Chronic pain among middle-aged and older adults in the United States: The role of everyday discrimination and racial/ethnic identity

Abstract

Chronic pain disproportionately affects middle-aged and older adults in the United States. Everyday discrimination is associated with worse pain outcomes and is more prevalent among adults from racial/ethnic minoritized groups. Yet, there is limited evidence on relationships between everyday discrimination and chronic pain among middle-aged and older adults, as well as how discrimination and racial/ethnic identity may interact to influence this relationship. We used the 2018 Health and Retirement study to evaluate associations between exposure to everyday discrimination and odds to experience any, severe, and high impact chronic pain among 5,314 Hispanic, non-Hispanic Black, and non-Hispanic White adults over the age of 50. Logistic regression was used to evaluate the main and interaction effects of everyday discrimination on the odds of chronic pain (any, severe, and high impact) across racial/ethnic groups. Results showed that Hispanic and non-Hispanic Black middle-aged and older adults had a higher, unadjusted prevalence of severe and high impact chronic pain and reported more exposure to everyday discrimination compared to non-Hispanic White middle-aged and older adults. In fully adjusted models, exposure to everyday discrimination predicted higher odds to experience each type of chronic pain. In addition, study findings showed that exposure to everyday discrimination significantly raised pain risk among Hispanic and non-Hispanic White, but not non-Hispanic Black, middle-aged and older adults. Findings underscore the influential role of everyday discrimination on the chronic pain experiences of middle-aged and older adults, as well as differential effects across racial/ethnic groups.

INTRODUCTION

Chronic pain disproportionately affects middle-aged and older adults in the United States (U.S.), who are also more likely to experience high impact chronic pain – or chronic pain that limits functional activities and social participation.^12,44 The chronic pain burden among aging adults is particularly concerning as it has been linked to higher levels of disability, worse quality of life, accelerated cognitive decline, and premature death.^13,31,42,43 Given the rapid projected growth of an increasingly racially and ethnically diverse older adult population, there is a substantial need for research that can identify mutable factors which enhance the health and well-being outcomes of aging adults and mitigate the high economic and societal cost of chronic pain.

There is robust evidence that pain experiences differ for people in racial and ethnic minoritized groups, and that these differences persist throughout middle and older adulthood.^{1,20, 40} For example, certain racial and ethnic minoritized groups have been found to report more severe pain and pain-related disability, particularly non-Hispanic Black adults, when compared to non-Hispanic White adults.⁴² Yet, there is conflicting evidence regarding racial/ethnic differences in pain specifically among older adults. In some studies, non-Hispanic Black and Hispanic older adults have been found to have more severe and disabling pain²⁵ , while others indicate that non-Hispanic Black older adults show similar or even lower pain burden, when compared to non-Hispanic White older adults, after controlling for socioeconomic factors.^21,33 A potential pain equivalence or advantage has also been shown among Hispanic older adults.^25,42

Research on social determinants of health has shown that between 40% to upwards of 70% of health outcomes can be attributed to social and environmental factors.^5,24 Thus, the specific influence of social factors on racial/ethnic differences in pain outcomes has been an emerging area of research, particularly the role of discrimination. While specific evidence on the influence of racial/ethnic identity on relationships between discrimination and chronic pain among middle-aged and older adults is lacking, there is prior evidence that has established an association between exposure to discrimination and an increased likelihood to experience myriad negative health outcomes^17,30,38, including chronic pain and disability more broadly.^6,7,14,15,37 In particular, there is an abundance of evidence on everyday discrimination – or the self-reporting of exposure to unfair and differential treatment on a day-to-day basis – as a key stressor within the human experience.³⁸ Indeed, repeated exposure to discrimination is thought to alter the body’s physiological response to stress through mechanisms such as higher allostatic load and concentration of inflammatory cytokines, which could contribute to the development and severity of chronic pain.^11,29 In addition, psychological mechanisms, such as perceived stress and pain catastrophizing, could explain relationships between discrimination and chronic pain outcomes.³⁷ Given these associations and prior evidence that racial and ethnic minoritized groups with pain report more exposure to discrimination³⁷, there remains a need to investigate relationships between discrimination and chronic pain among middle-aged and older adults, overall, as well as to explore interactions between discrimination and racial/ethnic identity.

