difficulties assessing/treating pain cross-culturally⁷⁸, and/or provider biases^31,32 may contribute to this disparity; but, it could also stem from differences in the way that pain is processed and experienced³⁷.

Differences in pain experience can be studied in reaction to clinical pain (e.g., fibromyalgia); but, this approach suffers from an internal validity problem because it is unclear whether the source of pain is consistent across persons/groups being studied. As such, experiential differences could simply be due to differences in the disease state. This problem is addressed by assessing reactions to quantitative sensory testing (QST) that involves the delivery of controlled, painful stimuli on healthy, pain-free individuals.

QST studies have been conducted to examine racial/ethnic differences in pain, but most have studied African-Americans (and to a lesser degree Hispanics), because these minority groups also experience high rates of chronic pain^52,63. This research suggests they may be at a higher pain risk due, in part, to heightened affective-motivational reactivity to pain, because group differences (when compared to non-Hispanic whites, NHW) are largest in response to suprathreshold stimuli (e.g., pain tolerance) that evoke strong affective reactions^52,63. Notably, these suprathreshold measures are believed to have the greatest relationship with, and relevance to, clinical pain¹⁹. Given evidence that affective factors (e.g., anxiety) can enhance pain^{21,44,59,64,66}, greater affective-motivational reactions may create a vicious cycle to enhance and promote future pain (e.g., pain➔anxiety➔pain)^14,30,66,71. Importantly, it is pain-related affect (e.g., fear/anxiety) that is the primary cause of pain-related suffering³⁸, and clinical studies have shown that racial/ethnic minorities report greater pain-related negative affect in response to clinical pain^29,63.

To our knowledge, only 4 published studies have used QST to study pain in NAs^41,47,48,65. In contrast to studies of other minorities, 3 noted reduced pain sensitivity and/or a lack of pain facilitation^47,48,65, and 1 noted no group differences⁴¹. This suggests mechanisms for pain risk may be different for NAs. However, all 4 studies suffered from small sample sizes, or assessed only one pain modality^41,65.

To address these gaps in the literature, the present study assessed responses to electric, heat, cold, ischemic, and pressure stimuli in 137 NA and 145 NHW participants. Immediately following each pain task, sensory reactions were assessed from the McGill Pain Questionnaire-Short Form (MPQ-SF) sensory scale and the MPQ-SF visual analog scale (VAS) for pain intensity, whereas affective reactions were assessed from the MPQ-SF affective scale and from a VAS measuring pain-related anxiety. Finally, because pain catastrophizing is a cognitive-affective reaction that is known to promote pain⁵⁰, the Pain Catastrophizing Scale was administered after each task with instructions to report on catastrophizing that occurred during the painful tasks (i.e., situation-specific pain catastrophizing)¹¹. Given that group differences in the distribution of sex between racial/ethnic groups were found, sex was controlled for in the analyses.

Materials and Methods

Participants

These data were collected as part of the Oklahoma Study of Native American Pain Risk (OK-SNAP). Healthy, pain-free participants were recruited from newspaper ads, tribal newspapers, fliers, personal communications with NA groups, email announcements, and online platforms (e.g., Facebook). Exclusion criteria included: 1) <18 years old, 2) history of cardiovascular, neuroendocrine, musculoskeletal, neurological disorders, 3) chronic pain or current acute pain, 4) BMI≥35 (due to difficulties recording electromyogram for other tasks), 5) use of anti-depressants, anxiolytic, analgesic, stimulant, or anti-hypertensive medication, 6) current psychotic symptoms (assessed by Psychosis Screening Questionnaire (35)) or substance abuse problems, and/or 7) an inability to read and speak English. Data collection occurred between March 2014 and February 2018. The study was approved by Institutional Review Boards of The University of Tulsa, Cherokee Nation, and the Indian Health Service Oklahoma City Area Office. Participants were given an overview of all procedures and told they could withdraw at any time. All participants provided verbal and written informed consent prior to enrollment and received a $100 honorarium for the completion of each testing day (or $10/hour of non-completed days). Table 1 presents the response completion rates by race/ethnicity and sex, whereas Table 2 presents characteristics of the final sample by race/ethnicity and sex. NA participants in the current study represent tribal nations predominately from southern plains and eastern Oklahoma tribes. NA status was verified from Certificate of Degree of Indian Blood (CDIB) or tribal membership cards.

Table 1.

Study completion rate by race and sex

	NHW		NHW		NA		NA
	male		female		male		female
	N	%	N	%	N	%	N	%
Total N	77		68		50		87
Quit during day 1	3	3.9%	5	7.4%	5	10.0%	16	18.4%
Completed 1 day (or quit during day 2)	14	18.2%	10	14.7%	3	6.0%	15	17.2%
Completed both days	60	77.9%	53	77.9%	42	84.0%	56	64.4%

Note. NHW=non-Hispanic white. NA=Native American

Table 2.

Participant Characteristics by Race/Ethnicity and Sex

	non-Hispanic white						Native American
	Male			Female			Male			Female
Continuous Variable	N	M	SD	N	M	SD	N	M	SD	N	M	SD	F	η2
Age (years)	77	28.31	13.81	68	28.69	13.19	50	30.14	12.69	87	31.94	13.60	1.22	0.01
BMI (kg/m2)	75	24.73	3.59	66	23.70	3.98	49	26.20	4.21	85	25.95	4.84	4.91*	0.05
Dispositional Pain Catastrophizing (PCS; 0–52)	77	10.65	7.98	68	8.93	8.52	50	11.66	8.99	86	9.29	7.80	1.43	0.02
State Anxiety (STAI; 20–80)	76	32.25	6.72	68	32.00	8.01	50	34.10	7.88	86	33.00	7.34	0.92	0.01
Negative Affect (PANAS; 10–50)	76	2.97	3.25	68	2.79	2.92	50	3.68	3.72	86	3.09	2.71	0.84	0.01
Positive Affect (PANAS; 10–50)	76	18.92	7.09	68	17.87	7.35	50	20.08	6.27	86	18.08	8.54	1.05	0.01
Psychological Distress (SCL-90 GSI; 0–4)	77	0.33	0.29	65	0.33	0.37	48	0.47	0.49	82	0.39	0.34	2.05	0.02
General Health Scale (SF-36; 0–100)	77	64.38	10.85	65	64.34	12.48	48	61.95	10.56	83	62.47	12.05	0.75	0.01

