Pain severity and emotion dysregulation among Latinos in a community health care setting: relations to mental health

Abstract

Background:

Although pain severity is often related to poorer mental health and is one of the most common presenting complaints in community health care settings, there is little understanding of the pain experience in relation to anxiety/depressive symptoms and disorders among Latino populations in medical contexts.

Method:

To address this gap, the current study explored an interactive model of pain severity and emotion dysregulation in relation to anxiety/depressive symptoms and psychopathology among 274 Latinos who attended a community-based primary health care clinic [86.9% female; M_age=39.3 (SD=11.2); 96.0% indicated Spanish as their first language].

Results:

Results indicated a statistically significant interaction between pain severity and emotion dysregulation for suicidal symptoms, social anxiety symptoms and number of mood/anxiety disorders, such that more severe pain and greater levels of emotion dysregulation related to poorer mental health. Both pain severity and emotion dysregulation were significant predictors of depressive symptoms, but only pain severity was a significant predictor of anxious arousal symptoms.

Conclusions:

These novel findings suggest a clinically significant interplay between pain severity and emotion dysregulation among Latinos in. The results are discussed in relation to the need for new screening and intervention tactics that address interrelations between pain severity and emotional dysregulation among Latinos seeking treatment in community health-care-based settings.

Latinos are among the largest and most rapidly growing racial/ethnic groups in the United States [1]. Unfortunately, US Latinos also face significant mental health care disparities [2]. For example, when compared with non-Latino Whites, Latinos are less apt to seek and utilize mental health services [3,4]. Of mental health problems, anxiety/depressive symptoms are particularly prevalent among Latinos [5–7].

Primary care/community health care medical settings are the most common health care domain. Due partially to such factors as stigma for seeking mental health care, community health care settings represent an ideal point of contact for early intervention in Latino mental health [8]. Numerous studies have demonstrated the feasibility and initial efficacy of primary care interventions for Latinos suffering from depression (e.g., [9–11]) and, to a lesser extent, anxiety (e.g., [12]). However, there remains a lack of information about risk factors for anxiety/depressive symptoms and psychopathology among Latinos in general and in community health care setting settings specifically [13].

Pain is one factor that is both highly prevalent and frequently co-occurs with anxiety, depression and general medical problems [14]. For example, “pain complaints” account for more than half of all outpatient primary care visits [15]. Primary care patients who endorse pain are also significantly more likely to suffer from anxiety or depressive disorders [16], and pain severity covaries with anxious/depressive symptom severity [17,18]. Although pain research among Latinos is limited [19], available studies have reported that pain severity and overall distress are greater among Latinos compared to Caucasians (e.g., [20]). Among Latinos, pain severity has been associated with greater anxiety and depressive symptoms [21,22]. Moreover, there is some evidence that Latinos tend to endorse a greater number of somatic symptoms associated with mental health problems relative to other cultural groups (e.g., [23]). Yet, we are not aware of any research that has examined pain severity in relation to anxiety and depressive symptoms and psychopathology among Latinos in community health care settings.

Beyond pain severity, there has been an increasing recognition of the importance of understanding how one reacts to emotional distress in the expression of aversive internal states, including pain as well as anxiety/depressive symptoms and psychopathology [24]. Emotion regulation reflects a multidimensional construct with maladaptive emotion regulation (emotion dysregulation) implicated as a transdiagnostic risk factor for anxiety and depressive disorders [25]. Although there are various definitions of emotion dysregulation, most converge on the idea that it reflects difficulties in employing a set of abilities wherein one can observe, understand, evaluate and differentiate one’s emotions and subsequently access strategies to regulate them and to control behavioral responses [24]. While there has been very little work examining emotion dysregulation among Latinos, greater levels of emotion dysregulation are associated with anxiety and depressive symptoms among this population [26]. There also is some empirical evidence that lesser levels of emotion dysregulation may buffer against effects of discrimination on mental health [27] and serve as a protective factor for stress more generally [28]. Despite these observations, studies have not yet evaluated the explanatory role of emotion dysregulation in relation to anxiety and depressive symptoms and psychopathology among Latinos in community health care settings.