The current cross-sectional study used nationally-representative data of middle-aged and older adults (aged 51 and older) in the U.S. to address two aims: 1) to characterize exposure to everyday discrimination and the prevalence of any, severe, and high impact chronic pain among middle-aged and older adults, overall, and across three racial/ethnic groups (Hispanic, non-Hispanic Black, non-Hispanic White); and 2) to investigate associations between everyday discrimination and chronic pain, including the degree to which interactions between everyday discrimination and racial/ethnic identity influence this relationship. We hypothesized that more exposure to everyday discrimination would increase the odds of experiencing each type of chronic pain among middle-aged and older adults. We also expected that there would be significant interactions between everyday discrimination and racial/ethnic identity in predicting the odds of chronic pain, particularly between non-Hispanic Black and non-Hispanic White respondents, based on prior evidence indicating that experiences of discrimination may have a stronger influence on pain among non-Hispanic Black adults.^14,15

MATERIALS AND METHODS

Overview

Data from the 2018 Health and Retirement Study (HRS) were used for the current study. HRS is an ongoing, biennial, longitudinal panel study on health and aging among a representative sample of more than 20,000 noninstitutionalized adults over 50 years of age in the U.S.²³ The HRS is sponsored by the National Institute on Aging (grant number U01AG009740) and Social Security Administration. It is conducted by the University of Michigan’s Institute for Social Research. The HRS study procedures have been approved by the University of Michigan Institutional Review Board and informed consent was obtained from each respondent. The first cohort was interviewed in 1992 and biennially thereafter. HRS oversamples NHB and HSP older adults and study participants are asked about physical and mental health, cognitive functioning, health care usage and coverage, employment, and financial information. The full details of the HRS , including information on the study sample and data collection procedures, are described elsewhere.³⁵

HRS data have been used extensively to characterize how psychosocial factors influence the health status and trajectories of older adults in the U.S. Beginning in 2006, a rotating subgroup of half of HRS respondents were selected at random to complete an enhanced face-to-face survey (EFTF), including a leave-behind questionnaire which focuses on the collection of psychosocial data such as experiences of discrimination.³⁴ The current study includes the random subsample of respondents who completed the leave-behind questionnaire in 2018 to analyze relationships between exposure to everyday discrimination and experiences of any, severe, and high impact chronic pain, as well as interactions between discrimination and racial/ethnic identity on these relationships. For 2018, there was an 8,806 random subsample of respondents who were eligible for the leave-behind questionnaire of which 5.695 were completed and returned by mail. We analyzed data from the Hispanic, non-Hispanic Black, and non-Hispanic White respondents who completed and returned the leave-behind questionnaire and were 51 years of age or older (n=5,314)

Outcome Variables

Any, severe, and high impact chronic pain.

The HRS captures self-reported pain in multiple ways. The survey first asks this question of all respondents: “Are you often troubled with pain?”. For the current study, respondents who answered “Yes” to this question were coded as having any chronic pain, an approach that has previously been used to estimate chronic pain prevalence among older adults.²¹ In addition, for respondents who answer “Yes” to this initial question, follow-up questions are asked with respect to pain severity and impact. For pain severity, respondents are asked, “How bad is the pain most of the time: mild, moderate or severe?” For the current study, respondents who indicated that their pain was ‘severe’ most of the time were coded as having severe pain. For pain impact, respondents are asked, “Does the pain make it difficult for you to do your usual activities such as household chores or work?”. Respondents who answered “Yes” to this question were coded as having high impact chronic pain.

Independent Variable

Everyday Discrimination Scale (6-item).

The HRS utilizes a shortened version of the Everyday Discrimination Scale which assesses participants’ perceptions of routine experiences of mistreatment within their lives.³⁹ HRS participants are asked how often the following six events have happened to them in their day-to-day life: 1) “You are treated with less courtesy or respect than other people”; 2) “You receive poorer service than other people at restaurants or stores”; 3) “People act as if they think you are not smart”; 4) “People act as if they are afraid of you”; 5) “You are threatened or harassed”, and 6) “You receive poorer service or treatment than other people from doctors or hospitals.” The potential reasons for mistreatment were not exclusive to race or ethnicity and could be attributed to a host of other reasons, such as the participant’s gender, age, or financial status. The response options for each item ranged from 1 (“almost everyday”) to 6 (“never”). To score the scale, items were reverse-coded, summed, and divided by six to establish a mean everyday discrimination score ranging from one to six with higher scores reflecting greater exposure to everyday discrimination. We next created a binary variable to account for the relatively low reported exposure to everyday discrimination within the sample. To do this, we coded participants whose mean score was less than two as having no exposure (‘none’) and participants whose mean scores were at least two as having ‘any’ exposure to everyday discrimination.

Demographic variables

We accounted for socio-demographic characteristics and health status in the present study. Racial/ethnic identity was categorized based on self-identification into three groups: Hispanic, non-Hispanic Black, and non-Hispanic White. Given their salience to pain, we also included age, sex, educational attainment, and employment status as covariates. We divided age into four categories: 51-59, 60-69, 70-79, or 80 years of age or older. Sex was dichotomized as either female or male. Education was divided into three categories: less than a high school education, high school graduate or equivalent, or any college. Employment status was dichotomized as either currently working part-time or full-time or not currently working. In addition, we accounted for self-rated health status based on the question, “Would you say your health is excellent, very good, good, fair, or poor?” Responses were divided into three categories: excellent to very good, good, or fair to poor.

