Experiences of Discrimination and Urinary Catecholamine Concentrations: Longitudinal Associations in a College Student Sample

Abstract

Background

Experiences of discrimination are a risk factor for subsequent cardiovascular disease. However, there is a lack of longitudinal research examining associations between discrimination and urinary catecholamines. This is surprising given the likely mediating role of sympathetic nervous system dysregulation in the association between psychosocial stress and cardiovascular morbidity.

Purpose

The current study examined the 3 year longitudinal association between experiences of discrimination and urinary catecholamines.

Methods

The sample included 149 college students (mean age at baseline = 18.8, standard deviation = 0.96; 45% Black/African American; 55% White/European American). Concentrations of epinephrine and norepinephrine—urinary catecholamines with established links to psychosocial stress exposure and subsequent morbidity—were determined from 12 hr overnight samples.

Results

Results indicated that experiences of discrimination were associated with increases in both epinephrine (β = .284, standard error [SE] = .117, p = .015) and norepinephrine (β = .306, SE = .114, p = .001). These longitudinal associations persisted after adjusting for negative affect, depression, and rejection sensitivity and did not vary as a function of race/ethnicity.

Conclusions

Results suggest that examination of overnight urinary catecholamines as a biological mediator of associations between experiences of discrimination and cardiovascular morbidity is warranted.

Young adults who experienced more discrimination had a greater increase, over a three year period, in stress hormones that put them at risk for chronic disease.

Experiences of discrimination have been linked to a wide range of adverse physical health outcomes in both racial/ethnic minority and majority populations, including increased risk of cardiovascular disease morbidity [1–5]. However, the biological pathways through which experiences of discrimination influence heart disease remain unclear. Catecholamine concentrations—markers of sympathetic nervous system activation—are an important consideration in the context of discrimination and cardiovascular disease for several reasons. Chronic elevated sympathetic activation has established links to blood pressure [6, 7], which, in turn, is robustly associated with cardiovascular disease and premature mortality [8–10]. Furthermore, catecholamines have been associated with several other risk factors for heart disease, including dysregulated lipid metabolism [11], cardiovascular reactivity [12], and allostatic load [13–15]. Elevated catecholamines, such as norepinephrine and epinephrine, are, thus, important markers of risk for cardiovascular morbidity and mortality.

Experiences of discrimination have been associated with various indexes of physiologic dysregulation, including allostatic load, cardiovascular reactivity, and cortisol output [5, 13, 14, 16–18]. There have been extensive analyses on these topics with various meta-analyses documenting the effect of discrimination on biologic markers of health [18, 19]. However, to our knowledge, only one study has examined the association between discrimination and urinary catecholamines [20]. Results of this cross-sectional study indicated that experiences of discrimination did not mediate the associations between SES and catecholamine dysregulation. The sample consisted of Black and White adults, aged 40, and discrimination was measured using the Experiences of Discrimination Scale, which aggregates single-item measures across domains (e.g., at school, getting a job, and at work) [21]. With only one cross-sectional study reporting on this topic addressing the relationship between socioeconomic status and catecholamines mediated by discrimination, whether discrimination may be directly associated with catecholamines remains to be known.

The college years represent a time in which many key health and life outcomes are determined. In a review of disease epidemiology across the life course, blood pressure and body mass index (BMI) measurements taken in adolescence or early adulthood were found to be predictive of cardiovascular disease 50 years later [22]. Weight gain, decreases in physical activity, and increases in stress are often observed during the college years [23, 24]. Furthermore, young adults may establish lasting health behavior patterns, indicating the importance of examining health determinants in this developmental period [25].

The influence of stressors during college, including discrimination, can also have a lasting impact [26, 27]. Several studies suggest that college may be a time when experiences of discrimination are prevalent, particularly for minority students attending predominantly White universities [26, 28–30]. Isolation and alienation are commonly associated with these experiences and hold consequences for college retention [31–35]. However, college student populations remain understudied in the association between discrimination and health.