Categorical Variable	N	%	N	%	N	%	N	%	X2	P
Marital Status (% single)	62	80.5%	46	67.6%	32	66.7%	50	57.5%	24.21	0.06
Education (% partial college)	42	55.3%	33	48.5%	24	48.0%	35	40.7%	23.50	0.17
Employment (% <40hrs/week)	35	47.3%	30	44.1%	16	33.3%	34	39.1%	5.17	0.82
Income Category									24.01	0.63
<$9,999	33	44.6%	22	32.8%	11	22.9%	25	29.8%
$10,000-$14,999	7	9.5%	9	13.4%	6	12.5%	9	10.7%
$15,000-$24,999	8	10.8%	8	11.9%	6	12.5%	13	15.5%
$25,000-$34,999	7	9.5%	5	7.5%	8	16.7%	8	9.5%
$35,000-$49,999	4	5.4%	9	13.4%	9	18.8%	11	13.1%
≥$50,000	15	20.3%	14	20.9%	8	16.7%	18	21.4%

Note: BMI = body mass index. PCS=Pain Catastrophizing Scale. STAI=State Trait Anxiety Inventory. PANAS=Positive and Negative Affect Schedule. SCL-90=Symptom Checklist 90. GSI=Global Severity Index. SF-36=Medical Outcomes Study 36 item Short Form Health Survey.

^*P<.05. Bolded tests are statistically significant.

Sample Size Determination

The parent study was powered to detect racial/ethnic differences in pain tolerance, temporal summation of the nociceptive flexion reflex, and conditioned pain modulation. That power analyses suggested 120 per group would provide power=.80 at alpha=.05, so that was the targeted sample. Based on this sample size, a sensitivity analysis was conducted in G*Power (version 3.1.9.2) to determine what size group difference (i.e., effect size) could be detected (by rejecting the null hypothesis) in the present study. That analysis indicated that, with alpha level set to .05, group differences with d≥.36 would be detected 80% of the time (i.e., power=.80), whereas with an alpha level set to .01, group differences with d≥.44 would be detected 80% of the time (i.e., power=.80). Thus, our targeted sample size provided adequate power to detect a small-to-medium effect size.

General Overview of Procedures/Testing

Testing was conducted over a 2-day period, each lasting 4–6 hours. Informed consent and inclusion/exclusion screening were conducted on the first day. If found to be eligible, participants filled out several questionnaires to assess background characteristics in order to ensure that groups did not differ on important variables (e.g., dispositional pain catastrophizing, state affect, general health perceptions, psychological functioning). On one of the testing days, electric tolerance, heat tolerance, mechanical pain threshold, and the Pain45 heat series were administered in a testing block (with a minimum of a 2-min break between tasks). There was a 10-min break after the first testing block and then cold tolerance and ischemia tolerance were administered in a second testing block (with 20-min break between tasks). The mandatory breaks were employed between tasks to minimize carryover effects. On the other day, suprathreshold electric pain, suprathreshold electric pain series, and the 52°C heat series were administered in a block of tests (with a minimum of a 2-min break between tasks). On both testing days, pain tasks were randomized within each block.

Immediately following each pain task, the McGill Pain Questionnaire-Short Form, pain-related anxiety questionnaire, and the Pain Catastrophizing Scale (using situation-specific instructions) were administered with instructions to rate their experience of the preceding pain task. The order of the two testing days was counterbalanced across participants, blocking for sex and race.

Apparatus

Questionnaire presentation was controlled by a computer with dual monitor capacity. Custom built LabVIEW software (National Instruments, Austin, TX) was used to control timing and order of the experimental protocol. One computer monitor was used by the experimenter to monitor experimental timing, whereas the second monitor was used by the participant to complete electronic questionnaires and to make ratings of stimuli. In order to reduce experimenter influences on testing, all reported experimental outcomes (except pressure pain) were conducted with participants in a sound attenuated and electrically shielded testing room and experimenters monitored the testing from an adjacent control room via a video camera (with a microphone) connected to a flat panel monitor. Participants wore a pair of sound attenuating headphones that allowed them to hear the experimenter and the computer-recorded instructions for each task.

Background Questionnaires

These questionnaires were administered to assess eligibility (e.g., Demographic and Health Status) and to examine whether groups differed on important background characteristics (e.g., mood, anxiety, psychological functioning, health, pain catastrophizing).

Demographics and Health Exclusion.

A custom-built demographic and health status questionnaire was used to obtain standard background information, as well as information regarding health problems. It was administered immediately following informed consent. The questionnaire asked about demographic information such as: age, sex, marital status, education level, employment, and income level, as well as potential exclusionary criteria such as cardiovascular problems, neurological problems, chronic pain, and medication use. Weight and height were assessed from a medical scale to calculate body mass index (BMI).

Current Affect.

Positive and negative affect prior to testing were assessed from the Positive and Negative Affect Schedule (PANAS)⁸⁸. Each positive and negative affect subscale consists of 10 items that measure positive and negative emotions, with subscales ranging from 10 to 50. Higher scores on each scale indicate greater positive or negative affect. In the current study, Chronbach’s alpha (reliability) for the negative affect scale was .76 for NHWs and .73 for NAs, and for the positive affect scale it was .88 for NHWs and .88 for NAs.

State Anxiety.

State anxiety prior to testing was assessed by the 20-item state anxiety subscale of the State Trait Anxiety Inventory (STAI)⁶⁹. The subscale ranges from 20–80, with higher scores indicating greater current anxiety. In the current study, Chronbach’s alpha (reliability) was .89 for NHWs and .88 for NAs.

Psychological Functioning.

The Symptom Checklist-90-Revised (SCL-90-R) was used to assess general psychological functioning¹⁶. The scale consists of 90 items that assess various psychological symptoms (e.g., somatization, obsessive-compulsive, depression, phobic anxiety, paranoia). The Global Severity Index (GSI) of the SCL-90-R was used to assess overall psychological distress. Total GSI scores range from 0 to 4 with higher scores indicating greater distress. In the current study, Chronbach’s alpha (reliability) was .97 for NHWs and .97 for NAs.

Health Perceptions.

The 5-item General Health subscale of the Medical Outcomes Study 36-item Short Form Health Survey (SF-36)⁸⁷ was used to determine whether groups differed in their perceptions of their health. Example items from this subscale are: “I seem to get sick a little easier than other people,” “I expect my health to get worse,” and “My health is excellent.” Scores are standardized to range from 0 to 100, with higher scores indicating better health. In the current study, Chronbach’s alpha (reliability) of the General Health subscale was .57 for NHWs and .56 for NAs.

Traditional (Dispositional) Pain Catastrophizing.