Analytic Approach

Everyday discrimination and chronic pain prevalence estimates (any, severe, high impact) were calculated for the total sample and across each racial/ethnic category. Differences among the categorical variables were analyzed using the chi-square test (X²). We next calculated four models for each chronic pain status (any, severe, high impact) using logistic regression. The first model (Model 1) calculates the main effects of everyday discrimination and racial/ethnic identity for each outcome, adjusting for age and sex. The second model (Model 2) calculates the main effects of everyday discrimination and racial/ethnic identity for each outcome, adjusting for age, sex, educational attainment, employment status, and self-rated health. The third model (Model 3) calculates both main effects and interaction effects of everyday discrimination and racial/ethnic identity for each chronic pain status, adjusting for age and sex. The fourth model (Model 4) calculates the main effects and interaction effects of everyday discrimination and racial/ethnic identity, adjusting for age, sex, educational attainment, employment status, and self-rated health. We present the adjusted odds ratio (aOR) and standard error for each model, with an aOR higher than 1.0 indicating greater odds to experience chronic pain versus an aOR less than 1.0 indicating lower odds of chronic pain. Since there is not universal agreement on whether survey weights should be used in multivariate models due to the larger standard errors that are produced, we present both the unweighted and weighted estimates for each model.^40,41 In addition, for each model we calculated the average marginal effects of exposure to everyday discrimination on the probability for each chronic pain status across racial/ethnic groups using post estimation commands. Marginal effects allow for the examination of whether the exposure has the same or differential effects on the probability of an outcome for the groups of interest.²⁷ Finally, we stratified the sample by racial/ethnic identity to look at within-group differences using the subpop command. We calculated two models for each racial/ethnic group across each chronic pain status. The first model calculates the main effects of everyday discrimination for each chronic pain status after adjusting for age and sex. The second model incorporates the remaining covariates – educational attainment, employment status, self-rated health – to assess the main effects of everyday discrimination for each chronic pain status within each racial/ethnic group. For weighting the U.S. population, we implemented survey design packages in all models which integrate sampling weights, stratum and psu clustering to account for the random selection into the subsample and the complex design of the HRS. Statistical significance was considered at p<0.05. Stata 17.0 was used to complete all analytic procedures.

RESULTS

Shown in Table 1, the weighted sample was predominantly female (53.5%) and made up of respondents between 51 and 59 years old (29.2%) or 60 and 69 years old (36.6%). In addition, the majority of the sample had attended college (58.5%), were not currently working (58.8%), and rated their health as either ‘excellent to very good’ (43.4%) or ‘good’ (32.3%). There were differences across racial/ethnic groups. There were a significantly higher percentage of non-Hispanic White respondents in the older age categories, who attended college, and who rated their health as ‘excellent or very good’ relative to Hispanic and non-Hispanic Black respondents.

Table 1.

