Introduction
The Latina/x/o population is the largest, fastest-growing, and youngest ethnic/racial minority group in the United States (U.S.; Funk & Lopez, Reference Funk and Lopez2022). They currently account for 19% of the total population in the U.S. (U.S. Census Bureau, 2019) and are projected to grow to 29% by 2060 (U.S. Census Bureau, 2017). It is also a population defined by its youth, with nearly half of U.S.-born Latina/x/o being younger than the age of 18 (Patten, Reference Patten2016). However, Latina/x/o youth face salient disparities in mental and sleep health compared to their White counterparts. Specifically, Latina/x/o youth are at a greater risk for depressive symptoms and anxiety disorders (Isasi et al., Reference Isasi, Rastogi and Molina2016), and Latina/x/o youth experience shorter sleep duration, poor sleep quality, and higher rates of sleep disturbances than their White counterparts (Yip et al., Reference Yip, Cheon, Wang, Cham, Tryon and El-Sheikh2020). Understanding and addressing such health disparities for Latina/x/o youth is paramount to the future health outlook of the Latina/x/o population, as research suggests that the health status of adolescents has significant implications for health during adulthood (Paula Braveman & Barclay, Reference Braveman and Barclay2009). Therefore, it is critical to identify early risk factors that contribute to health disparities.
One salient contributor to health disparities for Latina/x/o youth is their experiences of discrimination (Stein et al., Reference Stein, Gonzales, García Coll, Prandoni, Crockett and Carlo2016). It is well established that Latina/x/o youth frequently experience discriminatory treatment in various social contexts, and discrimination is a pervasive source of stress in their daily lives (Delgado et al., Reference Delgado, Nair, Zeiders, Jones, Fitzgerald, Johnson, Qin, Villarruel and Norder2019). Being exposed to discrimination is consistently found to be associated with poor mental and physical health such as depressive symptoms, anxiety, hypertension, and sleep dysregulation (Benner et al., Reference Benner, Wang, Shen, Boyle, Polk and Cheng2018; Pascoe & Smart Richman, Reference Pascoe and Smart Richman2009; Yip, Cheon, et al., Reference Yip, Cheon, Wang, Cham, Tryon and El-Sheikh2020). The biopsychosocial model of minority health (Myers, Reference Myers2009) posits that discrimination influences health through wear and tear on the systems and tissues in the body, which is linked to the biological stress responses, including the hypothalamic-pituitary-adrenal (HPA) axis, which is a primary stress response system in the body (Clendinen & Kertes, Reference Clendinen and Kertes2022). Emerging evidence suggests that experiences of discrimination can alter the secretion of cortisol, the end product of the HPA axis, as well as diurnal cortisol rhythms (Huynh et al., Reference Huynh, Guan, Almeida, McCreath and Fuligni2016; Korous, Reference Korous2017; Seaton & Zeiders, Reference Seaton and Zeiders2020), but whether the daily processes between discrimination and diurnal cortisol response influence overall health remains unanswered. To fill this gap in the literature, the current study integrates both daily diary and post-diary survey data to examine whether diurnal cortisol responses following discriminatory experiences influence Mexican-origin adolescents’ mental health (i.e., depressive symptoms, anxiety) and sleep health (i.e., sleep quality, sleep duration). As Mexican heritage accounts for the majority (62%) of Latina/x/o in the U.S. (Funk & Lopez, Reference Funk and Lopez2022), the current study focuses on examining the development of Mexican-origin youth.
