Perceived Racial Discrimination and Pain Intensity/Disability Among Economically Disadvantaged Latinos in a Federally Qualified Health Center: The Role of Anxiety Sensitivity

Abstract

The present study examined the role of anxiety sensitivity (AS; fear of the negative consequences of anxiety) in the relation between perceived racial discrimination and pain-related problems among Latinos seeking health services at a Federally Qualified Health Center. Participants included 145 adult Latinos (87.80% female, M_age = 38.07 years, SD = 11.98, and 96.2% reported Spanish as their first language). Results indicated that perceived racial discrimination was indirectly related to the pain intensity and pain disability through AS. These effects were evident above and beyond the variance accounted for by gender, age, marital status, educational status, employment status, years living in the United States, and number of axis I diagnoses. Overall, the present findings highlight the merit in focusing further scientific attention on the interplay between perceived racial discrimination and AS to better understand and inform interventions to reduce pain problems among Latinos in primary care.

Introduction

The Latino population has grown over 800% since the 1950s in the United States (U.S.) and now represent one of the largest minority groups [1]. Latinos experience high rates of health inequalities across numerous public health indicators (e.g., treatment utilization, access to care, severity of medical problems, psychiatric conditions) relative to many other groups [2]. One of the most common and costly public health problems among the Latino population is the experience of pain and clinical pain disorders [3]. In fact, Latinos are at greater risk for poorer quality of pain management (e.g., lower rates of prescribed medication), and moreover, pain is often not assessed during routine medical visits [4]. Additionally, compared to non-Latino Whites, Latinos report fewer medical visits and longer waiting times for pain care [4-6]. Research also suggests that the experience of pain is frequently viewed as a ‘common problem’ in the Latino community that does not warrant treatment [7]. Such perspectives, in conjunction with personal impact of pain on life functioning, contribute to the large-scale financial burden of pain complaints among the Latino population in the U.S. relative to other groups [7].

Among Latinos, the experience of pain can be influenced by a wide range of cultural beliefs, such as simpatía (seeking harmony within interpersonal relationships), respect and family involvement, and culturally influenced beliefs [8], among others. For example, research suggests that Latinos common belief systems include ideas that pain is predestined, pain is a necessary part of life, pain should be endured with stoicism, and there is value in maintaining simpatía by not letting your pain problem interfere with family and friend relationships [9-11]. Occupational demands from manual labor jobs (e.g., cleaning, construction), wherein Latinos represent a large percentage of the U.S. workforce [12], may also increase the chance of pain-inducing injury and chronic bodily stress. In fact, empirical research among Latinos suggests that pain severity and overall distress are greater among Latinos compared to non-Latino Whites e.g., [13, 14]. Pain severity also is associated with greater physical impairment and disability among Latinos [15]. Yet, despite the public health importance of pain among Latinos, there is little understanding of the psychological mechanisms related to their pain experience.

One psychological construct that may be relevant to better understanding the expression, development, and maintenance of pain is perceived racial discrimination [16, 17]. Perceived discrimination reflects a negative attitude, judgment, or unfair treatment toward members of a group [18, 19]. Perceived racial discrimination among Latinos is highly common, with estimates suggesting more than half of the population report that they have experienced discrimination or have been treated unfairly because of their race or ethnicity [20]. A large body of research suggests, similar to other minority groups [21], perceived racial discrimination among Latinos is related to greater chronic stress, more severe medical problems, and psychological disorders [22, 23]. Among Latinos, although the study of perceived racial discrimination and pain experience is highly limited, available work suggests perceived discrimination [16, 17] is related to more severe pain [24]. For example, in a multi-site longitudinal study designed to examine the health of Latina women during middle age, Dugan et al. [24] found perceived racial discrimination was related to greater bodily pain. The findings of this initial investigation prompt further study of the linkages between perceived racial discrimination and pain experience among Latinos.

One clinically important area in need of further study pertains to the mechanisms linking perceived racial discrimination to pain experience (intensity and disability) among Latinos. Anxiety sensitivity (AS), which reflects the extent to which individuals believe anxiety and anxiety-related sensations have harmful consequences [25, 26], is one possible mechanism linking perceived discrimination and pain. AS is a relatively stable, yet malleable, cognitive vulnerability that is associated with pain in clinical and nonclinical samples [27, 28]. Models of AS-pain relations are based on the role of fear and avoidance processes involved in managing pain experience [29-31]. According to such models, due to a fear of pain, individuals with high AS evince more pain-related avoidance, thereby exacerbating pain severity. High AS persons are also more apt to engage in catastrophic thinking about the negative consequences of somatic sensations, which worsens pain severity and disability [32, 33]. Despite the wealth of knowledge regarding the relation between AS and pain experience in non-Hispanic Whites, studies among Latinos are highly limited. Of two available investigations, AS was related to more severe pain among Latinos in primary care [34-36].