Descriptive statistics, overall and stratified by racial/ethnic identity

	Overall (n=5,314)	Hispanic (n=715)	Non-Hispanic Black (n=992)	Non-Hispanic White (n=3,607)	p-value*
	N Weighted % (95% CI)	N Weighted % (95% CI)	N Weighted % (95% CI)	N Weighted % (95% CI)
Age (years)					<.001
51-59	1,215 29.2 (.27, .32)	230 42.4 (.35, .50)	297 37.5 (.32, .43)	688 26.2 (.23, .30)
60-69	1,818 36.6 (34, .39)	305 33.8 (.27, .41)	399 33.7 (.29, .38)	1,114 37.5 (.34, .41)
70-79	1,321 22.1 (.20, .24)	119 16.9 (.13, .22)	193 19.4 (.16, .24)	1,009 23.2 (.21, .25)
80-89	960 12.0 (.11, .13)	61 7.0 (.05)	103 9.4 (.07, .12)	796 13.1 (.12, .15)
Sex					.078
Female	3,163 53.5 (.52, .55)	427 53.1 (49, .58)	638 58.3 (.54, .62)	2,098 52.9 (.51, .55)
Male	2,151 46.5 (.45, .48)	288 46.9 (.42, .51)	354 41.7 (.38, .46)	1,509 47.1 (.46, .49)
Education					<.001
Less than high school	834 13.7 (.12, .15)	308 45.0 (.39, .51)	211 26.2 (.22, .31)	315 7.5 (.06, .09)
High school graduate	1,587 27.8 (.26, .30)	168 23.6 (.20, .27)	315 29.2 (.26, .32)	1,104 28.2 (.26, .31)
Any college	2,878 58.5 (.56, .61)	237 31.4 (.26, .37)	463 44.7 (.40, .49)	2,178 64.3 (.61, .67)
Missing	15	2	3	10
Employment					.076
Working full- or part-time	1,657 41.2 (.39, .43)	264 47.3 (.40, .55)	305 36.3 (.32, .41)	1,088 41.1 (.38, .44)
Not currently working	3,366 58.8 (.56, .61)	362 52.7 (.45, .60)	660 63.7 (.59, .68)	2,344 58.9 (.56, .62)
Missing	291	89	27	175
Self-rated health					<.001
Excellent or very good	2,064 43.4 (.41, .45)	159 23.7 (.19, .29)	261 28.2 (.24, .33)	1,644 48.3 (.46, .50)
Good	1,865 32.3 (.30, .34)	247 30.6 (.26, .36)	412 37.2 (.33, .42)	1,206 31.8 (.30, .34)
Fair or poor	1,404 24.3 (.22, .26)	309 45.7 (.38, .54)	318 34.5 (.30, .39)	756 19.9 (.18, .22)
Missing	2	0	1	1
Everyday discrimination					.004
None	2,817 54.0 (.52, .56)	374 49.7 (.45, .54)	462 47.6 (.43, .53)	1,981 55.4 (.53, .58)
Any	2,280 46.0 (.44, .48)	296 50.3 (.46, .55)	461 52.4 (.47, .57)	1,523 44.6 (.42, .47)
Missing	217	45	69	103
Any chronic pain					.049
No	3125 60.0 (58, .62)	412 55.3 (.50, .61)	562 56.1 (.51, .61)	2,151 61.2 (.59, .63)
Yes	2185 40.0 (.38, .42)	301 44.7 (.39, .50)	430 43.9 (.39, .49)	1,454 38.8 (.37, .41)
Missing	4	2	0	2
Severe chronic pain					<.001
No	4,919 93.0 (92, .94)	625 87.2 (.84, .90)	883 86.6 (.83, .90)	3411 94.7 (.93, .96)
Yes	383 7.0 (.06, .08)	86 12.8 (.10, .16)	108 13.5 (.10, .17)	189 5.3 (.04, .07)
Missing	12	4	1	7
High impact chronic pain					.008
No	3,944 74.5 (.73, .76)	510 68.0 (.60, .75)	709 69.0 (.64, .74)	2,725 76.2 (.74, .78)
Yes	1,366 25.5 (.24, .27)	203 32.0 (.25, .40)	283 31.0 (.26, .36)	880 23.8 (.22, .26)
Missing	4	2	0	2

Notes:

^*p-value based on differences between racial/ethnic identity groups for the weighted sample;

CI confidence interval

Table 1 also shows weighted estimates for everyday discrimination and chronic pain status, overall and across each racial/ethnic group. Less than half of the weighted sample reported any exposure to everyday discrimination (46.0%) and there were racial/ethnic differences in everyday discrimination exposure. Fewer than half of non-Hispanic White respondents (44.6%) reported exposure to everyday discrimination, compared to 50.3% and 52.4% of Hispanic and non-Hispanic Black respondents, respectively (p=.004). Regarding chronic pain status, 40.0% and 25.5% of respondents reported any chronic pain and high impact chronic pain, respectively, while just 7.0% reported severe chronic pain. The unadjusted chronic pain prevalence also differed across racial/ethnic groups with significantly higher estimates of severe (p<.001) and high impact (p=.008) chronic pain among Hispanic and non-Hispanic Black respondents.

Table 2 shows unweighted and weighted adjusted odds ratio for each chronic pain status (any, severe, high impact) using two separate models.

Table 2.

Logistic regression of any chronic pain, severe chronic pain, and high impact chronic pain

	Model 1				Model 2
	Unweighted		Weighted		Unweighted		Weighted
	aOR	SE	aOR	SE	aOR	SE	aOR	SE
Any chronic pain
Everyday discrimination (REF: none)
Any	1.50 ***	.09	1.60 ***	.13	1.36 ***	.09	1.43 ***	.12
Racial/ethnic identity (REF: Non-Hispanic White)
Hispanic	1.08	.09	1.29	.18	.68 ***	.07	.74 *	.10
Non-Hispanic Black	1.12	.09	1.20	.15	.78 **	.07	.79	.11
Severe chronic pain
Everyday discrimination (REF: none)
Any	1.84 ***	.2	1.73 **	.28	1.65 ***	.20	1.55 *	.27
Racial/ethnic identity (REF: Non-Hispanic White)
Hispanic	2.45 ***	.35	2.56 ***	.53	1.47 *	.25	1.35	.26
Non-Hispanic Black	2.04 ***	.27	2.50 ***	.53	1.35 *	.20	1.54	.34
High impact chronic pain
Everyday discrimination (REF: none)
Any	1.60 ***	.10	1.70 ***	.14	1.45 ***	.11	1.56 ***	.14
Racial/ethnic identity (REF: Non-Hispanic White)
Hispanic	1.20	.12	1.51 *	.29	.69 **	.08	.78	.14
Non-Hispanic Black	1.14	.10	1.33	.20	.74 **	.07	.81	.14

Notes:

^*p< 0.05,

^**p < 0.01,

^***p < 0.001.