Longitudinal studies examining biological underpinnings of associations between discrimination and health are also scarce. Increases in catecholamines are associated with weight gain, blood pressure elevation, and cardiovascular disease [6, 36, 37]. These conditions also develop over time, indicating the importance of including longitudinal measures of catecholamines that can help to elucidate the long-term trajectories of the diseases. As a chronic stressor, discrimination may also proliferate throughout the life course, leading to stronger prospective effects [38, 39]. However, longitudinal associations between experiences of discrimination and catecholamines have not been previously studied.

With respect to differences by race/ethnicity in the effects of experiences of racial discrimination on health outcomes, reported findings have been mixed [40–42]. Although some studies have found significant differences in the strength of the association between racial discrimination and health across racial/ethnic groups [43, 44], a number of studies have reported no significant differences between African Americans and European Americans [42, 45, 46]. The frequency of discrimination is, however, consistently higher in African Americans than European Americans across studies [44, 47].

The primary aim of this study was to elucidate the longitudinal effects of discrimination on catecholamines in a college sample. Our main hypothesis was that experiences of discrimination would be associated with elevated catecholamine concentrations and would predict increases in catecholamines across the college years. Race was also examined as a moderator of associations between discrimination and catecholamines; however, based on prior research, we did not hypothesize a significant moderation effect.

Methods

Participants

Data were derived from the College Student Health Study [47, 48]. Participants were recruited from a large predominantly White university (73% White/European American, 4% Asian, 4% Hispanic/Latino, 3% Black/African American, and 1% American Indian). All first and second year African-American students and an equally sized stratified random sample of European-American students were invited to participate. The initial assessment (T1) took place in the clinical research unit of an on-campus university hospital in 2012 where researchers and nursing staff took physiologic measures and self-report measures were administered on a laptop computer. Overnight urine samples were also collected by participants in their homes on the night prior to the hospital visit and transported to the visit in an insulated container with cold packs. A total of 150 students participated in the baseline assessment. The analytic sample for the current study (N = 149; mean age at baseline = 18.8, standard deviation = 0.96; 45% African American, 55% European American) included all but one participant from the baseline sample who had missing urinary catecholamine data at T1. Two thirds of the baseline sample (n = 97) also participated in a follow-up survey assessment (T2) administered 3 years later, and 81 of these participants had urine samples assayed at T2. The protocol was approved by a university’s institutional review board and informed consent was obtained from all individual participants included in the study.

Measures

Experiences of discrimination

Discrimination was assessed using 13 items from the Racism and Life Experiences Scale (RaLES) [49]. Items were selected from the original 20-item scale due to their unidimensionality and high correlations with the total scale score. This shortened version of the original measure has been used in prior research [47]. Respondents indicated how often in the past semester they had experienced each of the 13 types of unfair treatment because of their race or ethnicity (e.g., “How often have you been treated as if you were stupid or talked down to?” and “How often have you been left out of conversations or activities?” [49]. Response options were on a six-point scale ranging from never (coded as 1) to several times a day (coded as 6). The RaLES has been used extensively in research on discrimination, and psychometric properties have been documented [41, 50]. Items were averaged to create a composite score, with higher values indicating higher levels of racial discrimination (α = .91).

Catecholamines

Supplies for 12 hr overnight urine collection were dropped off to participants on the day prior to their study visit. Collection instructions were described face-to-face and written instructions were provided. Start time of urine collection was set 12 hr before normal wake time. The participant was instructed to void at the start time and collect all urine until the stop time, being sure to include the first void upon morning awakening and any voids during the night. Reminder labels were also provided to avoid missed samples.

Participants were instructed to store the urine collection container in their refrigerator at all times and bring the container in a provided insulated package with cold packs to their appointment the next day. University of Wisconsin Health Laboratory Services conducted initial sample processing and then sent the samples to ARUP Laboratories to complete assay procedures following established protocols. Samples were acidified to stabilize the analyses of interest and assays were conducted using quantitative high-performance liquid chromatography/tandem mass spectrometry [51]. Intra-assay and interassay coefficients of variation for samples at ARUP are 6.4% and 7.1% for epinephrine and 6.2% and 6.8% for norepinephrine, respectively [52]. Norepinephrine and epinephrine values were adjusted for creatinine concentrations in order to account for differences in urine dilution and measured in micrograms of catecholamine/gram of creatinine (µg/g).