Discrimination, diurnal cortisol rhythms, and health
Discrimination, a form of unfair and negative treatment based on an individual’s real or perceived differences, has been consistently linked to a wide range of mental and physical health problems including depression, anxiety, cardiovascular disease, hypertension, and sleep dysregulation (Benner et al., Reference Benner, Wang, Shen, Boyle, Polk and Cheng2018; Pascoe & Smart Richman, Reference Pascoe and Smart Richman2009; Slopen et al., Reference Slopen, Lewis and Williams2016). Moreover, growing evidence has shown that discrimination likely explains health disparities not attributable to socioeconomic status (Williams, Reference Williams1999). The biopsychosocial model of minority health (Myers, Reference Myers2009) suggests that discrimination can get under the skin to influence health through elevating biological stress processes. One of the major stress response systems in the body is the HPA axis (Johnson et al., Reference Johnson, Kamilaris, Chrousos and Gold1992). When an individual encounters a stressor, the HPA axis is activated to prepare the body for the “fight or flight” response by triggering a cascade of stress hormones, including corticotrophin-releasing hormone in the hypothalamus, adrenocorticotropic hormone in the pituitary gland, and glucocorticoid hormones (cortisol) from the adrenal cortex (Gunnar & Quevedo, Reference Gunnar and Quevedo2007). Cortisol is the end product of the HPA axis, and it plays a key role in the metabolic system, the immune system, and the central nervous system in the body (Sapolsky et al., Reference Sapolsky, Romero and Munck2000). Although short-term activation of cortisol response is adaptive to cope with acute stress, frequent and chronic activation of cortisol responses can create wear and tear on the body, contributing to elevated allostatic load (McEwen, Reference McEwen2003).
Most of the early work on cortisol has focused on cortisol reactivity, which refers to the response process of the HPA axis to immediate stressors. For example, researchers have used the laboratory-based stress paradigm (e.g., Tier Social Stress Test) to elicit cortisol responses (Dickerson & Kemeny, Reference Dickerson and Kemeny2004). More recent research has moved beyond the reactivity framework and focuses on the basal activity of the HPA axis in naturalistic settings (Adam et al., Reference Adam, Quinn, Tavernier, McQuillan, Dahlke and Gilbert2017; Seaton & Zeiders, Reference Seaton and Zeiders2020). Basal cortisol activity follows a strong diurnal rhythm, which is regulated by the body’s internal circadian clock. The diurnal rhythm is characterized by high cortisol levels upon waking, a surge in the first 30–40 min after waking, and a decline throughout the day, reaching a nadir around bedtime (Pruessner et al., Reference Pruessner, Wolf, Hellhammer, Buske-Kirschbaum, von Auer, Jobst, Kaspers and Kirschbaum1997). This diurnal rhythm of cortisol is important for regulating various physiological processes in the body, including metabolism, immune response, and cardiovascular activity (Chung et al., Reference Chung, Son and Kim2011). High levels of cortisol in the morning help to promote wakefulness and alertness, while lower levels in the evening help to prepare the body for sleep (Adam et al., Reference Adam, Quinn, Tavernier, McQuillan, Dahlke and Gilbert2017).
To quantify diurnal patterns, multiple parameters have been examined in prior research, including waking cortisol, bedtime cortisol, the size of the cortisol awakening response (CAR), and the degree of change in cortisol from morning to evening (diurnal slope). Specifically, waking cortisol refers to the cortisol level upon wakening and elevated waking cortisol levels have been associated with mental health problems (e.g., depression; Mannie et al., Reference Mannie, Harmer and Cowen2007). Bedtime cortisol refers to the cortisol level before bedtime and high levels of evening cortisol secretion has been linked to mental health disorders and sleep disturbances (Kamali et al., Reference Kamali, Saunders, Prossin, Brucksch, Harrington, Langenecker and McInnis2012; Kumari et al., Reference Kumari, Badrick, Ferrie, Perski, Marmot and Chandola2009). The CAR refers to the rise in cortisol immediately post-awakening and prepares the body for the day’s challenges (Pruessner et al., Reference Pruessner, Wolf, Hellhammer, Buske-Kirschbaum, von