Work among the majority population suggests that emotion dysregulation also may impact the association between pain severity and the experience of negative moods [29] by amplifying the experience of pain (e.g., [30]). Indeed, the emotional impact of pain depends on emotion intensity, which can be mitigated through emotion regulation [31]. As informed by the initial empirical observations, pain severity and emotion dysregulation may operate with one another to increase the probability of the greater expression of anxiety and depressive symptoms and psychopathology. Accordingly, greater pain severity may be exacerbated by an individual’s lack of emotional regulatory skills. Conversely, an individual’s emotion regulatory processes may become more disrupted in the context of elevated pain severity. Therefore, these processes may function synergistically to confer greater risk for anxiety and depressive symptoms and psychopathology. From this perspective, a logical next step in research is to further explore the potential interplay of current pain severity and emotion dysregulation as an integrative explanatory process for vulnerability in the expression of anxiety and depressive symptoms and disorders among Latinos in community health care settings.

The aim of the current study was to examine main and interactive effects of current pain severity and emotion dysregulation in relation to anxiety and depressive symptoms/psychopathology among Latinos attending a community health care clinic. It was expected that there would be an interaction between pain severity and emotion dysregulation, such that greater pain severity and emotion dysregulation would be associated with increased anxiety/depressive symptoms and psychopathology. Effects were expected to occur over and above theoretically relevant covariates, including ethnicity, gender, age, number of years in the United States, educational attainment, marital status and negative affectivity. These demographic variables have been shown to be related to anxiety and depression in past work (e.g., [32,33]). Negative affectivity was included as a covariate as it is widely considered a “general distress” variable highly associated with anxiety and depression [34].

1. Method

1.1. Participants

Individuals (n=289) for this study were recruited from a community health care facility located in an urban southwestern area. Potential participants were excluded if there was a history of psychosis (n=10). Further, five multivariate outliers (see below) were discovered and excluded from analyses, resulting in a total of 274 adults.

The inclusion criteria included 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.

1.2. Measures

Validated, Spanish-language versions of all measures were employed in the present study.

1.2.1. Demographics questionnaire

Demographic information collected included gender, age, race, educational level and marital status.

1.2.2. Positive and Negative Affect Scale (PANAS [35])

The PANAS is a self-report measure asking participants to rate the extent to which they experience each of 20 different feelings and emotions (e.g., interested, nervous) based on a Likert scale that ranges from 1 (very slightly or not at all) to 5 (extremely). The measure yields two factors (negative and positive affectivity) with strong documented psychometric properties [35]. The current study utilized the trait version of the PANAS, inquiring about how participants feel “in general,” as has been done in past work among Spanish-speaking populations [36]. The negative affectivity subscale (PANAS-NA) was used in the present investigation (Cronbach’s α=.89).

1.2.3. Graded Chronic Pain Scale (GCPS [37])

The GCPS has acceptable psychometric properties, providing a reliable and valid method of assessing global pain severity [38] and has previously been used among Spanish-speaking samples (e.g., [39]). Items (e.g., “How would you classify your pain?”) assess the severity of respondents’ pain on average during the past 3 months using separate 0–10 numerical rating scales. Total scores were calculated by summing, yielding a continuous composite score of characteristic pain severity. This measure demonstrated excellent internal consistency in the present sample (Cronbach α=.91).

1.2.4. Difficulties in Emotion Regulation Scale (DERS [40])

The DERS is a 36-item self-report measure on which respondents indicate, on a 5-point Likert-style scale (1=almost never to 5=almost always), how often each item applies to them [40]. The DERS has demonstrated high levels of internal consistency (α=.93 [40]) and adequate test–retest reliability over a 4–8-week period (ρ=.88 [40]). In the current investigation, the DERS-total score was used to indicate a global composite index of emotion dysregulation [40]. The DERS-total score demonstrated good internal consistency in the current sample (Cronbach’s α=.89), which is consistent with other work utilizing the DERS among Spanish speakers [41].

1.2.5. Inventory of Depression and Anxiety Symptoms (IDAS [42])

The IDAS is a 64-item self-report instrument that assesses distinct affect symptom dimensions within the past 2 weeks. Items are answered on a 5-point Likert scale ranging from “not at all” to “extremely.” The IDAS subscales show strong internal consistency and convergent and discriminant validity with psychiatric diagnoses and self-report measures as well as short-term retest reliability with both community and psychiatric patient samples [42,43]. The present study used the general depression subscale (20 items; e.g., “I felt exhausted” or “I did not have much of an appetite”), the suicidality subscale (6 items; e.g., “I had thoughts of suicide”), the social anxiety subscale (5 items; e.g., “I was worried about embarrassing myself socially”) and the anxious arousal subscale (8 items; e.g., “I felt a pain in my chest”). As in past work among Spanish speakers [44], these subscales demonstrated good level of internal consistency among the present sample (Cronbach’s α=.94, .79, .90 and .91 for depressive, suicidal, social anxiety and anxious arousal subscales, respectively). These four subscales were selected as dependent variables given that they represent some of the most common and impairing symptoms among this population (e.g., [45]).