Model 1 adjusts for age and sex; Model 2 adjusts for age, sex, educational attainment, employment status, and self-rated health.

REF reference group; aOR adjusted odds ratio; SE standard error

Model 1

After adjusting for age and sex, both the unweighted and weighted models show that any exposure to everyday discrimination predicted a higher odds for any chronic pain, severe chronic pain, and high impact chronic pain. In addition, racial/ethnic identity predicted an increased odds of severe chronic pain independent of exposure to everyday discrimination. In the weighted model, compared to non-Hispanic White respondents, both Hispanic (aOR=2.56; p <.001) and non-Hispanic Black respondents (aOR=2.50, p <.001) had significantly higher odds of experiencing severe chronic pain, independent of exposure to discrimination, while only Hispanic respondents had a significantly higher odds of experiencing high impact chronic pain (aOR=1.51, p <.05).

Average marginal effects of exposure to everyday discrimination on the probability of any, severe, and high impact chronic pain were estimated for each racial/ethnic group and significant racial/ethnic differences are illustrated in Figures 1–2. There were no significant differences between groups with respect to any chronic pain. However, Hispanic and non-Hispanic Black respondents with and without exposure to everyday discrimination had a significantly higher probability of severe chronic pain, relative to their non-Hispanic White counterparts (Figure 1). Additionally, while there were no Hispanic-White differences among respondents without exposure to everyday discrimination, Hispanic respondents with exposure to everyday discrimination had a significantly higher probability of high impact chronic pain compared to their non-Hispanic White counterparts (Figure 2).

Figure 1.

Probability of severe chronic pain based on exposure to everyday discrimination by racial/ethnic identity (weighted sample)

Figure notes: Model estimates are based on the weighted sample and are adjusted for age and sex.

Figure 2.

Probability of high impact chronic pain based on exposure to everyday discrimination by racial/ethnic identity (weighted sample)

Figure notes: Model estimates are based on the weighted sample and are adjusted for age and sex.

Model 2

The second model shows the fully adjusted odds of any, severe, and high impact chronic pain after accounting for age, sex, educational attainment, employment status, and self-rated health. Exposure to everyday discrimination remained an independent predictor of any, severe, and high impact chronic pain in both the unweighted and weighted models, although the odds are slightly attenuated relative to Model 1. Notably, and contrary to Model 1, Hispanic respondents had a significantly lower odds of experiencing any chronic pain in both the unweighted and weighted models, relative to non-Hispanic White respondents, after accounting for education, employment, and self-rated health. In addition, while Hispanic and non-Hispanic Black respondents had a significantly greater odds of severe chronic pain, independent of everyday discrimination, in the unweighted models, this relationship was no longer significant in the weighted models.

Average marginal effects of exposure to everyday discrimination on the probability of any, severe, and high impact chronic pain were also estimated for each racial/ethnic group. Based on the weighted estimates, there were no significant racial/ethnic differences between the unexposed and exposed groups with respect to severe and high impact chronic pain. However, Hispanic respondents with and without exposure to everyday discrimination had a significantly lower probability of any chronic pain, relative to their non-Hispanic White counterparts (Figure 3).

Figure 3.

Probability of any chronic pain based on exposure to everyday discrimination by racial/ethnic identity (weighted sample)

Figure notes: Model estimates are based on the weighted sample and are adjusted for age, sex, educational attainment, employment status, and self-rated health.

Interaction effects

Table 3 shows the unweighted and weighted estimates after we assessed for potential interactions between exposure to everyday discrimination and racial/ethnic identity. In both the unweighted and unweighted models, exposure to discrimination remained an independent predictor of higher odds to experience each type of chronic pain, although it was attenuated in models that adjusted for socio-demographic and health factors. In the unweighted models, the main effects in Model 1 (adjusted for age and sex) were qualified by significant interactions between exposure to everyday discrimination and racial/ethnic identity. For non-Hispanic Black respondents, exposure to everyday discrimination had a significantly lesser effect in predicting the odds of any chronic pain (aOR=.72, p <.05) and severe chronic pain (aOR=.56, p<.05) relative to non-Hispanic White respondents. There were no significant interaction effects between Hispanic and non-Hispanic White respondents with respect to exposure to everyday discrimination predicting any, severe, or high impact chronic pain. The unweighted estimates for Model 2 (adjusted for age, sex, education, employment, and self-rated health) also show that the main effects for severe chronic pain continued to be qualified by significant interactions between exposure to everyday discrimination and racial/ethnic identity. For non-Hispanic Black respondents, exposure to everyday discrimination had a significantly lesser effect in predicting the odds of severe chronic pain (aOR=.56; 95% CI=.34, .95) relative to non-Hispanic White respondents. Supplementary Figures S1–3 illustrate the racial/ethnic differences in the influence of exposure to everyday discrimination on the probability of any and severe chronic pain when interaction effects are considered based on the unweighted estimates. Notably, although the estimates were similar in both the unweighted and weighted models in terms of direction, the interaction effects were not statistically significant in the weighted analyses.