Auer, Jobst, Kaspers and Kirschbaum1997), with the magnitude of CAR associated with psychosocial stress and health, although the direction of the effects remains inconclusive (Boggero et al., Reference Boggero, Hostinar, Haak, Murphy and Segerstrom2017). For example, research has shown that an elevated CAR is linked to chronic stress, general life stress, and depression (Chida & Steptoe, Reference Chida and Steptoe2009; Wüst et al., Reference Wüst, Federenko, Hellhammer and Kirschbaum2000), while other studies have found that lower levels of CAR are associated with depression, fatigue, and burnout (Chida & Steptoe, Reference Chida and Steptoe2009; O’Donnell et al., Reference O’Donnell, Badrick, Kumari and Steptoe2008). Diurnal cortisol slope is used to capture the rate of decline from waking levels of cortisol to levels assessed before bedtime. Flattening patterns of diurnal slopes with lower morning and higher evening cortisol levels are maladaptive and have been associated with poor overall mental and physical health such as loneliness, depression, fatigue, sleep disturbance, and cardiovascular disease (Adam et al., Reference Adam, Quinn, Tavernier, McQuillan, Dahlke and Gilbert2017; Doane & Adam, Reference Doane and Adam2010; Kumari et al., Reference Kumari, Badrick, Ferrie, Perski, Marmot and Chandola2009). The current study aims to comprehensively examine cortisol activity by including these four cortisol parameters (i.e., waking cortisol, bedtime cortisol, CAR, and cortisol slope).
The bourgeoning research examining the link between discrimination and the HPA axis is thus far inconclusive. Research emerging from animal models argues that chronic stressors that are social-evaluative and uncontrollable, such as discrimination, can activate diurnal cortisol secretion as a way to prepare individuals to mobilize resources and adjust to the situations in which their social standing is threatened (Dickerson & Kemeny, Reference Dickerson and Kemeny2004; Selye, 1956). However, another group of researchers proposed an attenuation hypothesis (Susman, Reference Susman2006; Trickett et al., Reference Trickett, Noll, Susman, Shenk and Putnam2010), which posits that the HPA axis adapts to chronic stress such as discrimination by downregulating cortisol secretion, which serves as a way to prepare the body for later stress. Empirical evidence linking discrimination and HPA axis functioning has yielded similarly conflicting results. Specifically, some studies have found that discrimination is associated with altered physiological functioning that may indicate poor cortisol functioning, such as lower waking cortisol (Huynh et al., Reference Huynh, Guan, Almeida, McCreath and Fuligni2016), more pronounced CAR (Doane & Zeiders, Reference Doane and Zeiders2014; Zeiders et al., Reference Zeiders, Doane and Roosa2012), flatter diurnal slope (Huynh et al., Reference Huynh, Guan, Almeida, McCreath and Fuligni2016; Skinner et al., Reference Skinner, Shirtcliff, Haggerty, Coe and Catalano2011; Zeiders et al., Reference Zeiders, Hoyt and Adam2014), and greater overall diurnal output (Huynh et al., Reference Huynh, Guan, Almeida, McCreath and Fuligni2016; Seaton & Zeiders, Reference Seaton and Zeiders2020; Zeiders et al., Reference Zeiders, Doane and Roosa2012), whereas others have observed associations between discrimination and altered physiological functioning that may indicate adaptive cortisol functioning, including higher waking levels (Fuller-Rowell et al., Reference Fuller-Rowell, Doan and Eccles2012), less pronounced CAR (Seaton & Zeiders, Reference Seaton and Zeiders2020), steeper diurnal slope (Fuller-Rowell et al., Reference Fuller-Rowell, Doan and Eccles2012; Seaton & Zeiders, Reference Seaton and Zeiders2020), and lower overall diurnal output (Kaholokula et al., Reference Kaholokula, Grandinetti, Keller, Nacapoy, Kingi and Mau2012; Lee et al., Reference Lee, Eisman, Stoddard, Peckins, Goldstick, Hsieh, Muñoz-Velázquez and Zimmerman2018). There are also studies that have yielded null findings between discrimination and diurnal slope (Doane & Zeiders, Reference Doane and Zeiders2014; Martin et al., Reference Martin, Bruce and Fisher2012), CAR (Zeiders et al., Reference Zeiders, Hoyt and Adam2014), and overall secretion throughout the day (Brody et al., Reference Brody, Lei, Chae, Yu, Kogan and Beach2014; Lee et al., Reference Lee, Peckins, Heinze, Miller, Assari and Zimmerman2018).