1.2.6. MINI International Neuropsychiatric Interview 6.0 (MINI [46])

Diagnostic assessments were performed using the MINI, which provides reliable Diagnostic and Statistical Manual of Mental Disorders diagnoses within a short time frame which is applicable to research settings [46]. The MINI has demonstrated sound interrater and test–retest reliability and validity [47]. The interviews were administered by trained, Spanish-speaking staff and supervised by an independent doctoral-level rater. Approximately 12% of randomly selected interviews were checked for accuracy; no cases of diagnostic disagreement were noted. The MINI was used to obtain all Axis I diagnoses to describe the sample. Further, the dependent variable (number of mood/anxiety disorders) was calculated as the total number of mood and/or anxiety diagnoses.

1.3. Procedure

Participants for the study attended a community-based health care clinic in Texas between April 2014 and May 2015. Individuals interested in participating provided informed written consent (in Spanish) and were given a diagnostic interview and completed self-report questionnaires. Participants were compensated with $20. The study protocol was approved by the Institutional Review Board.

1.4. Analytic strategy

First, bivariate correlations between study variables were evaluated to examine associations. Second, data were screened for univariate normality and multivariate outliers. Then, main and interactive effects of pain severity (GCPS) and emotion dysregulation (DERS) were examined among five outcome variables: depressive, suicidal, social anxiety and anxious arousal symptoms (IDAS subscales), and number of mood and anxiety disorders (via MINI). Hierarchical regression analyses were conducted with predictors centered at their respective means. Covariates of ethnicity, gender, age, number of years in the United States, educational attainment, marital status and negative affectivity were entered in the first step. Pain severity and emotion dysregulation were entered concurrently in the second step with the interaction term entered in the final step. Post hoc simple slope analyses were conducted for significant interaction terms using values of ±1 SD (high/low) from the mean of the moderator variable emotion dysregulation. Finally, as recommended by Hayes [48]), the Johnson–Neyman technique [49] was used to statistically identify regions of significance among the moderator variable where the effect of the predictor (pain severity) on the criterion variables is statistically significant. Specifically, it identifies the point(s) where the effect of the predictor on the criterion changes from being statistically nonsignificant to statistically significant (and vice versa), which provides more precision than the “pick-a-point” approach used in simple slope analysis.

2. Results

2.1. Descriptive statistics

Data were normally distributed with acceptable levels of skewness (< 3) and kurtosis (<10 [50]). Multivariate outliers (n=5) were identified using Mahalanobis distance tests (e.g., [51]). Correlations are presented in Table 1. Pain severity had a statistically significant and positive association with emotion dysregulation and all six dependent variables. Likewise, emotion dysregulation was significantly and positively associated with all dependent variables. Ratings of pain intensity (GCPS; M=17.8; SD=13.6) in the current sample were comparable to samples of those who report nonchronic pain [52].

Table 1.

Means, standard deviations and bivariate correlations among variables (n=274).

Variable	Mean/n (SD/%)	1	2	3	4	5	6	7	8	9	10	11	12	13
1. Ethnicity (% Mexican-American)	169 (61.7)	–
2. Sex (% female)	238 (86.9)	−.02	–
3. Age	39.3 (11.2)	−.04	.04	–
4. Years in US	18.6 (12.3)	.06	−.02	.44**	–
5. Education (% with 12 years plus)	130 (47.5)	−.01	−.03	−.06	.02	–
6. Marital status (% with partner)	184 (67.2)	.09	.03	.02	−.06	−.07	–
7. PANAS-NA	17.4 (5.6)	−.11	.02	−.01	−.04	.03	−.10	–
8. GCPS	17.8 (13.6)	−.06	.09	.08	.10	−.10	−.01	.39**	–
9. DERS	72.2 (18.9)	−.16**	−.02	−.03	−.06	−.10	−.12	.48**	.38**	–
10. IDAS-DEP	34.8 (11.8)	−.10	.04	.02	.03	−.06	−.15*	.67**	.48**	.61 **	–
11. IDAS-SUI	6.4 (1.5)	−.08	−.04	.02	.08	.05	−.13*	.44**	.30**	.41 **	.56**	–
12. IDAS-SOC	6.8 (2.7)	−.01	.05	−.01	.03	−.02	−.12*	.55**	.37**	.36**	.68**	.48**	–	.
13. IDAS-ANX	11.2 (5.1)	−.15*	.05	−.02	.01	−.11	−.15*	.59**	.45**	.35**	.71**	.42**	.62**	–
14. Number of Dx	0.5 (1.1)	−.09	.05	.06	.03	−.01	−.10	.49**	.41**	.48**	.63**	.44**	.50**	.46**