The AS construct may help explain the relation between perceived racial discrimination and pain intensity and disability among Latinos. Past work has demonstrated that perceived racial discrimination may be associated with greater life stress [37-40]. Exposure to such stressful life events, in theory, may lead to increased levels of AS [41, 42]. Specifically, the cumulative experience of stress related to repeated exposure to perceived discrimination may progressively amplify AS-based cognitive-affective responses, such as negative interpretation of the physical sensations [43, 44]. Such elevated AS may, in turn, be related to greater expression of pain intensity and disability. To illustrate, a Latino individual who perceive experiences of chronic discrimination may learn to be more attuned to aversive internal sensations and fear their personal consequences, or exacerbate concerns about the displaying visible signs of anxiety in public. These types of fear concerns (i.e. AS), in turn, may be related to increased pain intensity and disability.

Together, the current study tested the hypothesis that among adult Spanish-speaking Latinos in primary care, AS would partially explain the associations between perceived racial discrimination and pain intensity and disability (see Fig. 1). It was further expected that any explanatory effects would not be explained by theoretically relevant covariates, including gender [45], age [46], marital status [47], educational status [48], employment status [49], years living in the U.S. [50], and number of axis I diagnoses [51]. The present study involved a primary care sample at a Federally Qualified Health Center (FQHC), as “pain complaints” account for more than half of all outpatient primary care visits [52]. Primary care clinics also are the most common service domain for Latinos to seek healthcare for pain and related health problems [53].

Fig. 1.

Proposed model: Anxiety Sensitivity as the Proposed Mediator of the Relationship between Perceived Discrimination and Pain problems. a Effect of X on M; b Effect of M on Y_i; c Total effect of X on Y_i; c’ Direct effect of X on Y_i controlling for M; a*b Indirect effect of M

Note: a = Effect of X on M; b = Effect of M on Y_i; c = Total effect of X on Y_i; c’ = Direct effect of X on Y_i controlling for M; a*b = Indirect effect of M.

Method

Participants were recruited from a FQHC primary care health facility located in an urban southwestern community. The inclusion criteria consisted of the ability to read, write and communicate in Spanish, and being between 18 and 64 years old. Participants were excluded if they exhibited limited mental competency and/or inability to provide informed, voluntary, written consent, or if they endorsed current or past psychotic-spectrum symptoms via structured interview screening.

Measures

Demographics

Participants provided demographic information including age, sex, educational attainment, marital status, employment status, and amount of time living in the US.

MINI International Neuropsychiatric Interview MINI 6.0

Diagnostic assessments were performed using the MINI. The MINI provides reliable DSM-IV diagnoses within a short time frame, which is applicable to research settings [54]. The MINI has demonstrated sound inter-rater and test–retest reliability and validity [55]. The MINI has been successfully employed among Latino samples e.g. [56]. The interviews were administered by Spanish-speaking staff who were trained on DSM-based diagnosis and MINI diagnostic interviewing. Interviewers were supervised by an independent doctoral-level rater. Approximately 12% of the MINI interview videos (including audio) were randomly checked through a trainer-review process; no cases of diagnostic coding disagreement were noted. For this study, the total number of current axis I disorders per MINI for each individual was used as a covariate.

Perceived Racial Discrimination

Discrimination was operationalized using items from the Social, Attitudinal, Familial, and Environmental Acculturation Stress scale SAFE; [57]. SAFE has been shown to be a reliable and valid measure of perceived racial discrimination in general and among Latinos [58]. The SAFE is a 24-item measure used to assess stressors associated with acculturation in social, attitudinal, familial, and environmental contexts as well as perceived discrimination towards acculturating populations. Items are rated on a 5-point Likert scale (1 = not stressful to 5 = extremely stressful). Items that directly tapped into stress associated with discrimination and stigma were used, as in past work [59]. These items were as follows: ‘‘Because of my ethnicity, people exclude me from participation in activities,’’ ‘‘Many people stereotype my culture or ethnic group, and they treat me as if they are in the right’’; ‘‘When I look for work, I feel limited due to my ethnicity/race,’’ and ‘‘I feel uncomfortable when people laugh at people from my ethnic group’’. A composite score was made by summing the perceived discrimination items (range 4–18; Cronbach α = 0.80).