A recent qualitative review (Busse et al., Reference Busse, Yim, Campos and Marshburn2017) on discrimination and HPA axis functioning and a meta-analysis (Miller et al., Reference Miller, Chen and Zhou2007) on chronic stress and HPA axis functioning both suggest that, depending on the timing and onset of discrimination experiences, cortisol activation and attenuation might both come into play. As such, cortisol activity is activated at the onset of discrimination experience as indicated by steeper cortisol diurnal slopes, higher waking cortisol, less pronounced CAR, and lower bedtime cortisol. Such short-term activation of cortisol response is adaptive and beneficial for everyday function (Adam & Kumari, Reference Adam and Kumari2009). As time passes, however, chronic experiences of discrimination attenuate cortisol secretion to below normal levels as indicated by flatter diurnal cortisol slopes, lower waking cortisol, more pronounced awakening response, and higher bedtime cortisol. Chronic activation of cortisol response is associated with negative health outcomes (Adam & Kumari, Reference Adam and Kumari2009). Therefore, the mixed empirical findings may be due to the fact that most prior studies asked participants to report the incidence of discrimination within the past year or lifetime, leaving the time intervals between the onset of discrimination event and measures of HPA activity varying to a large extent. Given that timing and discrimination onset matter in the link between discrimination and the HPA axis, scholars have called for studies that use shorter intervals from stress onset to better disentangle the mixed findings (Korous, Reference Korous2017; Miller et al., Reference Miller, Chen and Zhou2007). Another possible reason for the inconsistent empirical findings is that most prior research focuses on examining between-person (BP) associations (i.e., nomothetic approach) between discrimination and HPA activities. Given that individuals often respond to stress in an idiosyncratic way, it is important to examine the within-person (WP) associations (i.e., idiographic approach) between discrimination and HPA activity.
To address these limitations, the current study adopts a daily diary approach, which captures the snapshot of an individual’s everyday life and permits the exploration of WP associations. A few empirical studies have utilized this approach to examine the association between discrimination and the HPA axis functioning, and these have generally involved adult populations (e.g., Seaton & Zeiders, Reference Seaton and Zeiders2020; Zeiders et al., Reference Zeiders, Landor, Flores and Brown2018). For example, Seaton and Zeiders (Reference Seaton and Zeiders2020) found that on days when Black college students experienced racial discrimination, they exhibited greater overall cortisol output the same day as well as less pronounced awakening response and steeper diurnal slope the next day. The daily processes of discrimination and HPA axis functioning remain relatively underexplored among adolescents, as only three studies have examined the link between discrimination and HPA activity during adolescence (Doane & Zeiders, Reference Doane and Zeiders2014; Huynh et al., Reference Huynh, Guan, Almeida, McCreath and Fuligni2016; Zeiders et al., Reference Zeiders, Doane and Roosa2012; results of these studies are described above). Adolescence, however, is a salient developmental period marked by the onset of puberty, which triggers a series of biological processes including hormonal and neuroendocrine changes, and the HPA axis is particularly sensitive and responsive to stressors during puberty (Romeo, Reference Romeo2013). Simultaneously, adolescents undergo rapid cognitive, emotional, and social changes, which make them more sensitive to social-evaluative stressors such as discrimination (Dickerson & Kemeny, Reference Dickerson and Kemeny2004; Romeo, Reference Romeo2013). Given that previous research has found spillover effects of daily experiences on cortisol responses (Adam et al., Reference Adam, Hawkley, Kudielka and Cacioppo2006; Doane & Adam, Reference Doane and Adam2010; Seaton & Zeiders, Reference Seaton and Zeiders2020), the current study aims to comprehensively examine the concurrent (e.g., same-day) and lagged (e.g., next-day) associations between discrimination and four indicators of diurnal cortisol patterns (i.e., waking cortisol, bedtime cortisol, CAR, and cortisol diurnal slope) among Mexican-origin adolescents.