Note: DEP=Depressive; SUI=Suicidal; SOC=Social Anxiety; ANX=Anxious Arousal; Number of Dx=number of mood/anxiety disorders as per the MINI. Ethnicity, sex, age, years in the United States, education, marital status and PANAS-NA were covariates. Numbers across header correspond with variables numbered 1–14.

^*P<.05.

^**P<.01.

Participants had a mean age of 39.3 (SD=11.2) years. Most (86.6%) were female, and nearly all (99.6%) reported Spanish as first language. The majority (61.7%) identified as Mexican-American, with 26.6% Central American, 4.0% South American, 1.1% Cuban American, and 6.6% as “Other.” Those identifying as Mexican-American reported significantly lower emotion dysregulation (r=−.16, P=.007) and anxious arousal (r=−.15, P=.012) relative to those identifying as other ethnicities.

In terms of education, 6.9% of participants reported less than 6 years of education, 45.6% 6–11 years, 28.5% 12 years (completion of high school) and 19.0% more than 12 years. Approximately half (51.1%) of participants were married, 16.1% were living with partner, 24.5% were single, 6.9% were divorced, and 1.5% were widowed. As for employment, 21.5% were employed full time (40 h a week), 12.8% were employed part time (20 h a week), 8.8% were employed less than 20 h a week, 45.2% were unemployed, and 11.8% were looking for employment. The reasons for attendance to clinic were as follows: family medicine (11.3%); dental (28.8%); psychiatric/psychological (4.4%); and lab test, physical exam or other reasons (45.5%).

As determined by the MINI 6.0 [46], 38.3% of the sample met criteria for current (past year) Axis I psychopathology. In the current sample, the average number of diagnoses per participant was 0.94 (SD=1.1). The diagnoses were as follows: major depressive disorder (23.4%), post-traumatic stress disorder (5.8%), generalized anxiety disorder (5.5%), agoraphobia (4.4%), dysthymia (4.0%), substance use disorder (3.6%), alcohol use disorder (2.9%), social phobia (2.9%), obsessive–compulsive disorder (2.9%) and panic disorder (2.6%).

2.2. Interactive analyses

2.2.1. Depressive and suicidal symptoms

Covariates in the first step accounted for 47% of variance in depressive symptoms. Negative affectivity was the only significant covariate (β=.67, P<.001). There were significant main effects for both pain severity (β=.17, P<.001) and emotion dysregulation (β=.33, P<.001). In contrast to prediction, there was no significant interaction between pain severity and emotion dysregulation. In terms of suicidal symptoms, covariates accounted for 21% of variance, with negative affectivity as the only significant covariate (β=.43, P<.001). There was a significant main effect for emotion dysregulation (β=.25, P<.001) but not pain severity. As expected, there was a significant interaction of pain severity and emotion dysregulation (β=.79, P=.002). Simple slope analyses revealed that pain severity was related to greater levels of suicidal symptoms among individuals with higher (β=.21, P=.003) versus lower (β=−.11, P=.212) levels of emotion dysregulation (Fig. 1). The Johnson–Neyman technique indicated that pain severity was significantly and positively associated with suicidal symptoms (β=.12–.44, P<.05) for DERS-total scores of 85.23 or higher (21.90% of the sample; see Fig. 2).

Fig. 1.