The 13-item Pain Catastrophizing Scale (PCS)⁷⁶ was used to assess dispositional catastrophizing at the beginning of the testing session to assess how each participant generally reacts to painful events. This was done to establish whether groups generally differed in pain catastrophizing. Scores ranged from 0 to 52, with higher scores indicating greater catastrophic thoughts. In the current study, Chronbach’s alpha (reliability) was .93 for NHWs and .93 for NAs.

Sensory, Affective, and Catastrophizing Reactions

The current study administered questionnaires to determine group differences in sensory (MPQ – Sensory), affective (MPQ – Affective, and Pain-Related Anxiety), and cognitive-emotional (Situation Specific Catastrophizing,) responses to pain tasks.

McGill Pain Questionnaire – Short Form (MPQ-SF).

Immediately following each painful task, the MPQ-SF⁴² was administered to assess sensory and affective reactions. The MPQ-SF consists of 15 descriptors (11 sensory, 4 affective) that describe the quality of pain. The sensory subscale includes throbbing, shooting, stabbing, sharp, cramping, gnawing, hot-burning, aching, heavy, tender, and splitting. The affective subscale includes tiring-exhausting, sickening, fearful, and punishing-cruel. Subscales are summed creating total scores that range from 0–33 (sensory) and 0–12 (affective). The MPQ visual analog scale (VAS) was also used to assess pain intensity (a sensory reaction). In the current study, Chronbach’s alpha (reliability) for the MPQ-SF sensory subscale was .76 for NHWs and .84 for NAs and for the MPQ-SF affective subscale it was .73 for NHWs and .81 for NAs.

Pain-related Anxiety.

Immediately following each painful task, a computer-presented visual analog scale (VAS) was administered to assess pain-related anxiety⁵⁷. Instructions asked the participant to rate the level of anxiety they experienced as a result of the task. Anchors of the VAS ranged from “not at all anxious” to “extremely anxious.” The computer converted the scores so that they ranged from 0 to 100, with higher scores indicating greater pain-related anxiety. These scores were used as a measure of pain-related affect.

Situation-Specific Pain Catastrophizing.

The Pain Catastrophizing Scale (PCS)⁷⁶ was also used to assess situation-specific (SS) pain catastrophizing. To assess SS pain catastrophizing, the SS-PCS was administered immediately following each painful task with altered instructions asking the participant to report the degree of catastrophizing they engaged in during the task. Scores ranged from 0 to 52, with higher scores indicating greater catastrophic thoughts. SS pain catastrophizing was used as a cognitive-emotional reaction to pain. In the current study, Chronbach’s alpha (reliability) was .95 for NHWs and .96 for NAs.

Exposure to Noxious Stimuli

To assess reactions to pain, participants underwent several suprathreshold pain tasks that involved electric, heat, cold, ischemic, and mechanical pressure stimuli. Following each pain task, participants were administered the MPQ-SF, PCS (situation-specific instructions), and the pain-related anxiety VAS (order randomized).

Electric Pain Tasks

Three tasks were used to assess responses to suprathreshold electric pain. To assess these responses, a bipolar electrode (Nicolet; 30 mm inter-electrode distance) filled with a conductive gel (EC60, Grass Technologies) was placed over the retromalleolar surface of the left ankle after the skin had been cleaned with alcohol and abraded with an exfoliating cream (NuPrep, Weaver and Company, Aurora, CO). Stimulations were delivered by an isolated, constant current stimulator (Digitimer DS7A; Hertfordshire, England). Each stimulus consisted of a train of five 1 ms rectangular wave pulses with a 3 ms inter-pulse interval (250 Hz); however, the train was always experienced as a single stimulation. The maximum stimulation intensity was set at 50-mA to ensure safety.

Electric Pain Tolerance.

Similar to previous studies⁵⁶, electric pain tolerance was assessed using a single ascending staircase of stimulations that started at 0-mA and increased in 2-mA steps. After each stimulus, the participant rated their pain intensity on a VAS that ranged from “no pain” to “maximum tolerable pain.” The stimulation intensity was increased until the participant rated the stimulus as the maximum tolerable pain. Afterwards, participants rated their reactions using the MPQ-SF, pain-related anxiety VAS, and SS-PCS.

Suprathreshold Electric Stimuli.

In the parent study, single suprathreshold electric stimuli were delivered to assess the size of individual nociceptive flexion reflexes (NFR), but the current analyses focus only on subjective reactions to these stimuli. First, validated procedures were used to determine the stimulus intensity that reliably elicits NFRs⁷⁹ (for detailed procedures,⁸⁰). In brief, stimulation intensity was set at 1.2X the highest of stimulation intensity resulting from: (1) NFR threshold^25,55 or (2) 3-stimulation threshold⁷⁹. NFR threshold assessment involved the delivery of stimulations starting at 0-mA and increased in 2-mA steps (8–12 s interstimulus interval, ISI) until an NFR was evoked. The intensity was decreased in 1-mA steps until an NFR was no longer present. This ascending-descending staircase procedure was repeated 2 more times in 1-mA steps until the NFR appeared and disappeared 2 more times. 3-stimulation threshold involved the delivery of a train of 3 electric stimulations (0.5 s ISI). The intensity of the train was increased in 2-mA steps (8–12 s ISI between trains) until the third stimulus in the train elicited an NFR.

Once the suprathreshold stimulus intensity was determined by the procedures above, participants received 5 individual suprathreshold electric stimuli (8–12 s ISI) and then rated their reactions using the MPQ-SF, pain-related anxiety VAS, and SS-PCS.

Suprathreshold Electric Series.

The current analyses focused on the subjective reactions to this task. However, in the parent study, these stimuli were used to assess temporal summation of the nociceptive flexion reflex. Stimulation intensity for this task was the same as the suprathreshold electric stimuli noted above. This task involved the delivery of 5 series (trains) of 3 suprathreshold stimulations with an ISI = 0.5 s. There was an 8–12 s break between trains. After all 5 trains were delivered, participants rated their reactions using the MPQ-SF, pain-related anxiety VAS, and SS-PCS.

Heat Tasks

Heat stimuli were generated using a Medoc (Haifa, Israel) Pathway device with a Contact Heat Evoked Potential Stimulator (CHEPS) thermode. The maximum intensity of any heat stimulus was set to 52°C (51°C for heat tolerance). Three different tasks were used to assess responsivity to painful heat. For all 3 tasks, the thermode was attached to the participants’ left volar forearm using a Velcro strap.

Heat Pain Tolerance.