Anxiety Sensitivity Index-3

The ASI-3 is an 18-item measure [60], based, in part, upon the original Anxiety Sensitivity Index [26] in which respondents indicate the extent to which they are concerned about possible negative consequences of anxiety-related symptoms. Responses are rated on a 5-point Likert scale ranging from 0 (very little) to 4 (very much) and summed to a total score (ASI-3-Total). The ASI-3 has strong and improved psychometric properties relative to previous measures of the construct and has been successfully employed among Latino samples [35, 56]. In the present investigation, internal constancy was excellent for the total score (Cronbach α = 0.91).

Graded Chronic Pain Scale

The Graded Chronic Pain Scale (GCPS) assesses pain-related intensity and disability [61]. The GCPS has had acceptable psychometric properties, providing a reliable and valid method of assessing global pain intensity [62]. This scale has also previously been used among Spanish-speaking samples e.g. [34, 56, 63]. Six items assess the average intensity of respondents’ pain (e.g., “How would you classify your pain”) as well as the average disability due to the experienced pain (e.g., “How much has pain interfered with your daily activities”) during the past 6 months using separate 0 to 10 numerical rating scales. Mean of the sum scores for intensity and disability were calculated, yielding continuous composite scores. Both GCPS subscales demonstrated excellent internal consistency in the present sample (Cronbach α’s = 0.91 and 0.93).

Procedure

Participants were attendees of a community-based primary health care clinic. Individuals provided informed written consent (in Spanish) prior to participation in study procedures. After providing consent, participants completed a semi-structured clinical interview (MINI) in a private office space and self-report measures. Participants were compensated with $20. The Institutional Review Board approved the study protocol.

Analytic Strategy

Descriptive statistics and bivariate correlations were calculated for all study variables. Analyses were conducted with the PROCESS Macro [64], a computational tool for observed variable analysis using IBM SPSS version 23.0. Five thousand bootstrap re-samplings were conducted to detect the indirect effects of the proposed predictor on dependent variables through AS (i.e., the product of the beta coefficients of path A and path B; see Fig. 1). As a non-parametric method, bootstrapping estimates the sampling distribution of an estimator based on resampling with replacement. A bootstrap-confidence interval that does not include zero provides evidence of a significant indirect effect [65]. It is generally agreed that a significant indirect effect, from the predictor variable, through the explanatory variable, to the dependent variable, is the only requirement necessary to demonstrate the indirect effects [65, 66]. The indirect effect was computed for each of the samples, resulting in an empirically generated sampling distribution [64]. Separate analyses were conducted for each dependent measure, with perceived racial discrimination as the predictor and AS as the proposed explanatory variable in each analysis (see Fig. 1). Covariates included gender, age, marital status, educational status, employment status, years living in the U.S., and number of axis I diagnoses. As recommended, the theoretical models for each dependent variable were compared with an alternative model in which the dependent variable was treated as an explanatory variable of the association between perceived racial discrimination and AS [65, 67, 68].

Results

Descriptive Data

The sample (N = 145) was largely female (87.80%) with an average age of 38.07 years (SD = 11.98). Nearly all participants (96.6%) reported Spanish as their first language. In terms of ethnic subgroups, more than half (54.5%) identified as “Mexican American,” 22.9% “Central American,” 11.6% “American/Born in the U.S.,” 5.2% “South American,” 2.2% “Cuban American,” 1.8% “Dominican American,” and 2.0% identified as “Other.” The study was approved by the university’s corresponding Institutional Review Board.

Approximately half (48.2%) of all participants were married, 29.2% were single, 11.9% were living with a partner, 7.5% were divorced, and 3.2% were widowed. Forty-five percent of the sample was unemployed with 31.2% employed full-time, 13.6% working half-time (~ 20 h per week), and 10.1% working part-time (less than 20 h per week). Over half (55.9%) reported earning less than $14,999, and 37.1% reported earning less than $34,999 annually. Less than 10 percent (6.4%) of participants reported earning between $35,000 and $74,999 annually. Participants indicated the following reasons for their visit to the primary care facility: family medicine (41.6%), dental (11.4%), psychiatric/psychological (8.1%), lab test or physical exam (5.0%), or accompanying someone/other reasons (32.5%); 10.9% did not disclose a reason.

Descriptive data and correlations between all variables included in the models are presented in Table 1. Both perceived racial discrimination and AS were related to the dependent measures (r’s range 0.21–0.35; all p’s < 0.01).