Plotting the conditional effect of pain severity on suicidal and social anxiety symptoms, and number of mood/anxiety disorders. Note: GCPS (Graded Chronic Pain Scale); DERS (Difficulties in Emotion Regulation Scale); IDAS (Inventory for Depression and Anxiety Symptoms); Number of Mood/Anxiety Disorders based on the MINI (Mini International Neuropsychiatric Interview). Plots represent the association between pain severity and suicidal symptoms (top left) social anxiety symptoms (top right), number of mood/anxiety disorders (bottom left) at high and low values of DERS.

Fig. 2.

Location of change in significance of the conditional effect of emotion dysregulation on the association between pain severity and suicidal symptoms, social anxiety symptoms and number of mood/anxiety disorders. Note: The solid black lines illustrate the conditional effect of pain severity on IDAS-suicide (top left), IDAS-social anxiety (top right), and number of mood/anxiety disorders (bottom left) at values of the moderator (DERS). Dotted lines represent the 95% confidence interval for the effect. The grey area represents the regions of DERS scores where the respective associations are significant. There was no significant interaction for depressive symptoms.

2.2.2. Social anxiety and anxious arousal symptoms

Regarding social anxiety symptoms, covariates accounted for 32% of variance. Negative affectivity was the only significant covariate (β=.55, P<.001). There was a main effect of pain severity (β=.15, P=.007) but not emotion dysregulation. As expected, there was a significant interaction of pain severity and emotion dysregulation (β=.88, P<.001). Pain severity was related to greater levels of social anxiety among individuals with higher (β=.16, P=.004) versus lower (β=−.07, P=.407) levels of emotion dysregulation (Fig. 1). Examination of the Johnson–Neyman results indicated that pain severity was significantly and positively associated with social anxiety symptoms (β=.14–.53, P<.05) for DERS-total scores of 77.34 or higher (37.23% of the sample; Fig. 2). In predicting anxious arousal symptoms, covariates accounted for 39% of variance. Education (β=−.14, P=.004) and negative affectivity (β=.58, P<.001) were significant covariates. There was a main effect of pain severity (β=.24, P<.001) but not emotion dysregulation. In contrast to prediction, there was no significant interaction.

2.2.3. Depressive and anxiety disorders

For number of mood and anxiety disorders, covariates accounted for 25% of variance. Negative affectivity was the only significant covariate (β=.48, P<.001). As expected, there were significant main effects of pain severity (β=.19, P<.001) and emotion dysregulation (β=.27, P<.001), and the interaction term was also significant (β=.67, P= .007). Simple slope analysis indicated that pain severity was related to a greater number of mood and anxiety disorders among individuals with higher (β=.20, P<.001) versus lower (β=.02, P=.797) levels of emotion dysregulation (Fig. 1). Pain severity was positively associated with number of mood and anxiety disorders (β=.13–.49, P<.05) for DERS-total scores of 71.39 or greater (48.18% of the sample; Fig. 2).

3. Discussion

As hypothesized, the interaction between pain severity and emotion dysregulation was significantly related to suicidal and social anxiety symptoms as well as number of mood/anxiety disorders in this community health care Latino sample. Given that a sizable portion of variance was accounted for by covariates and main effects in the first two steps for these tests (see Table 2), the presence of a significant interaction improving the fit in predictive power of these models is clinically noteworthy [53]. Moreover, the form of the interactions followed the same expected pattern. Specifically, pain severity predicted greater suicidal and social anxiety symptoms as well as number of mood/anxiety disorders at higher, but not lower, levels of emotion dysregulation (Fig. 1). Thus, the overall pattern of findings was consistent with the posited theoretical model (e.g., [29,30]) suggesting that there may indeed be clinically relevant interplay between pain severity and emotion dysregulation in terms of an array of anxiety/depressive symptoms and disorders among Latinos in community health care settings. Specifically, these data indicate that lesser ability to adaptively regulate one’s emotions may serve to exacerbate the relation between pain severity and certain anxiety/depressive symptoms and disorders.

Table 2.

Main and interactive effects of pain and emotion dysregulation in predicting depressive, suicidal and social anxiety symptoms (n=274).