Substance use

Students reported whether they currently smoke cigarettes (1 = yes, 0 = no) and the number of alcoholic drinks they consume on a typical day when they are drinking. Binge drinking was coded as 1 if participants reported consuming five or more drinks on a “typical day when drinking” and 0 otherwise [53].

Socioeconomic disadvantage

Income-to-needs ratio was calculated from youth reports of their parents’ combined household income from the year prior to the baseline assessment [54]. This value was divided by the federal poverty threshold for the household size using 2011 U.S. Federal Poverty Guidelines.

Psychological covariates

Depressive symptoms were assessed using the Beck Depression Inventory II [55], a validated and internally consistent 21-item scale (α = .82). Negative affects were assessed on the Positive and Negative Affect Schedule. This scale is an adjective checklist on which individuals use a five-point Likert-type scale to rate the degree to which they have experienced 20 mood states over the past 30 days. This study used a shortened version of the scale using the mean of five of the mood states (α = .66). Published data support the reliability and validity of this scale [56]. Rejection sensitivity was measured using six items from the racial rejection sensitivity questionnaire [34]. The six items were scored using an established algorithm to form a composite (α =.93).

Body mass index

Trained nursing staff collected anthropometric measurements of height and weight. BMI was calculated as weight (kilograms) divided by height squared (square meters).

Demographic variables

Age and sex were reported by participants. Race was taken from university records and confirmed by student self-reports.

Analyses and Missing Data

Regression analyses were conducted in Mplus (Version 8.0) using the maximum likelihood estimator. Independent samples t-tests were used to test for differences on continuous variables between groups. A series of linear regression models were fit considering norepinephrine and epinephrine at T2 as separate outcomes. Models adjusted for earlier levels of each catecholamine at T1. Analytic models included all 149 individuals who participated in the Wave 1 assessment, with 1 participant excluded from analyses because of missing catecholamine data at both time points. Missing data were dealt with using full-information maximum likelihood estimation. Of the 149 individuals included analyses, for T1 variables, 6% had missing data for income to poverty ratio, 2% for experiences of discrimination, and 3% for rejection sensitivity and, for T2 variables, 46% had missing data for both norepinephrine and epinephrine (81 of the original 149 participants provided overnight urine samples at both time points). Missing urine data at the 3 year follow-up assessment was primarily due to attrition (97 out of 149 participated in the follow-up). Reasons for attrition included no longer residing in the area or nonresponse to contacts for follow-up participation (e.g., due to changes contact information). An electronic error also led to loss of data for 17 urine samples (results were faxed to a machine that was not functional), leaving 81 with valid urinary data at both time points. Missingness of catecholamine data were not significantly correlated with any study variables; thus, any bias due to missing data in reported results is expected to be minimal. To further probe this assumption, additional analyses were conducted to examine whether the reported results differed when including only those who had catecholamine data at both T1 and T2 (see Supplementary Tables S1 and S2). The general pattern of findings and significance was equivalent to those reported in primary analyses. Specifically, the longitudinal association between discrimination and catecholamines remained statistically significant and was similar in magnitude. Additional models were also examined to consider nonlinear associations between experiences of discrimination and catecholamines. However, no evidence for nonlinear effects was found.

Results

Table 1 provides descriptive statistics and correlations for key variables. Norepinephrine and epinephrine levels were within normal ranges for urinary catecholamines adjusted for creatinine [57]. Participant age was correlated with norepinephrine at T2 but not norepinephrine at T1. Experiences of discrimination were correlated significantly with each catecholamine except for T1 epinephrine. Gender was correlated with norepinephrine at both time points.

Table 1.