Table 1.

Descriptive statistics and bivariate correlations between study variables

	1	2	3	4	5	6	7	8	9	10	11	Mean (SD) or %
1. Agea	1	0.08	−0.16*	−0.39**	0.12	0.22**	0.06	0.06	0.14	0.16*	0.08	38.77 (12.54)
2. Sexa		1	0.01	−0.02	0.08	0.01	0.01	0.14	0.03	−0.01	−0.04	87.8% female
3. Maritala			1	0.10	−0.21*	0.01	0.15	0.07	0.02	0.04	0.03	60.1% with partner
4. Edua				1	0.02	−0.03	−0.19*	−0.12	−0.31**	−0.21**	−0.21**	10.73 (3.74)
5. Employa					1	0.02	−0.01	0.06	−0.02	−0.05	−0.04	54.9% employed
6. Yrs- USa						1	−0.02	0.04	0.11	−0.08	0.01	18.82 (13.03)
7. N-AxisI a							1	0.23**	0.43**	0.16*	0.24**	0.42 (0.50)
8. Perceived Discriminationb								1	0.38**	0.21**	0.27**	6.50 (2.90)
9. ASc									1	0.27**	0.35**	12.78 (12.60)
10. Pain Intensityd										1	0.75**	7.56 (7.40)
1. Pain Disabilityd											1	6.00 (5.85)

N = 145; ** p < 0.01, * p < 0.05

^aCovariate

^bPredictor

^cExplanatory variable

^dOutcome

Sex: 1 = male and 2 = female; Age = age in years; Yrs- US = number of years living in US; N-Axis I = number of Axis I diagnosis; perceived discrimination = discrimination measured by the Social, Attitudinal, Familial, and Environmental Scale (SAFE); AS = anxiety sensitivity total score as measured by ASI-3. Pain intensity and pain disability = as measured by Graded Chronic Pain Scale (GCPS)

Tests of Indirect Effects

The full results for the tests of the indirect effects are presented in Table 2. Regarding pain intensity, the full model accounted for a significant amount of variance (R² = 0.38, F [9,135] = 2.65, p < 0.01). Bootstrap analysis revealed a significant positive indirect effect from perceived racial discrimination to pain intensity through AS (point estimate = 0.13, SE = 0.08; BC 95% CI [0.004–0.34]; completely standardized indirect effect = 0.05). Moreover, results for bootstrap analysis of the alternative model yielded non-significant indirect effects (point estimate = 0.15, SE = 0.12; BC 95% CI [− 0.01 to 0.52]).

Table 2.

Regression models

Y	Model	b	SE	t	p	Lower CI (lower)	Upper CI (upper)
Pain intensity	Discrimination → AS (a)	1.511	0.39	3.87	< 0.001	0.75	2.31
	AS → pain intensity (b)	0.08	0.04	1.87	0.06	−0.006	0.17
	Discrimination → pain intensity (c)	0.52	0.21	2.43	0.01	0.10	0.94
	Discrimination → pain intensity (c’)	0.39	0.22	1.75	0.08	−0.05	0.87
	Discrimination → AS → pain intensity (a*b)	0.13	0.08	–	–	0.004	0.34
Pain disability	AS ◊ Pain disability (b)	0.08	0.03	2.35	0.02	0.01	0.15
	Discrimination → pain disability (c)	0.51	0.16	3.12	0.002	0.18	0.85
	Discrimination → pain disability (c’)	0.39	0.17	2.24	0.02	0.04	0.73
	Discrimination → AS → pain disability disability (a*b)	0.12	0.07	–	–	0.02	0.34

^aEffect of X on M

^beffect of M on Yi

^ctotal effect of X on Yi

^c’direct effect of X on Yi controlling for M, CI (lower) lower bound of a 95% confidence interval, CI (upper) upper bound, → affects

Path a is equal across all models; therefore, it presented only in the model with Y₁ to avoid redundancies. The standard error and 95% CI for a*b are obtained by bootstrap with 5000 re-samples. Bolded responses are statistically significant. Discrimination (the perceived discrimination subscale measured by the Social, Attitudinal, Familial, and Environmental Scale) is the predictor, AS (anxiety sensitivity as measured by the ASI-3) is the explanatory variable, and Pain intensity and Pain disability (measured by Graded Chronic Pain Scale) are the outcome variables

The model examining pain disability accounted for a significant amount of variance (R² = 0.52, F [9,135] = 8.32, p < 0.001). Bootstrap analysis revealed a significant positive indirect effect of perceived racial discrimination for pain disability via AS (point estimates = 0.12, SE = 0.07; BC 95% CI [0.02–0.34]; completely standardized indirect effect = 0.07). However, bootstrap analysis of the alternative model also showed a significant indirect path from discrimination to AS through pain disability (point estimate = 0.24, SE = 0.15; BC 95% CI [−0.04 to 0.72]).