Table 3.

Logistic regression of any chronic pain, severe chronic pain, and high impact chronic pain including interaction between everyday discrimination and racial/ethnic identity

	Model 1 + interaction				Model 2 + interaction
	Unweighted		Weighted		Unweighted		Weighted
	aOR	SE	aOR	SE	aOR	SE	aOR	SE
Any chronic pain
Everyday discrimination (REF: none)
Any	1.61 ***	.11	1.60 ***	.15	1.43 ***	.11	1.41 ***	.13
Racial/ethnic identity (REF: Non-Hispanic White)
Hispanic	1.11	.13	1.14	.18	.69 **	.10	.63 *	.13
Non-Hispanic Black	1.32 **	14	1.37	22	.90	11	.87	15
Everyday discrimination x Racial/ethnic identity (REF: Any discrimination/Non-Hispanic White)
Any discrimination/Hispanic	.94	16	1.28	36	.98	19	1.37	45
Any discrimination/Non-Hispanic Black	.72 *	11	.78	15	.77	13	.84	17
Severe chronic pain
Everyday discrimination (REF: none)
Any	2.06 ***	32	1.88 **	40	1.84 ***	31	1.72 *	38
Racial/ethnic identity (REF: Non-Hispanic White)
Hispanic	2.19 **	35	2.44 **	76	1.25	33	1.33	42
Non-Hispanic Black	2.84 ***	56	3.33 ***	87	1.87 **	40	2.13 *	61
Everyday discrimination x Racial/ethnic identity (REF: Any discrimination/Non-Hispanic White)
Any discrimination/Hispanic	1.22	35	1.07	41	1.34	44	1.01	37
Any discrimination/Non-Hispanic Black	.56 *	15	.62	20	.56 *	16	.57	19
High impact chronic pain
Everyday discrimination (REF: none)
Any	1.69 ***	10	1.73 ***	16	1.51 ***	11	1.59 ***	14
Racial/ethnic identity (REF: Non-Hispanic White)
Hispanic	1.18	16	1.38	29	.67 *	11	.70	18
Non-Hispanic Black	1.35 *	16	1.57 *	28	.84	12	.95	18
Everyday discrimination x Racial/ethnic identity (REF: Any discrimination/Non-Hispanic White)
Any discrimination/Hispanic	1.03	.20	1.18	.37	1.05	.23	1.19	.44
Any discrimination/Non-Hispanic Black	1.35 *	16	.74	15	.78	15	.75	16

Notes:

^*p< 0.05

^**p < 0.01

^***p < 0.001.

Model 3 accounts for age, sex; Model 4 accounts for age, sex, educational attainment, employment status, and self-rated health.

REF reference group; aOR adjusted odds ratio; SE standard error.

To evaluate within-group differences, Table 4 shows the adjusted odds ratios for any, severe, and high impact chronic pain among each racial/ethnic group based on exposure to everyday discrimination. Model 1 shows that, after accounting for age and sex, any exposure to everyday discrimination increased the odds of experiencing each type of chronic pain among non-Hispanic White respondents, and for any and severe chronic pain among Hispanic respondents. There was no significant effect of everyday discrimination on odds of either type of chronic pain among non-Hispanic Black respondents. Model 2 further accounts for educational attainment, employment status, and self-rated health and shows that any exposure to everyday discrimination continues to predict higher odds of experiencing any, severe, and high impact chronic pain among non-Hispanic White respondents. There remains so significant effect among non-Hispanic Black respondents. In addition, Model 2 shows that, after further accounting for socio-demographic and health factors, everyday discrimination is no longer a significant predictor among Hispanic respondents regarding odds to experience any, severe, or high impact chronic pain.

Table 4.

Logistic regression of any, severe, and high impact chronic pain stratified by racial and ethnic identity

		Hispanic		Non-Hispanic Black		Non-Hispanic White
		aOR	SE	aOR	SE	aOR	SE
Any chronic pain
Model 1: Everyday discrimination (REF: none)
	Any	2.13**	.60	1.22	.20	1.60***	.15
Model 2: Everyday discrimination (REF: none)
	Any	1.96	.67	1.19	.23	1.41***	.13
Severe chronic pain
Model 1: Everyday discrimination (REF: none)
	Any	2.24*	.76	1.16	.33	1.85**	.41
Model 2: Everyday discrimination (REF: none)
	Any	1.99	.73	1.03	.30	1.68*	.38
High impact chronic pain
Model 1: Everyday discrimination (REF: none)
	Any	2.11*	.64	1.25	.22	1.74***	.17
Model 2: Everyday discrimination (REF: none)
	Any	1.96	.72	1.21	.23	1.59***	.16

Notes: Model estimates are weighted to account for survey design features; Model 1 adjusts for age and sex; Model 2 adjusts for age, sex, educational attainment, employment status, and self-rated health;

REF reference group; aOR adjusted odds ratio; SE standard error.