Descriptive statistics and zero-order correlations for study variables

	1	2	3	4	5	6	7	8	9	10
1. Sex (Female)	–
2. Age (T1)	.215**	–
3. Race (Black)	.079	.080	–
4. Norepinephrine (T1, μg/g)	.253**	−.054	.108	–
5. Norepinephrine (T2, μg/g)	.272*	−.272*	.067	.417***	–
6. Epinephrine (T1, μg/g)	−.052	.017	.047	.465***	.105	–
7. Epinephrine (T2, μg/g)	−.054	-.054	.122	.257*	.724***	.183	–
8. Exp. of discrimination (T1)	.140	.105	.629***	.177*	.253*	.108	.226*	–
9. Cigarette smoker	−.032	.220**	.025	.026	.218	−.034	.161	.050	–
10. Heavy drinker	−.206*	.084	−.256**	.066	−.102	.057	−.005	−.187*	.117	–
Mean (%)	(56)	18.81	(45)	22.64	25.60	2.95	3.15	1.58	(7)	(39)
SD		0.96		8.76	13.02	1.93	3.26	0.65

SD standard deviation; T1 baseline assessment; T2 3 year follow-up assessment.

*p < .05, **p < .01, ***p < .001.

Women excreted higher levels of norepinephrine at both T1 (t = 3.17, p = .003) and T2 (t = 2.52, p = .014); this is consistent with previous research [57]. Women were also less likely to be heavy drinkers (t = −2.57, p = .012). African-American students had higher BMI (t = 2.49, p = .021), lower parental income (t = −3.18, p = .002), were more likely to be depressed (t = 3.76, p ≤ .001), less likely to be heavy drinkers (t = −3.22, p = .001), and scored higher on anticipation of racial profiling (t = 18.04, p ≤ .001). For these reasons, age and sex were included as covariates in each model and health behaviors and psychological covariates were tested in later models.

African-American college students reported substantially more experiences of discrimination at T1 (t = 9.72, p ≤ .001). Approximately 98.5% of African-American and 71.3% of European-American students reported experiencing at least one discrimination event during the past semester. 74.2% of African-American and 32.1% of European-American students reported at least one discrimination event happening “About once a week” during the past semester. The most common discrimination events included “How often have you overheard or been told an offensive race-related joke or comment?” and “How often have you witnessed someone showing a lack of respect for people of your race/ethnicity?”

Experiences of Discrimination and Catecholamines

In regression analyses, discrimination was associated with norepinephrine at T2 adjusting for T1 norepinephrine (β = .306; p = .001; see Table 2, Model 2). Discrimination was associated with epinephrine at T2, adjusting for T1 epinephrine (β = .284 p = .015; see Table 3, Model 2). Longitudinal associations between experiences of discrimination and catecholamines remained after adjusting for depression, negative affect, and race rejection sensitivity (Tables 2 and 3, Model 3). The magnitude of the longitudinal association between discrimination on catecholamines increased by 36.6% for norepinephrine and 33.1% for epinephrine after adjusting for depression, negative affect, and race rejection sensitivity.

Table 2.

Results of longitudinal regression models examining predictors of urinary norepinephrine at Time 2

	Model 1		Model 2		Model 3
	B	SE	B	SE	B	SE
Intercept	7.339***	2.578	6.971**	2.607	7.281**	2.978
Norepinephrine at Time 1	0.391***	0.100	0.358***	0.094	0.414***	0.102
Age	−0.319**	0.115	−0.332**	0.115	−0.359**	0.107
Sex (female)	0.093	0.116	0.035	0.107	0.010	0.111
Race (Black)	0.031	0.093	−0.136	0.100	−0.008	0.162
Heavy drinker	−0.026	0.097	−0.051	0.090	−0.073	0.096
Cigarette smoker	0.171	0.134	0.145	0.113	0.136	0.106
BMI	−0.072	0.091	−0. 049	0.082	−0.053	0.095
Income-to-needs ratio	0.038	0.117	0.040	0.108	0.004	0.106
Experiences of discrimination			0.306**	0.114	0.418**	0.130
Depression					−0.132	0.100
Negative affect					0.077	0.118
Rejection sensitivity					−0.260	0.190

All continuously coded predictor variables were standardized to have a mean of 0 and standard deviation of one in all models.

BMI body mass index; SE standard error.

*p < .05, **p < .01, ***p < .001.

Table 3.