Linking the discrimination-cortisol response association to adolescents’ mental and sleep health
Theorists posit that development results from a continuous accumulation of many daily experiences, and these daily experiences can accumulate and impact children’s development (Barker, Reference Barker1968; Gallimore et al., Reference Gallimore, Goldenberg and Weisner1993). Research has also found that constant activation of the biological stress response can result in harmful wear and tear on the body and compromise health (McEwen, Reference McEwen2003). Building upon the developmental theories and research, it is possible that altered diurnal cortisol patterns in response to daily discrimination may influence mental (e.g., depressive symptoms, anxiety) and sleep health (e.g., sleep duration, sleep quality) of adolescents.
One approach to examining this postulation is to integrate daily diary and post-diary survey data. Daily diary data encompassing a series of data points of stress and cortisol responses over time can be combined to create an association-between-two-variables indicator that reflects the relation between daily discrimination and cortisol responses in a given individual using multilevel modeling, and this indicator can then be used to predict health (Charles et al., Reference Charles, Piazza, Mogle, Sliwinski and Almeida2013; Cole et al., Reference Cole, Martin, Sterba, Sinclair-McBride, Roeder, Zelkowitz and Bilsky2014; Mandel et al., Reference Mandel, Dunkley and Moroz2015; Wichers et al., Reference Wichers, Geschwind, Jacobs, Kenis, Peeters, Derom, Thiery, Delespaul and van Os2009). The advantage of this method is that it enables the association between discrimination and cortisol responses to vary across individuals, which captures the idiographic and dynamic nature of the physiological responses to stress (Abela & Hankin, Reference Abela and Hankin2009). Using this method, Charles et al. (Reference Charles, Piazza, Mogle, Sliwinski and Almeida2013) found that adults’ daily affective reactivity to stress was associated with increased depressive and anxious symptoms 10 years later. Similarly, Wichers et al. (Reference Wichers, Geschwind, Jacobs, Kenis, Peeters, Derom, Thiery, Delespaul and van Os2009) found that that daily stress reactivity in adult populations prospectively predicted affective symptoms over time. These findings highlight the cumulative effects of daily stress responses on individuals’ overall well-being. Thus, the current study adopts this strategy to explore how the associations between daily discrimination and cortisol response patterns influence overall well-being in Mexican-origin adolescents. The current study focused specifically on depressive symptoms, anxiety, sleep quality, and sleep duration as these developmental outcomes are sensitive to discrimination stress responses (Adam et al., Reference Adam, Quinn, Tavernier, McQuillan, Dahlke and Gilbert2017; Kumari et al., Reference Kumari, Badrick, Ferrie, Perski, Marmot and Chandola2009), and they are also integral domains of adolescent well-being and development (DeHart et al., Reference DeHart, Sroufe and Cooper2004; Short et al., Reference Short, Booth, Omar, Ostlundh and Arora2020).
Current study
To address the limitations of prior research and promote a better understanding of the cumulative effects of discrimination experiences on mental and physical health, the current study integrated both daily diary and post-diary survey data from a community-based sample of 282 Mexican-origin adolescents (see Fig. 1 for the conceptual model). First, the current study examined the day-to-day association between experiences of discrimination and diurnal cortisol levels. Given the short interval since stress onset, we hypothesized that the cortisol response patterns to discrimination would be more in line with the activation hypothesis (Busse et al., Reference Busse, Yim, Campos and Marshburn2017; Miller et al., Reference Miller, Chen and Zhou2007). That is, on days when adolescents reported discrimination experiences, they would have steeper cortisol diurnal slopes and lower bedtime cortisol the same day as well as higher waking cortisol and less pronounced CAR the next day.