DISCUSSION

To our knowledge, the present study is the first to use a nationally representative sample of middle-aged and older adults in the U.S. to investigate the effects of everyday discrimination and racial/ethnic identity on the chronic pain burden among this population. In unadjusted analyses, our results indicate that in 2018, 40% of middle-aged and older adults in the U.S. experienced any chronic pain, 26% experienced high impact chronic pain, with far fewer reporting severe chronic pain (7%). Unadjusted analyses also provide evidence to a greater chronic pain burden among Hispanic and non-Hispanic Black middle-aged and older adults. Regarding the relationship between everyday discrimination and chronic pain, the present study’s findings indicate that any exposure to everyday discrimination increases the odds to experience any, severe, and high impact chronic pain within this population, independent of racial/ethnic identify, socio-demographic, and health factors. In addition, study results show that there are notable differences regarding the influence of everyday discrimination on chronic pain among racial/ethnic subpopulations. These findings contribute to the growing body of evidence on the detrimental consequences of discrimination on health outcomes^8,17,30, including in the area of pain^7,37, as well as underscoring a need for additional research on the ways in which exposure to discrimination can have differential effects based on racial/ethnic identity.

One of the key findings of the current study, and consistent with our hypothesis, was that exposure to everyday discrimination predicted higher odds of experiencing all types of chronic pain among middle-aged and older adults, and this relationship held across all models. These findings lend support to prior research that has shown similar associations between everyday discrimination and pain.^6,7,15,19,37 Exposure to routine discrimination has been linked to alterations in the body’s response to stress through both physiological and psychological processes that can influence the development and severity of pain.^{6,11,29,37,38} It should be noted that the overall exposure to everyday discrimination within the study sample was lower than anticipated, with fewer than half of respondents reporting at least some exposure – which in many cases could be less than once a year on average. In addition, exposure to everyday discrimination in the present study was not exclusive to racial discrimination, and could be attributed to other factors, such as age, sex, or financial status. Indeed, we did find that racial discrimination was the most frequent reason reported for everyday discrimination among non-Hispanic Black respondents versus age and financial status being most frequently reported among non-Hispanic White and Hispanic respondents, respectively (data not shown). Future research that compares different measures of discrimination, such as major experiences of discrimination or which are exclusive to racial discrimination, for instance, can provide a more detailed understanding of the influence of discrimination experiences on chronic pain within this population. Regarding targets for intervention, strategies that consider how middle-aged and older adults cope with everyday discrimination provide one potential avenue to address potential influences of everyday discrimination on pain. A more fruitful approach would be to consider upstream strategies that address the policies and practices which create the conditions for discrimination to occur.

Another key takeaway was that exposure to everyday discrimination behaved differently within the racial/ethnic subgroups. Contrary to our hypothesis, exposure to everyday discrimination significantly raised pain risk among Hispanic and non-Hispanic White, but not non-Hispanic Black, middle-aged and older adults. The chronic pain burden was already relatively high among non-Hispanic Black middle-aged and older adults without exposure to everyday discrimination, and showed no significant difference to the group who was exposed to everyday discrimination across all chronic pain categories. There are several potential explanations for this finding. It is possible that other forms of discrimination are more salient for non-Hispanic Black middle-aged and older adults, such as structural forms of disadvantage—which disproportionately affect this population and have consistently been shown to have detrimental health consequences.³ Indeed, some scholars posit that racial identity should be considered a proxy for exposure to systemic racism.²⁶ In this same vein, it may be that our findings are in fact showing the effects of systemic racism, since the probability of chronic pain remained largely unchanged between the unexposed and exposed non-Hispanic Black respondents. It is possible that these structural forces are more explanatory regarding Black-White racial differences in chronic pain burden than interpersonal forms of discrimination as measured by the Everyday Discrimination Scale.²⁶ It is also possible that other measures of discrimination may yield different results than shown here, particularly measures that are explicitly focused on racial discrimination. Some scholars have critiqued the degree to which the Everyday Discrimination Scale is valid to use for comparisons in discrimination experiences across racial/ethnic identities.^4,22 Taken together, to advance health equity in the area of chronic pain, future research is warranted to more robustly evaluate how well different measures of discrimination perform across diverse populations of middle-aged and older adults.