	β	t	P	R2 change
Depressive symptoms (IDAS)
Step 1
Ethnicity	−.03	−0.64	.522
Gender	.03	0.72	.475
Age	−.01	−0.08	.940
Years in United States	.06	1.16	.249
Education	−.08	−1.84	.066
Marital status	−.08	−1.71	.088
Negative affectivity (PANAS-NA)	.67	14.62	<.001	.47***
Step 2
Pain (GCPS)	17***	3.72	<.001
Emotion dysregulation (DERS)	.33***	6.98	<.001	12***
Step 3
Pain × emotion dysregulation (GCPS × DERS)	.15	0.76	.447	<.01
Suicidal symptoms (IDAS)
Step 1
Ethnicity	−.03	−0.61	.540
Gender	−.04	−0.76	.448
Age	−.02	−0.35	.725
Years in United States	.11	1.73	.084
Education	.02	0.44	.662
Marital status	−.07	−1.32	.188
Negative affectivity (PANAS-NA)	.43***	7.84	<.001	.21***
Step 2
Pain (GCPS)	.09	1.57	.118
Emotion dysregulation (DERS)	.25***	3.87	<.001	.06
Step 3
Pain × emotion dysregulation (GCPS × DERS)	.79**	3.06	.002	.03**
Social anxiety symptoms (IDAS)
Step 1
Ethnicity	.05	0.89	.372
Gender	.04	0.86	.392
Age	−.04	−0.72	.473
Years in United States	.07	1.22	.225
Education	−.04	−0.82	.412
Marital status	−.07	−1.28	.203
Negative affectivity (PANAS-NA)	.55***	10.77	<.001	.32***
Step 2
Pain (GCPS)	.15**	2.70	.007
Emotion dysregulation (DERS)	.08	1.40	.163	.03**
Step 3
Pain × emotion dysregulation (GCPS × DERS)	.88***	3.61	<.001	.03***
Anxious arousal symptoms (IDAS)
Step 1
Ethnicity	−.09	−1.83	.068
Gender	.04	0.90	.370
Age	−.05	−0.88	.382
Years in United States	.05	0.99	.322
Education	−.14**	−2.90	.004
Marital status	−.09	−1.80	.074
Negative affectivity (PANAS-NA)	.58***	11.98	<.001	.39***
Step 2
Pain (GCPS)	.24**	4.55	<.001
Emotion dysregulation (DERS)	−.01	−0.17	.862	.05***
Step 3
Pain × emotion dysregulation (GCPS × DERS)	−.15	−0.65	.518	<.01
Number of mood/anxiety disorders
Step 1
Ethnicity	−.04	−0.72	.474
Gender	.04	0.77	.441
Age	.05	0.84	.400
Years in United States	.03	0.47	.640
Education	−.02	−0.42	.675
Marital status	−.05	−0.87	.386
Negative affectivity (PANAS-NA)	.48***	9.01	<.001	.25***
Step 2
Pain (GCPS)	19***	3.40	<.001
Emotion dysregulation (DERS)	27***	4.55	<.001	10***
Step 3
Pain × emotion dysregulation (GCPS × DERS)	.67**	2.73	.007	.02**

^**P<.01.

^***P<.001.

Contrary to expectation, there was no significant interaction between pain severity and emotion dysregulation for depressive or anxious arousal symptoms. For depressive symptoms, a substantial portion variance was accounted for by variables entered in the first two steps (Table 2). Specifically, while negative affectivity was a significant predictor of each outcome, it accounted for more variance in depressive symptoms than the other criterion variables, a finding in line with past work demonstrating that depression exhibits a stronger relation to negative affectivity than other mood/anxiety disorders [34]. Yet, the lack of significant interaction for depressive and anxious arousal symptoms suggests that the interplay between pain severity and emotion dysregulation may not be uniformly applicable to all manifestations of mental health. The closer interconnection between anxious arousal and pain severity (both predominately somatic in origin) may be a possible reason for the lack of a significant interaction. Future work may benefit by examining present pain severity and emotion dysregulation in a model with other somatic anxiety processes, such as panic attacks.

Greater pain severity was associated with more severe depressive, social anxiety and anxious arousal symptoms and more mood/anxiety disorders. These findings add to the limited work among Latinos linking pain severity with depressive and anxiety symptoms [21,22] and uniquely extend such findings to include psychopathology among a community health care sample. Importantly, the effects of pain severity represent unique relations over and above demographic factors, negative affectivity and emotion dysregulation. However, there was no main effect for pain severity in relation to suicide symptoms. Recent research suggests that individuals with suicidal ideation may have more tolerance for pain [54]. Thus, it may be useful to understand the role of distress tolerance for pain in relation to suicidal symptoms among Latinos.