Similar to prior studies⁹, heat pain tolerance was assessed 4 times after an initial practice trial. Each trial started from a baseline of 32°C and heated at a rate of 0.5°C/s. Participants were told to terminate the stimulus by pushing a button as soon as the heat became intolerable. In between trials, the thermode was moved slightly to avoid sensitization and then there was a 25–35 s randomly determined inter-trial interval. Heat pain tolerance was defined as the average of the 4 trials. Immediately after the 4^th trial, participants rated their reactions using the MPQ-SF, pain-related anxiety VAS, and SS-PCS.

Pain45 Heat Series.

In the parent study, 5 series of suprathreshold heat stimuli were delivered to assess temporal summation of heat, but the current analyses focus only on the global sensory, affective, and catastrophizing reactions to this task. This task involved 5 series of 10 heat stimuli delivered with a 2.5 s ISI⁷⁰. Each pulse peaked at an individually calibrated temperature corresponding to a rating of 45 on a numerical rating scale with the following labels⁷⁰: 10=warm; 20=a barely painful sensation; 30=very weak pain; 40=weak pain; 50=moderate pain; 60=slightly strong pain; 70=strong pain; 80=very strong pain; 90=nearly intolerable pain; 100=intolerable pain. The interval between each 10-pulse series was always 2 min. After the delivery of the 5^th series, participants rated their reactions using the MPQ-SF, pain-related anxiety VAS, and SS-PCS.

52°C Heat Series.

In the parent study, 6 blocks of 5 heat pulses (8–12 s ISI) were delivered to assess contact heat evoked potentials from EEG²⁸. However, the current analyses focused only on the subjective reactions to this task. Each pulse started at a baseline of 32°C and increased to 52°C at a rate of 70°C/s and then returned to baseline at a rate of 40°C/s²⁸. Each 65 s block of 5 pulses was separated by a short (~1-min) break. After the 5^th block of pulses, participants rated their reactions using the MPQSF, pain-related anxiety VAS, and SS-PCS.

Cold Pressor Tolerance

A cold pressor procedure^9,40,54 was used to assess responses to cold pain. Participants were asked to submerge their hand and forearm into a circulating water bath (Thermo Fisher Scientific, Pittsburgh, PA) held at 6±0.1°C. Participants were instructed to keep their fingers spread apart and to place their hand on the bottom of the water tank and keep it there for as long as they could tolerate it. The water level was kept constant (6” deep) across all participants to keep procedures standardized and maintain similar cold exposure to participants’ hands/forearms. During the task, a computer timed the length of the hand/arm immersion. The maximum cold water exposure was set to 5 min, but the participant was not informed of the limit. Immediately after tolerance was reached, participants rated their reactions using the MPQ-SF, pain-related anxiety VAS, and SS-PCS.

Ischemia Pain Tolerance

To measure responses to ischemia pain, a standard forearm tourniquet test was employed²⁴. First, participants used their non-dominant hand to conduct hand exercises with a dynamometer (Lafayette Hand Dynamometer, Lafayette Instrument Company, IN) at 50% grip strength for 2 min (1x/sec). Immediately after the last exercise, the non-dominant arm was raised for 15 s to allow the blood to drain from the forearm, then a blood pressure cuff was inflated to 220 mm/Hg around the non-dominant biceps to occlude blood flow to the forearm. Participants were instructed to keep the non-dominant arm still during the task and indicate by computer when they could no longer tolerate the ischemic pain. During the task, the computer timed the length of the arm occlusion until tolerance was reached. The maximum exposure to ischemic pain was set to 25 min, but the participant was not made aware of this limit. Immediately after tolerance was reached, participants rated their reactions using the MPQ-SF, pain-related anxiety VAS, and SS-PCS.

Pressure Pain Threshold

To measure responses to mechanical pressure, a Medoc-Wagner computerized algometer was used (Haifa, Israel). Similar to previous studies⁵³, pressure pain threshold was assessed at 3 body sites (masseter muscle, trapezius muscle, thumbnail) in random order. Pressure was applied at a rate of 98 kPa/s until the participant reached pain threshold by pressing a button. Pressure pain threshold was defined as the average of the 3 trials. For parsimony, only responses to the masseter muscle were used because this site elicited greater pain (and lower thresholds) than the other two sites. Immediately after the last pressure pain trial, participants rated their reactions using the MPQ-SF, pain-related anxiety VAS, and SS-PCS.

Data Analysis

Prior to analyses, variable distributions were screened using boxplots, histograms, and normality statistics. Those that were skewed were log or square transformed to reduce positive and negative skew, respectively. If a variable was transformed, this is noted in the units next to the variable name in the tables. Reverse transformed variables are presented and labeled when appropriate. Next, outliers were identified using Wilcox’s MAD-median procedure (using the recommended 2.24 cutoff) and then winsorized by replacing the outlier value with the next nearest non-outlier value⁹⁰. The following variables were windsorized as a result of outliers: 1) heat tolerance (17 cases) and pressure threshold (20 cases), 2) MPQ-SF sensory ratings for electric tolerance (13 cases), suprathreshold electric stimuli (19 cases), suprathreshold electric series (9 cases), heat tolerance (17 cases), Pain45 heat series (14 cases), 52°C series (5 cases), cold tolerance (14 cases), ischemia tolerance (22 cases), pressure threshold (22 cases), and electric tolerance (13 cases), 3) MPQ-SF VAS for 52°C heat series (6 cases), pressure threshold (19 cases), suprathreshold electric stimuli (4 cases), heat tolerance (12 cases), and ischemia tolerance (6 cases), 4) Pain-related anxiety for heat tolerance (20 cases), Pain45 heat series (21 cases), and ischemia tolerance (39 cases), and 5) SS pain catastrophizing for suprathreshold electric series (7 cases), heat tolerance (9 cases), and cold tolerance (26 cases).

Background Characteristics.

Preliminary analyses determined that there were significant group differences in sex (i.e., there were more women in the NA group), therefore to examine group differences on background characteristics a new independent variable (IV) was created that coded for race and sex in the same variable (i.e., a grouping variable that coded for: NHW men, NHW women, NA men, NA women). This IV was then used in 1-way ANOVAs (for continuous dependent variables [DV]) and chi-square analyses (for categorical DVs) to explore group differences. Significance level for group differences in background characteristics were set to α<.05 (2-tailed).

Primary Analyses.