Results of longitudinal regression models examining predictors of urinary epinephrine at Time 2

	Model 1		Model 2		Model 3
	B	SE	B	SE	B	SE
Intercept	3.595	3.122	3.568	3.067	3.551	3.062
Epinephrine at Time 1	0.161	0.151	0.111	0.152	0.146	0.152
Age	−0.146	0.139	−0.171	0.136	−0.175	0.135
Sex (female)	−0.103	0.120	−0.155	0.104	−0.188	0.103
Race (Black)	0.074	0.107	−0.075	0.108	0.088	0.226
Heavy drinker	0.005	0.107	−0.025	0.103	−0.058	0.117
Cigarette smoker	0.166	0.139	0.142	0.129	0.118	0.119
BMI	0.021	0.099	0.040	0.093	0.067	0.102
Income-to-needs ratio	−0.063	0.108	−0.063	0.103	−0.079	0.102
Experiences of discrimination			0.284**	0.117	0.378*	0.155
Depression					0.002	0.101
Negative affect					−0.033	0.118
Rejection sensitivity					−0.296	0.289

All continuously coded predictor variables were standardized to have a mean of 0 and standard deviation of one in all models.

BMI body mass index; SE standard error.

*p < .05, **p < .01.

While there were differences in experiences of discrimination for African-American and European-American college students, no differences were present in catecholamine levels at either time point. In sensitivity analyses, Race × Discrimination interaction terms were tested alongside the main effects of each variable and no significant moderation was detected for norepinephrine or epinephrine.

Discussion

The current study examined longitudinal associations between experiences of discrimination and catecholamines in a college student sample. Findings showed that discrimination was associated with increases in norepinephrine and epinephrine over a 3 year period and that these associations persisted after adjusting for negative affect, depression, and rejection sensitivity. The findings are consistent with cross-sectional research documenting associations between discrimination and allostatic load [13, 14, 16] and between discrimination and disease morbidity [58]. The results also extend this work by showing that chronic overactivation of the sympathetic nervous system may be one important mechanism for established associations between discrimination and indexes of physiologic dysregulation.

The one other study that has considered the association between discrimination and catecholamines found no significant association and, thus, is somewhat in contrast to our results [20]. However, differences in the population under study (college students vs. adults), measure of discrimination (racial discrimination vs. general unfair treatment), and study design (longitudinal vs. cross-sectional) prelude the possibility of detailed comparison and suggest the need for additional research to determine the contexts in which perceived discrimination is most associated with indicators of sympathetic nervous system activation. This study is the first to examine these associations in a young adult sample, with a focus on racial discrimination and using longitudinal catecholamine data.

Repeated exposure to anxiety-inducing events have been shown to progressively increase levels of norepinephrine [59]. Results of this study suggest that discrimination may operate in a similar manner. Furthermore, because catecholamines potentiate vasoconstriction and have been associated with hypertension, it is likely that elevated sympathetic activity could be operating as a mediator of established associations between discrimination and hypertension, as well as related cardiometabolic risk factors [60, 61]. One clear direction for future research will, therefore, be to explicate these mediating pathways.

Results of longitudinal analyses indicated that students experiencing higher levels of discrimination have greater increases in nocturnal sympathetic activation across the college years relative to peers who did not experience discrimination. As African-American students experience higher levels of discrimination, another future direction is additional research to examine whether cumulative discrimination during emerging adulthood over longer periods of time may account for race differences in adult sympathetic nervous system activation.

Although educational attainment is known to promote health and psychosocial adjustment [62], the process of obtaining higher education often comes with additional stressors for minority students that can partially offset these benefits [47, 63–65]. These additional stressors are also more prevalent at predominantly White universities [30, 66, 67]. The significance of this point is underscored by the fact that the college years are a salient time for establishing health behaviors and coping strategies that can have notable consequences across the life span [25]. In line with cumulative stress theory, stressors during college may compound, influencing cardiovascular functioning and mortality [68–70]. Evaluation of programs to support students who are part of negatively stigmatized or underresented groups is, thus, a promising direction for future work [28, 30, 71].