Discussion

The current study examined the indirect effect of perceived racial discrimination via AS on pain intensity and disability among Spanish-speaking Latinos in a FQHC primary care setting. As hypothesized, there were significant indirect effects of perceived racial discrimination on pain intensity and disability through AS. These effects were evident above the variance accounted for by gender, age, marital status, educational status, employment status, years living in the U.S., and number of axis I diagnoses. The results suggest that AS may help explain the relations between perceived racial discrimination and pain among Latinos. Specifically, among Latinos, greater experience of negative emotions related to repeated exposure to racial discrimination could progressively intensify anxious and fearful responding to negative internal states, and such sensitivity, in turn, may be related to greater pain intensity and pain disability. This type of finding implicates AS as an individual difference factor that may impact the relation between perceived racial discrimination with pain. Importantly, the rejection of the indirect effects for the alternative model of pain intensity explaining the relation between perceived racial discrimination and AS increases confidence in direction of the findings for pain intensity. However, the alternative model for pain disability explaining the relation between perceived racial discrimination and AS was significant. This finding is in line with evidence among non-Latino samples pertaining to a reciprocal, forward-feeding relationship between pain-related disability and AS [69, 70].

There are several non-mutually exclusive pathways that may link perceived racial discrimination to pain through AS. First, discrimination experiences, especially when more frequent or severe, may contribute to threat-based catastrophic thinking e.g., “I have no control”; [71, 72]. Such catastrophizing could, in turn, lead to misinterpreting innocuous bodily sensations as harmful, thereby increasing vulnerability for pain [27]. Second, considering that Latinos often report higher levels of somatic sensitivities [73, 74], it is possible that the accumulation of stress related to discrimination increases threat-based attention (AS) to interoceptive cues among this population [17] which may, in turn, influence pain experience [30]. Finally, discrimination commonly co-occurs with the experience of actual or perceived social exclusion [75]. In this context, negative experiences related to social exclusion could contribute to withholding aspects of one’s identity and subsequent negative internal experiences [72]. The cumulative effects of these negative internal experiences may contribute to an exaggerated focus on them and their possible negative consequences (AS), linking perceived discrimination with pain [76].

Clinically, findings of the present study could serve to conceptually inform the development of specialized intervention strategies for Latinos in primary care with elevated risk for pain problems. Research among non-Latino samples suggests reducing AS can improve pain severity [77-79]. Therefore, there may be merit to modifying or culturally adapting such interventions for the Latino community and explicitly integrating culturally-relevant factors such as perceived racial discrimination. Thus, future work could usefully develop and subsequently examine culturally-informed AS reduction programs to target the impact of perceived racial discrimination on pain experience.

There are several study caveats that warrant comment. First, the data were cross-sectional. Such a methodological design precludes the ability to establish causal relations between the tested variables. Longitudinal design studies need to be conducted to further strengthen these findings. For example, it may be advisable to use time sampling tactics to track daily discrimination experiences in relation to fluctuations in AS and their corresponding impact on pain experience. Second, the sample was largely female and seeking services for a range of health/mental health issues in primary care. Replication using samples with greater parity for gender and possibly other sociodemographic variables would strengthen the generalizability of the study results. Third, there is some evidence that Latino’s acculturation status is associated with pain related coping e.g. use of diverting attention and praying/hoping; [50]. Future work may therefore benefit from exploring the potential utility of the present explanatory model across Spanish-speakers with different levels of acculturation. Finally, there is evidence that the presentation of pain among Latinos differ by country of origin [80]. Future studies are needed to further examine within-group differences (e.g., potential moderating role of country of origin) in the relation between perceived racial discrimination and pain through AS.

Together, the present study highlights the explanatory utility of AS in the relation between perceived racial discrimination and a pain intensity and disability among Spanish-speaking Latinos in primary care. These findings suggest the importance of assessing AS and indicate it may be advantageous to address AS in the context of experience of perceived racial discrimination and pain among Latinos. Such targeted intervention programs could theoretically help offset the notable health disparities well-documented among this underserved population.