Analyses for stimulus intensity (e.g., pain tolerance) outcomes were conducted using 2-way factorial ANOVAs with race (NHW vs. NA) and sex (men vs. women) as IVs. Significance level for group differences in stimulus intensity were set to α<.05 (2-tailed). For analyses of sensory, affective, and SS catastrophizing outcomes, ANCOVAs were used to control for the fact that stimulus intensity was individually calibrated to the participant (except for 52°C heat series where the stimulus was constant, so ANOVAs were used). Significance levels for group differences for primary analyses used a Bonferroni adjusted alpha of .01 that accounted for the 5 outcomes (intensity, sensory, affect, anxiety, SS catastrophizing) measured in response to each stimulus (.05 / 5 = .01). However, because this was the first study of its kind we also note when results would have been significant at an alpha level of .05.

Results

Final Sample

338 persons consented to participate, but 56 did not meet inclusion criteria; thus, 282 persons were enrolled (Table 1). Of those, 75% (N=211) completed both days of testing. 15% (N=42) completed only one day and 10% (N=29) quit during the first testing day. Given this, Ns across analyses differed and will be reported. Non-completers were more likely to be NA women (P=.045) and have a lower education level (P=.015); however, there were no differences on age, BMI, marital status, positive affect, negative affect, state anxiety, psychological functioning, or pain catastrophizing (all Ps>.18).

Background Characteristics

Chi-square analysis indicated that groups differed by sex, χ²(df=1)=7.85, P=.005. 64% of the NA group were women, whereas only 47% of the NHW group were women. Thus, all background characteristic analyses used a 4 level IV that coded for race/ethnicity and sex. Inferential statistics, as well as means and SDs (or N’s and %’s), for these analyses are reported in Table 2. As shown, groups were similar on most variables except for BMI. NHW women had a significantly lower BMI than NA men and NA women (Ps<.005).

Ethnic/Racial Differences in Stimulus Intensity/Stimulus Exposure

Prior to examining group differences in sensory, affective, and catastrophizing reactions, stimulus intensity/duration was analyzed. As shown in Table 3, there were significant group differences in suprathreshold electric stimulus intensity (P=.010) and cold pressor pain tolerance (P=.015). The average intensity of the electric stimuli was higher for the NA group than the NHW group. Cold pressor tolerances were lower in the NA group than the NHW group (P=.015), but this was qualified by a Group X Sex interaction (P=.032) that indicated NA women had lower pain tolerances than NHW women (M=36.84 sec, SD=1.74 vs. M=62.29 sec, SD=2.76; P=.001; means/SD reversed transformed), whereas NA men and NHW men did not differ (M=63.73 sec, SD=2.46 vs. M=65.98 sec, SD=2.56; P=.834; means/SD reversed transformed).

Table 3.

Group differences in painful stimulus intensity/duration

	non-Hispanic white			Native American			Cohen’s	Group	Sex	GxS
Stimulus Variable	N	M	SD	N	M	SD	d	F	F	F
Electric Tolerance Stimulus (mA)	126	31.01	13.10	111	30.82	13.17	0.01	0.01	0.68	0.05
Suprathreshold Electric Intensity (mA)	131	21.70	10.02	124	25.02	10.05	−0.33	6.73*	0.01	0.03
Heat Pain Tolerance Stimulus (°C)	126	45.89	2.07	111	45.52	1.66	0.19	2.44	30.98*	0.03
Heat Pain45 Series Intensity (°C)	126	47.34	1.79	112	47.30	1.85	0.02	0.03	0.59	1.05
Cold Pressor Pain Tolerance Stimulus (Log[sec+1])	127	1.81	0.42	110	1.69	0.33	0.32	6.04*	7.07*	4.64*
Ischemia Pain Tolerance Stimulus (Log[sec+1])	126	2.21	0.44	111	2.11	0.43	0.23	3.21	5.56*	1.16
Pressure Pain Threshold (kPa)	127	162.24	61.32	111	155.31	57.49	0.12	0.79	2.60	1.14

Note. G × S = Group x Sex interaction.

^*P<.05. Bolded F-tests are statistically significant. Cohen’s d is reported for the effect size for the group mean comparisons.

Sex Differences in Stimulus Intensity/Stimulus Exposure

Sex differences were also found for heat, ischemia, and cold pain tolerance. Compared to men, women had lower heat tolerances (M=45.05°C, SD=1.62 vs. M=46.36°C, SD=1.95; P<.001; d=0.77), lower ischemia tolerances (M= 125.20 sec, SD= 2.77 vs, M= 170.15 sec, SD= 2.63; P=.019; d=0.33; means/SD reversed transformed), and lower cold pressor tolerances (M=47.90 sec, SD=2.33 vs. M=64.84 sec, SD=2.51; P=.008; d=0.36; means/SD reversed transformed). This last effect was moderated by race/ethnicity as described in the previous paragraph. There were no group or sex differences in electric pain tolerance, heat Pain45 intensity, or pressure pain threshold.

Ethnic/Racial Differences in Subjective Reactions to Painful Stimuli

Table 4 reports Ms, SDs, Cohen’s d, and inferential statistics for pain intensity ratings, sensory ratings, affective ratings, pain-related anxiety, and situation-specific catastrophizing in response to the various painful stimulus modalities. Figure 1 depicts these group differences after using reverse operations (e.g., 10^X, square root) to put transformed variables back into their original units. Results indicated that, compared to NHW participants, NAs reported: 1) higher pain intensity in response to Pain45 stimuli (P=.041), 2) higher MPQ-SF sensory ratings in response to suprathreshold electric stimulations (P=.008), suprathreshold electric series (P=.004), and Pain45 stimuli (P=.009); 3) higher MPQ-SF affective ratings to the Pain45 heat series (P=.017); 4) higher pain-related anxiety in response to suprathreshold electric stimuli (P=.025), suprathreshold electric series (P=.047), heat pain tolerance stimuli (P=.003), Pain45 heat series (P=.004), cold pressor tolerance stimuli (P=.030), and ischemia tolerance stimuli (P=.002); and 5) greater situation-specific catastrophizing in response to suprathreshold electric stimuli (P=.033), suprathreshold electric series (P=.008), and heat tolerance stimuli (P=.004).

Table 4.