Figure 1. The conceptual figure for the proposed model. Post-diary survey was conducted around 1–2 weeks after the completion of the diary survey.
Second, the current study examined whether the daily association between discrimination and diurnal cortisol response was linked with adolescents’ mental health (i.e., depressive symptoms, anxiety) and sleep health (i.e., sleep quality, sleep duration). Based on deleterious effects of discrimination and altered diurnal cortisol response for health (Adam et al., Reference Adam, Quinn, Tavernier, McQuillan, Dahlke and Gilbert2017; Pascoe & Smart Richman, Reference Pascoe and Smart Richman2009), we hypothesized that adolescents who had a stronger association between discrimination and cortisol alteration would report more compromised health in the form of more depressive symptoms and anxiety as well as lower sleep quality and shorter sleep duration.
Method
Participants
Data for the present study were drawn from a three-wave longitudinal study of Mexican immigrant families living in and around a metropolitan area in central Texas (W1: N = 604, W2: N = 483, W3: N = 334). At Wave 1, adolescents were in middle school (sixth through eighth grades), ranging from 11.08 to 15.29 years old (M age = 12.41, SD = .97). Wave 2 data were collected 1 year later (M age = 13.22, SD = .95) and Wave 3 data were collected about 3 years after Wave 2 (M age = 17.10, SD = 1.12). Among the 344 families who consented to participate in the Wave 3 data collection, 290 adolescents participated in the 4-day daily diary study that included a salivary cortisol data collection. Post-diary survey was completed approximately 1–2 weeks after the daily diary data collection. Of the 290 daily diary participants, 8 were eliminated because of their use of steroid-based medication, leaving a final analytic sample size of 282 participants. For the analytic sample, approximately 55% were female (n = 159), and most adolescent participants were U.S.-born (n = 217, 75%). Adolescents came from families in which the median household income fell within the range of $20,001–$30,000 at Wave 1 (between years 2012 and 2015), and the median highest education level of adolescents’ parents was some middle school/junior high school.
Procedures
At Wave 1, families were recruited through public records, school presentations, and community recruitment in and around a metropolitan area in central Texas from 2012 to 2015. To qualify for participation, parents were required to be of Mexican origin with a child in middle school who translated for at least one parent. A family visit was scheduled to obtain parents’ consent and adolescent assent after families were screened as eligible. Bilingual and bicultural interviewers went on family visits, reading questions out loud to participants and entering participants’ responses on a laptop computer. All questionnaires were prepared in both English and Spanish. Questionnaires were first translated to Spanish and then back-translated to English by bilingual and bicultural research assistants. The survey took approximately 2 hours to complete. About 1 year later, families were approached to participate in the Wave 2 study; about three years after Wave 2, families were approached to participate in the daily diary and Wave 3 study.
Adolescents who consented to participate in the daily diary study at Wave 3 were trained to use the Micro Motionlogger Watch (Ambulatory Monitoring, Ardsley, NY) to log their sleep data and the tablet to answer a short nightly survey before bedtime. In addition, adolescents were also trained to collect their own salivary samples immediately after waking, 30 minutes after waking, and at bedtime. Adolescents followed the same protocol for four consecutive school days from Monday to Thursday. On each of the sampling days, phone calls, text messages, or emails were made to remind adolescents to complete the survey on their assigned tablet, provide salivary samples, and wear the Micro Motionlogger Watch on their non-dominant wrist before going to bed. Another follow-up family visit was scheduled after each participant finished the daily diary study to collect all materials (including saliva samples) and equipment as well as family members’ yearly health and demographic information. Most of the participants (77%) completed their follow-up yearly survey within a week after the completion of the diary survey, 9% completed within 2 weeks, 4% completed within 3 weeks, and 2% completed within 4 weeks. Families that participated were compensated $60 at Wave 1 and $90 at Wave 2 and Wave 3. Adolescents who participated in the daily diary component of Wave 3 were compensated up to an additional $40. The compliance rate for daily diary completion was high (M = 3.79, SD = .57), such that 85% of participants completed all four survey days, 11% of participants completed three survey days, 3% of participants competed 2 days and 2% of participants completed 1 day. The compliance rate for the saliva sample was also high (M = 11.67, SD = 1.15) such that 98% of participants provided at least 9 saliva samples and 86% of participants provided all 12 saliva samples.