Emotion dysregulation also demonstrated unique associations with all dependent variables except social anxiety and anxious arousal symptoms. Greater levels of emotion dysregulation were associated with more severe depressive and suicide symptoms and a higher number of mood/anxiety disorders, findings that are consistent with past work [24,25], including Latino samples [26]. A lack of findings for social anxiety and anxious arousal was surprising given past work linking these symptoms to emotion dysregulation among non-Latino samples [55,56]. Future research should examine whether, among Latinos, emotion dysregulation has less of a direct impact on anxiety symptoms or whether the effect of pain severity (at the same level) minimized the variance to be accounted for by emotion dysregulation.

These findings highlight the importance of assessing emotion dysregulation in community health care settings among Latinos. Importantly, a growing body of work has shown the malleability of emotion dysregulation in response to interventions (e.g., [57,58]). Further, targeted therapies such as emotion regulation therapy [59] have shown promise in treating anxiety and depression via direct focus on emotion regulation enhancement. Future work should examine the efficacy/feasibility of such interventions in community health care settings, potentially in brief (e.g., one to two session) formats, and whether culturally appropriate modifications (e.g., [60]) are needed to suit unique needs of Latinos experiencing pain. Results of the current study suggest that decreasing emotion dysregulation (i.e., improving emotion regulation) may buffer against the adverse impact of pain and result in fewer symptoms/disorders.

Although not the primary aim of the study, the current study found ethnic differences in levels of emotion dysregulation and anxious arousal. Specifically, participants identifying as Mexican-American reported significantly lower emotion dysregulation and anxious arousal. These findings are consistent with the “Hispanic paradox,” a phenomenon wherein Latinos, specifically acculturated Mexican-Americans, demonstrate better health outcomes than other Latino ethnicities and European Americans despite barriers to care and other health inequities (e.g., [61]). Despite these differences, ethnicity (Mexican-American vs. other ethnicities) was included as a contrast covariate and did not significantly impact the pattern of findings.

Several limitations should also be noted. First, due to the cross-sectional nature of these data, it is not possible to make causal statements concerning the relations under study. One important next step would be to employ prospective methodologies to evaluate the consistency of the present findings over time. Another approach would be to experimentally manipulate pain severity in the laboratory to test singular and interactive effects with emotion dysregulation in terms of anxious and depressive reactions to theoretically relevant stressors (e.g., bodily sensations, discrimination). Second, the present Latino sample was largely female and seeking medical services for a wide range of issues. Future work could evaluate the generalizability of the present model to other sectors of the Latino community, including samples with a larger percentage of males and persons not seeking medical services. Third, Latinos often employ an extreme and acquiescent response style relative to other groups [62]. Accordingly, there is the possibility that a response bias may have influenced in the present observations, although this issue should be minimized in the current study because the tests conducted were within-group (rather than between-group) in nature. Fourth, we employed a general measure of pain severity. Therefore, it is not clear whether similar relations would be evident among specific clinical/chronic pain populations. Future work would benefit by exploring the generalizability of the current results to specific Latino pain populations. Fifth, we collapsed across anxiety and depressive disorders to index the number of diagnoses due to the sample size. Future work may benefit by testing whether the interactive model is particularly applicable to certain types of anxiety/depressive psychopathology. Sixth, we focused our investigation on general manifestations of anxiety/depressive phenomenology. However, it is possible that the same type of interactive model between pain intensity and emotion dysregulation is applicable to other more culturally specific forms of distress or stress. Future work may benefit by exploring the interactive model in relation to these processes, such as acculturation anxiety or subjective social status. Although Johnson–Neyman results in the current study provided ranges of emotion dysregulation where the association of pain intensity/mental health was significant, larger studies are needed to examine sensitivity and specificity of cutoff scores. Finally, the current study was descriptive in nature. Based upon these findings, there is a need for intervention-oriented work addressing emotion dysregulation in the context of pain.

Together, the present findings suggest that interventions that address anxiety/depressive symptoms and disorders among Latinos in community health care settings may benefit from screening for pain and emotion dysregulation. For example, tailoring community health-care-based treatments for Latinos with high levels of pain and emotion dysregulation may help isolate a high-risk segment of the Latino population for anxiety/depressive problems that could benefit from brief, psychosocial interventions that target reducing pain severity and emotion dysregulation through psychoeducation and skills training. In such work, there will likely be a need to further culturally adapt psychosocial intervention programs and explore their efficacy among Latinos attending community health care clinics.