Sensory, affective, and catastrophizing reactions to pain tasks

		ANCOVA F-tests			non-Hispanic white			Native American			Cohen’s
Stimulus Type	Reactivity Variable	Group	Sex	G x S	M	SD	N	M	SD	N	d
Electric Tolerance	Pain Intensity (squared MPQ VAS)	0.74	0.21	2.27	4312.0 4	2672.82	12 6	4616.2 7	2731.7 3	11 1	–0.11
	Sensory (MPQ; 0–33)	0.15	0.11	0.27	10.97	4.70	12 6	11.24	5.73	11 1	–0.05
	Affect (Log[MPQ+1])	0.15	2.99	0.17	0.45	0.32	12 6	0.43	0.33	11 1	0.05
	Anxiety (squared VAS)	3.04	0.52	0.16	4185.7 7	2967.82	12 6	4892.8 3	3254.7 2	11 1	–0.23
	SS Catastrophizing (Log[PCS+1])	0.96	<.01	0.21	0.92	0.48	12 6	0.98	0.47	11 1	–0.13
Suprathr Electric	Pain Intensity (MPQ VAS; 0– 100)	2.89	0.02	0.26	38.54	24.68	12 9	44.024	26.41	11 4	–0.22
	Sensory (MPQ; 0–33)	7.11*	1.17	2.83	7.57	4.83	12 9	9.33	5.67	11 4	–0.34
	Affect (Log[MPQ+1])	1.37	0.77	3.53	0.29	0.33	12 9	0.34	0.36	11 4	–0.15
	Anxiety (0–100)	5.06#	0.20	1.07	39.74	29.24	12 9	48.56	31.62	11 4	–0.29
	SS Catastrophizing (Log[PCS+1])	4.61#	0.00	1.61	0.70	0.50	12 9	0.84	0.50	11 4	–0.28
Suprathr Electric Series	Pain Intensity (MPQ VAS; 0– 100)	3.29	0.04	<.01	51.78	26.90	12 8	58.02	27.02	11 2	–0.23
	Sensory (MPQ; 0–33)	8.34*	0.12	1.30	10.41	5.81	12 7	12.74	6.92	11 2	–0.37
	Affect (Log[MPQ+1])	2.12	2.09	2.72	0.38	0.35	12 8	0.45	0.38	11 2	–0.19
	Anxiety (squared)	3.98#	1.49	2.09	3600.4 8	3088.85	12 8	4447.6 8	3569.5 5	11 2	–0.26
	SS Catastrophizing (Log[PCS+1])	7.23*	0.97	0.89	0.90	0.49	12 7	1.05	0.42	11 1	–0.34
Heat Tolerance	Pain Intensity (squared MPQ VAS)	0.67	8.63*	1.33	4436.9 3	2404.65	12 5	4718.1 8	2859.8 5	11 0	–0.11
	Sensory (MPQ; 0–33)	1.15	4.87*	2.21	9.70	4.89	12 5	10.45	5.76	11 0	–0.14
	Affect (Log[MPQ+1])	1.11	4.60*	0.20	0.25	0.29	12 5	0.29	0.31	11 0	–0.14
	Anxiety (Log[Anx+1])	8.98*	6.56*	3.41	1.28	0.49	12 5	1.47	0.46	11 0	–0.40
	SS Catastrophizing	8.66*	9.62*	0.57	0.74	0.50	12	0.92	0.43	11	–0.38
	(Log[PCS+1])						5			0
Pain45 Heat Series	Pain Intensity (MPQ VAS; 0–100)	4.22#	4.12*	0.07	42.58	19.07	12 6	47.88	20.66	11 2	–0.27
	Sensory (MPQ; 0–33)	6.76*	0.77	3.11	7.81	3.97	12 6	9.41	5.42	11 2	–0.34
	Affect (Log[MPQ+1])	5.77#	0.39	0.54	0.17	0.23	12 6	0.25	0.31	11 2	–0.31
	Anxiety (Log[Anx+1])	8.61*	0.01	0.82	1.25	0.50	12 6	1.43	0.42	11 2	–0.37
	SS Catastrophizing (Log[PCS+1])	3.01	1.62	0.16	0.59	0.44	12 6	0.69	0.48	11 2	–0.23
52°C Heat Series§	Pain Intensity (Log[MPQ VAS+1])	1.13	1.39	0.07	1.17	0.42	13 0	1.23	0.45	11 2	–0.14
	Sensory (Log[MPQ+1])	0.75	1.56	<.01	0.71	0.27	13 0	0.74	0.30	11 2	–0.11
	Affect (Log[MPQ+1])	0.75	3.02	0.25	0.13	0.23	13 0	0.16	0.26	11 2	–0.11
	Anxiety (Log[Anx+1])	0.89	0.17	0.11	1.02	0.58	13 0	1.09	0.58	11 3	–0.12
	SS Catastrophizing (Log[PCS+1])	0.89	0.29	0.02	0.43	0.43	13 0	0.48	0.48	11 2	–0.12
Cold Pressor Tol	Pain Intensity (squared MPQ VAS)	1.38	7.00*	0.70	4875.1 1	2653.93	12 6	5272.3 2	2936.1 2	10 9	–0.14
	Sensory (MPQ; 0–33)	2.75	1.83	0.89	9.70	5.58	12 6	10.92	5.70	10 9	–0.22
	Affect (Log[MPQ+1])	<.01	2.90	1.10	0.31	0.32	12 6	0.31	0.32	10 9	0.00
	Anxiety (0–100)	4.75#	1.60	0.89	39.34	29.29	12 6	47.83	33.31	10 9	–0.27
	SS Catastrophizing (Log[PCS+1])	3.76	6.73*	4.07*	0.99	0.46	12 6	1.10	0.40	10 9	–0.24
Ischemia Tolerance	Pain Intensity (MPQ VAS; 0– 100)	1.71	13.34*	<.01	56.89	22.75	12 6	60.70	23.14	11 0	–0.17
	Sensory (MPQ; 0–33)	2.96	6.04*	0.73	9.65	4.91	12 6	10.92	6.59	11 0	–0.22
	Affect (Log[MPQ+1])	0.17	7.36*	0.10	0.43	0.29	12 6	0.41	0.32	11 0	0.05
	Anxiety (Log[Anx+1])	9.72*	6.23*	0.25	1.38	0.51	12 6	1.55	0.34	11 0	–0.39
	SS Catastrophizing (Log[PCS+1])	0.02	4.09*	0.38	0.90	0.51	12 6	0.91	0.50	11 0	–0.02
Pressure Threshold	Pain Intensity (Log[MPQ VAS+1])	0.28	0.33	0.52	1.21	0.31	12 5	1.18	0.39	11 1	0.09
	Sensory (MPQ; 0–33)	0.19	3.43	0.05	4.98	3.05	12 5	4.80	3.21	11 1	0.06
	Affect (Log[MPQ+1])	0.34	2.09	0.11	0.08	0.17	12 5	0.09	0.20	11 1	–0.08
	Anxiety (Log[Anx+1])	3.80	0.04	0.22	0.69	0.57	12 5	0.84	0.63	11 1	–0.26
	SS Catastrophizing (Log[PCS+1])	0.82	2.27	0.52	0.26	0.37	12 5	0.31	0.38	11 1	–0.12

Note. MPQ=McGill Pain Questionnaire – Short Form. SS=situation-specific. PCS=Pain Catastrophizing Scale. Cohen’s d is reported for the effect size for the group mean comparisons.