While several studies have found that African-American individuals experience more frequent discrimination than European Americans, there have been mixed findings surrounding the associations between discrimination and health outcomes across races [42, 45, 46]. Results of this study add to this body of research, suggesting that racial discrimination as a stressor has deleterious consequences for both minority and majority racial/ethnic groups. While African-American students reported more discrimination experiences at T1, the effects on health were similar across race, indicating that, while there may be more racial discrimination experiences reported by minority students, the effects of the perceived racism are similar across races. Why measures of unfair treatment and discrimination are just as predictive of adverse health outcomes among European Americans as among African-American individuals is not completely clear and warrants further investigation. One possible explanation for reports of discrimination among European Americans is that changes in demographics of the United States may have caused some European Americans to perceive that they are a part of a marginalized minority group. Consistent with this perspective, studies have shown that European Americans living in areas of minority populations report more cases of discrimination [72]. Additionally, our results are consistent with prior research suggesting that many Whites report significant levels of racism, which may be due to conservative ideologies that are in opposition to programs designed to assist historically disadvantaged groups [73]. Last, the presence of diversity policies have implications for minority students at predominantly White institutions, including decreased reporting of discrimination and an increased sense of belonging, possibly leading to reduced associations between discrimination and health [74]. The relevance of this interpretation in the context of college students at a predominantly White institution is unclear and is worthy of further examination.

Consideration of depression, negative affect, and rejection sensitivity also provides insight into the reported results. Results of these analyses suggested that personality and mental health constructs are not driving the associations between experiences of discrimination and catecholamines. Accounting for effects of psychological well-being when considering effects of discrimination is important as predisposition to report negative experiences and having heightened sensitivity are a central critique of research on racial disparities [75]. As the results did not substantively change when the psychological constructs were included, confidence in the findings is bolstered.

Some limitations of the current study should also be noted. The study focused on a sample of African-American and European-American college students at a Midwestern predominantly White university. The degree to which findings may be similar in other samples and among other racial/ethnic groups remains to be determined and will be an important direction for future research. In addition, only one measure of discrimination is examined in this study, limiting the ability to understand all nuances of racial discrimination experiences and their relation to catecholamines. Various measures of discrimination have been used in previous studies to capture the varying nature of these experiences, addressing vicarious discrimination [76], daily hassles or microaggressions versus major experiences of discrimination [77], and the combination of discrimination with other stressors [78]. Perceptions of racism at an institutional level and the transgenerational transmission of racism-related stress can also have impacts on health [49]. Each of these contexts of discrimination are important to understand in the development of measures of racial discrimination and future studies should explore the many facets of these experiences within the context of the association between discrimination and health.

Additional research is needed to further explicate the psychosocial mechanisms for the effects of discrimination on health. For example, peer social support, which has been found to mitigate cardiovascular reactivity to various stressful events [79] or family and community support, which was found to buffer the effects of racial discrimination on health outcomes [13, 80], are both important constructs worthy for additional consideration. These, along with other societal and institutional approaches to addressing the effects of discrimination, are key areas of research as a growing literature examines the effects of discrimination on health outcomes.

Results of this study indicate that, in a college student sample, discrimination is associated with increases in norepinephrine and epinephrine over a 3 year period and that these associations persisted after adjusting for negative affect, depression, and rejection sensitivity. The findings suggest that research examining chronic sympathetic nervous system overactivation as a mediator of established associations between discrimination and cardiometabolic risk are warranted.

Acknowledgments

Funding: Support for this project was provided by the Robert Wood Johnson Health and Society Scholars Program at the University of Wisconsin-Madison and by the Clinical and Translational Science (CTSA) Award program through the National Institutes of Health National Center for Advancing Translational Sciences (NCATS) (grant UL1TR000427).

An earlier version of this paper was presented at the Annual Meeting of the American Psychosomatic Society, March 2018.

Compliance With Ethical Standards

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Authors’ Statement of Conflict of Interest and Adherence to Ethical Standards The authors declare that they have no conflict of interest.

Authors’ Contributions L.K.H. conceptualized the research questions, analyzed the data, and wrote the manuscript. T.E.F.R. designed the study, directed data collection, assisted with data analysis, and contributed to writing of the manuscript.

Ethical Approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed Consent Informed consent was obtained from all individual participants included in the study.