Third, the present study initially found that Hispanic middle-aged and older adults had higher odds to experience severe and high impact chronic pain, relative to non-Hispanic White middle-aged and older adults, when only adjusting for discrimination exposure, age and sex. However, when factoring education, employment, and self-rated health into the weighted analyses, Hispanic middle-aged and older had significantly lower odds of any chronic pain – and no notable difference in severe and high impact chronic pain – relative to non-Hispanic White middle-aged and older adults. These results align with some prior research, which has shown that Hispanic Americans had lower or equivalent pain prevalence to non-Hispanic White adults once accounting for socio-demographic factors.⁴² In addition, previous studies using HRS data have similarly found that Hispanic respondents were more likely to experience severe pain relative to non-Hispanic White respondents prior to accounting for other socio-demographic and health factors.^21,33 In contrast to non-Hispanic Black middle-aged and older adults, when looking at relationships between everyday discrimination and odds to experience chronic pain among Hispanic respondents, the current study findings’ indicate that exposure to everyday discrimination does significantly increase their odds of chronic pain until accounting for sociodemographic and health factors. In sum, the current study’s findings suggest that the chronic pain experiences among Hispanic middle-aged and older adults share similarities and differences to both their non-Hispanic Black and White counterparts, yet our understanding of these nuances remains quite limited. Given projections that growth among the Hispanic Americans will increase at a more substantial pace than both non-Hispanic Black and non-Hispanic White populations in the coming decades, further research is warranted to promote healthy aging for this group, particularly that which considers the significant diversity that exists within this population.¹⁸

Notably, we indicate above how we observed an equivalent or lower odds of any and high impact chronic pain among Hispanic and non-Hispanic Black respondents, relative to non-Hispanic White respondents, once accounting for socio-demographic and health factors. Prior pain research investigating racial/ethnic differences in pain status has similarly found that the presence of pain disparities disappeared once accounting for socio-demographic and health factors as covariates. While some of these studies have suggested an absence of pain disparities or that adults from racial/ethnic minoritized groups may actually have an “advantage” when it comes to pain—we urge against this interpretation.^21,33 Alternatively, we recommend that future studies begin to investigate the effects of socioeconomic and health status variables on relationships between discrimination, racial/ethnic identity, and chronic pain, as potential moderators or mediators of these relationships versus covariates. Indeed, some scholars have posited that when socioeconomic and health status factors are considered, they should be included as intermediate variables along the causal pathway between race and health outcomes to avoid the potential interpretation that race is less important relative to these other factors.^2,28 When these additional analyses are not done, as in the case of the present study, investigators should then acknowledge the potential influence of socioeconomic and health factors along the causal pathway between discrimination and chronic pain.

The present study does have several limitations. Because this analysis was cross-sectional, the directionality of relationships between everyday discrimination and chronic pain are unable to be ascertained here, and it is possible that middle-aged and older adults with chronic pain are more likely to experience everyday discrimination. In addition, while the chronic pain definitions and prevalence estimates in the current study align with some prior work²¹, there remains significant variability in how chronic pain is defined, which can limit comparisons of estimates across studies.³⁶ For instance, Pitcher et al. (2019) reported that 4.8% of U.S. adults experienced high impact chronic pain using data from the National Health Interview Survey, which is substantially lower than the estimate provided here.³² While this is likely explained, at least in part, by our target population being middle-aged and older adults, there are also differences in how the authors defined chronic pain and high impact chronic pain within their study. Given the societal impact of chronic pain, the lack of a clear definition of chronic pain across population-based studies remains an area for improvement nationally. Finally, the potential influence of intersectional discrimination, or different types of discrimination that are experienced simultaneously (e.g., ageism, racism), should be considered and was not examined in the present study. Some emerging research has used analytic approaches, such as latent class analysis, to better understand these intersections and we believe that this is a fruitful area for subsequent investigations that examine associations between discrimination and chronic pain among middle-aged and older adults.^9,10,16

CONCLUSION

The current study’s findings provide additional evidence on the substantial burden of chronic pain among middle-aged and older adults in the U.S., especially among racial and ethnic minoritized groups. A key finding was that exposure to everyday discrimination predicted a higher odds of any, severe, and high impact chronic pain, independent of socio-demographic and health factors. Another key takeaway was that exposure to everyday discrimination was a significant predictor of odds to experience chronic pain among Hispanic and non-Hispanic, but not non-Hispanic Black, middle-aged and older adults, suggesting that other factors are more salient in our understanding of chronic pain within this latter group.

PERSPECTIVE.

Using national data, we examined associations between discrimination and chronic pain among middle-aged and older adults, including interactions between discrimination and race/ethnicity. Exposure to discrimination predicted a higher chronic pain burden, overall. Differential effects within racial/ethnic groups underscored a need for more nuanced investigations into pain disparities among this population.

Highlights.

• Exposure to everyday discrimination predicted a higher odds of chronic pain among the study sample

• Everyday discrimination had differential effects on odds of chronic pain by racial/ethnic identity

• Nuanced research is needed to better understand the effects of discrimination on pain disparities

Availability of data and materials:

Data for the 2018 Health and Retirement Study core public survey data is available here: https://hrsdata.isr.umich.edu/data-products/public-survey-data?_ga=2.8288378.953035295.1674504685-1731101969.1674504685