^#P<.05

^*P<.01. Bolded F-tests are statistically significant.

^§analysis of this variable did not include stimulus intensity as a covariate because the intensity was the same for all participants.

Figure 1.

Racial/ethnic differences in pain intensity ratings, sensory ratings, affective ratings, pain-related anxiety, and situation-specific (SS) catastrophizing in response to electric tolerance testing (Elect Tol), suprathreshold electric stimulations (Supra Elect), 3 stimulation series of suprathreshold electric stimulations (Elect Series), heat tolerance testing (Heat Tol), 10-pulses series of suprathreshold heat stimuli (Pain45 series), series of 52°C heat stimuli (52°C series), cold pressor tolerance testing (Cold Tol), ischemia tolerance testing (Isch Tol), and pressure threshold testing (Pressure Thr). For variables that were transformed (e.g., log, square) to address non-normality, the reverse operation was conducted on the means (e.g., 10^X, square root) in order to display them in the original units. MPQ=McGill Pain Questionnaire-Short Form. ^#Racial/ethnic group differences at P<.05. *Racial/ethnic group differences at P<.01.

Sex Differences in Subjective Reactions to Painful Stimuli

Also, as noted in Table 4, sex differences were also found. Figure 2 depicts these results after using reverse operations (e.g., 10^X, square root) to put transformed variables back into their original units. Results indicated that, compare to women, men reported: 1) more pain intensity in response to heat tolerance stimuli (P=.004, d=.41), Pain45 series (P=.044, d=.26), cold tolerance stimuli (P=.009, d=.32), and ischemia tolerance stimuli (P=.0003, d=.47), 2) higher MPQ-SF sensory ratings in response to heat tolerance stimuli (P=.028; d=0.31) and ischemia tolerance stimuli (P=.015; d=0.32), 3) higher MPQSF affective ratings in response to heat tolerance stimuli (P=.033; d=0.30) and ischemia tolerance stimuli (P=.007; d=0.36), 4) higher pain related anxiety in response to heat tolerance stimuli (P=.011; d=0.35) and ischemia tolerance stimuli (P=.013; d=0.30); and 5) higher situation-specific catastrophizing in response to heat tolerance stimuli (P=.002; d=0.43), cold tolerance stimuli (P=.010, d=.23), and ischemia tolerance stimuli (P=.044; d=0.26). The only significant Group X Sex interaction was noted for log situation-specific catastrophizing during cold pressor (P=.045). NA women reported greater situation-specific catastrophizing than NHW women (M= 12.02, SD= 2.34 vs. M= 7.40, SD= 3.21; P=.005; d=0.63, means reverse transformed to ease interpretation), whereas there was no difference in situation-specific catastrophizing between NA men and NHW men (M= 12.94, SD= 2.79 vs. M= 13.06, SD= 2.45; P=.956; d=.01, means reverse transformed to ease interpretation).

Figure 2.

Sex differences in pain intensity ratings, sensory ratings, affective ratings, pain-related anxiety, and situation-specific (SS) catastrophizing in response to electric tolerance testing (Elect Tol), suprathreshold electric stimulations (Supra Elect), 3 stimulation series of suprathreshold electric stimulations (Elect Series), heat tolerance testing (Heat Tol), 10-pulse series of suprathreshold heat stimuli (Pain45 series), series of 52°C heat stimuli (52°C series), cold pressor tolerance testing (Cold Tol), ischemia tolerance testing (Isch Tol), and pressure threshold testing (Pressure Thr). For variables that were transformed (e.g., log, square) to address non-normality, the reverse operation was conducted on the means (e.g., 10^X, square root) in order to display them in the original units. MPQ=McGill Pain Questionnaire-Short Form. ^#Sex differences at P<.05. *Sex differences at P<.01.

Discussion

This study used QST to examine racial/ethnic group differences in reactions to painful stimuli that might contribute to chronic pain risk in NAs^4,35,83. To achieve this, we measured pain intensity ratings, sensory ratings, affective ratings, pain-related anxiety, and situation-specific catastrophizing (SS-catastrophizing) following suprathreshold exposure to electric, heat, cold, ischemic, and pressure stimuli.

Race/Ethnic Differences in Pain Reactions

Sensory differences were noted between NAs and NHWs, but were restricted to electric stimuli and Pain45 (heat) series. In all cases, NAs reported higher pain intensity (although non-significant at the P<.01 level) and higher MPQ-SF sensory ratings than NHWs. Thus, these electric and heat stimuli evoked a stronger sensory experience in NAs. Interestingly, all of these stimuli were delivered in a series. 2 of the 3 stimuli (electric series, Pain45 series) were designed to evoke temporal summation of pain (i.e., had inter-stimulus intervals ≤3 s), a phenomenon associated with temporary sensitization of spinal neurons^2,22,72. Therefore, these racial/ethnic differences could reflect a heightened sensory experience in NAs resulting from amplification of spinal nociception. However, this cannot fully explain the group differences because NAs also had higher sensory ratings in response to the 5 suprathreshold electric stimuli whose inter-stimulus intervals were too long to evoke temporal summation. Moreover, we have not observed racial/ethnic differences in temporal summation of pain itself (⁷⁴; data to be reported elsewhere). Because sensitized spinal pathways assessed by temporal summation are not likely to explain the results, it would be interesting to determine whether these heightened sensory experiences correspond to pain signaling differences within the primary or secondary somatosensory cortices^43,86, or whether they reflect differences in the interpretation of the pain signal. Whatever the mechanism, these data suggest differences in the sensory experience of NAs that occurs during suprathreshold electric and heat stimuli.

There was only one group difference in MPQ-SF affective ratings and it occurred in response to the Pain45 heat series, but it did not survive after adjusting for familywise Type I error. Some of the most consistent differences were noted for pain-related anxiety. A close examination of Fig 1 finds that NAs reported more anxiety to all stimuli, but only heat tolerance, Pain45, and cold tolerance survived familywise type I error adjustment. Some of the non-significant group differences may be due to ceiling and floor effects because stimuli evoking moderate anxiety showed the strongest group differences. Given that anxiety is known to enhance pain^3,39,49,59, this could be a significant risk factor for NAs.

NAs also reported greater SS-catastrophizing in response to suprathreshold electric stimuli, electric series, and heat tolerance stimuli (although suprathreshold electric stimuli did not survive familywise Type I error adjustment). This is interesting given that there were no group differences in traditionally-measured (dispositional) pain catastrophizing (Table 2) and suggests that NAs experienced greater catastrophic thoughts during specific pain tasks. This implies that SS-catastrophizing may be more sensitive in detecting racial/ethnic differences. Given that prior research has found SS-catastrophizing to be a stronger predictor of pain outcomes than dispositional catastrophizing^10,12,18,58, greater SS-catastrophizing could put NAs at greater risk of pain.

Implications for Racial/Ethnic Group Differences in Pain Reactions

These differences in pain experience have several implications. First, compared to NHWs, NAs reported heightened sensory, anxiety, and catastrophizing reactions to multiple pain stimuli, suggesting amplification of the pain experience across different stimulus modalities. Notably, these experiential differences were more consistent than differences in pain thresholds/tolerances (Table 3). NAs had lower cold pain tolerances, but there were no group differences for electric/heat tolerance or pressure thresholds. Thus, pain risk in NAs may not stem from a general reduction in pain tolerance, but rather the resulting psychological impact of the noxious event (e.g., anxiety, catastrophizing). These observations are in contrast to prior (small N) studies of NAs that reported dampened pain sensitivity or reduced pain facilitation^47,48,65. So, the present study underscores the importance of replicating findings from small studies using larger, more diverse samples and indicates that pain risk in NAs may be similar to other minorities. Specifically, others have noted that ethnic/racial differences are often attributed to psychosocial variables like hypervigilance, anxiety, and/or depression.^{9,19,20,29,36,52,63,91} It is noteworthy though that we did not find group differences in general psychological status (Table 2; e.g., state anxiety, distress). Thus, future research is needed to explore cultural factors (e.g., historical trauma) that could contribute to heightened pain reactivity in NAs.

Second, the heightened pain reactivity in NAs cannot be attributed to differences in nociceptive input. Even though most stimuli (except 52°C heat series) were individually-calibrated to participants, individual differences in stimulus intensity/exposure were controlled in the analyses. Moreover, group differences cannot be explained by the unequal number of men/women across racial/ethnic groups, because sex was also controlled.

Third, the most consistent group differences across stimuli were noted for pain-related anxiety, sensory ratings, and to a lesser extent SS-catastrophizing. Both anxiety^{3,13,44,45,49,59,68,85} and catastrophizing^{10,12,50,57,73} are strong pain enhancers. Anxiety can activate descending, brain-to-spinal cord facilitation circuits^33,81 and catastrophizing can facilitate pain via supraspinal circuits^{17,26,27,57,80,81}. Given this, these factors may play a significant role in pain risk for NAs due to their sensitizing effects, perhaps actually promoting the eventual onset of chronic pain. Moreover, anxiety and catastrophizing are related to greater pain-related disability and suffering in those with pain^39,50,75; thus, they may also help maintain pain in NAs. This could ultimately form a vicious cycle, in which pain leads to anxiety/catastrophizing, that then enhances pain^{8,12,39,51,84}. Hence, treatments that target pain-related anxiety and catastrophizing may be particularly helpful for NA patients^15,34,67,82.

Sex Differences in Pain Reactions

Sex differences in pain are often observed with women showing greater experimental pain sensitivity and a higher likelihood of developing chronic pain^{5–7,23,60–62,77,89} (although effect sizes can be quite variable⁶²). Consistent with this, we found that women had lower heat pain tolerances, ischemia pain tolerances, and cold pain tolerances, but cold tolerance was moderated by race/ethnicity such that NA women were less tolerant than NA men (P=.001). Interestingly, sex differences in pain experience did not mirror these differences. Rather, men reported greater pain intensity (heat, cold, & ischemia tolerance), MPQ-SF affective pain (ischemia tolerance), and SS-catastrophizing (heat & cold tolerance). Although SS-catastrophizing during cold pain was moderated by race/ethnicity, men in both racial/ethnic groups reported greater SS-catastrophizing than women (Ps<.05).

It is unclear why men reported greater pain reactivity in our study, but it is interesting to note that higher reactivity was only found in response to stimuli that men tolerated longer. By contrast, NAs reported greater pain reactivity in response to stimuli in which their pain tolerances were lower or similar to NHWs. Together, this suggests increased pain reactivity in men may reflect a different mechanism than those observed in NAs. Indeed, the racial/ethnic differences are more likely to reflect pain amplification processes that could promote pain in NAs (i.e., a risk factor), whereas the sex differences seem to reflect that men were able to tolerate pain longer despite experiencing greater pain. Taken together, one could argue this latter observation reflects a resiliency factor for men. However, this hypothesis requires further testing.

Strengths and Limitations

This study had a number of strengths. It is currently the largest study of experimental pain in a NA sample. It also used numerous pain stimulus modalities that were pseudorandomly ordered to minimize order effects. We also assessed several measures of pain experience. Analyses also controlled for stimulus intensity and sex. Despite these strengths, a few limitations should be noted.

First, this study was conducted in healthy, pain-free individuals in order to determine whether group differences in pain processing exist that could contribute to chronic pain risk. This helps ensure that observed group differences are not due to differences in disease status or access to health care. As a result, we cannot know whether these findings will generalize to NAs experiencing chronic pain. Moreover, it meant that many of our participants were young adults; thus, results may not generalize to older individuals. Second, NA women were more likely to quit before completing all tasks. Thus, it is possible that our results will not generalize to all NA women. That said, non-completers were not systematically different than completers on other background variables. Related, our NA sample was recruited mostly from northeastern Oklahoma where NAs are not reservation-dwelling. Future studies should determine if our results generalize to NAs from other geographical regions. Third, although we varied the order of pain tasks and scheduled mandatory breaks in between most tasks, several tasks occurred each day. Therefore, it is possible that some carryover occurred, perhaps masking other racial/ethnic differences. And finally, although we adjusted our alpha level, it is possible that the numerous hypothesis tests resulted in inflation of Type I error rate.

Summary

In sum, NAs experienced greater sensory, pain-related anxiety, and catastrophizing reactions to laboratory pain. Given the potential for anxiety and catastrophic thoughts to amplify pain, these enhanced reactions in NAs may, at least partly, explain the greater prevalence of pain in this population.

Highlights.

• Native Americans (NAs) experienced greater pain reactivity than non-Hispanic whites

• Reactivity differences were noted for sensory ratings, anxiety, and catastrophizing

• This might place them at risk for pain disorders