Measures
Person-level outcome measures were assessed using Wave 3 surveys, and daily measures were assessed using daily diary surveys that occurred approximately 1–2 weeks prior to the Wave 3 survey.
Daily-level measures
Daily Discrimination. Discrimination was assessed daily by four items adapted from the daily discrimination scale (Kessler et al., Reference Kessler, Mickelson and Williams1999). A sample item is, “I was treated with less courtesy than other people today.” Adolescents reported daily on a scale ranging from 1 to 5: “strongly disagree” (1), “disagree” (2), “neutral” (3), “agree” (4), and “strongly agree” (5). Each item was recoded as 0 if adolescents answered 1–3, indicating no such experience or 1 if adolescents answered 4–5, indicating they had such experiences. Scores were summed each day (range: 0–4). Consistent with other research employing daily diary methods, daily discrimination was a relatively low-frequency event (M = .16 SD = .56; Goosby et al., Reference Goosby, Cheadle and Mitchell2018; Torres & Ong, Reference Torres and Ong2010; Yip et al., Reference Yip, Cheon, Wang, Cham, Tryon and El-Sheikh2020). Therefore, discrimination was dichotomized to indicate whether adolescents experienced any of the four discriminatory experiences that day (0 = no discrimination, 1 = discrimination).
Salivary Cortisol. Participants were instructed to provide saliva samples via the passive drool method by spitting into a Sali Cap vial (IBL International GMBH, Hamburg, Germany). Participants provided saliva at three time points each day, resulting in a total of 12 samples: immediately upon waking up (sample 1), 30 minutes after waking up (sample 2), and at bedtime (sample 3). Participants were advised to not eat, drink, or brush their teeth 30 minutes before sampling. The 12 vials were stored in Medication Event Monitoring System (MEMS) caps, which contain an embedded computer chip that digitally records when the bottles are opened to dispense the saliva collection vial. In addition to using the MEMS caps, participants self-reported the timing of each sample using a tablet questionnaire. We prioritized using the sample timing recorded by the MEMS caps, but in cases where the timing was missing, we relied on self-reported sample timing. Participants’ self-reported awakening times were validated by the Micro Motionlogger data. Participants were instructed to store their samples in the freezer section of their household refrigerators. At the conclusion of the saliva collection period, research assistants transported the samples to the research laboratory, where they were stored in a −32°C freezer before being sent to Biochemisches Labor in Trier, Germany, for cortisol assay. In total, 3,480 (4 days × 3 samples × 290 participants) saliva samples were collected during the 4-day diary period of 290 participants. However, eight participants reported taking steroid-based medication during the study period, and they were excluded, leaving the final analytic saliva samples to be 3,384. The compliance with the time interval for taking Sample 1 and 2 was high, as 94% of the second samples were taken within 45 minutes of Sample 1 (DeSantis et al., Reference DeSantis, Adam, Mendelsohn and Doane2010). Cortisol concentration was determined using a time-resolved fluorescence immunoassay (see Dressendörfer et al., Reference Dressendörfer, Kirschbaum, Rohde, Stahl and Strasburger1992, for a detailed description). Intra-assay coefficients of variation (CVs) were between 4.0% and 6.7%, and inter-assay CVs ranged from 7.1% to 9.0%. Consistent with prior cortisol studies (Kertes & van Dulmen, Reference Kertes and van Dulmen2012), cortisol values were winsorized at 50 nmol/L to reduce the effect of outliers on the analyses. In total, there were 5 data points that were winsorized. We examined four different cortisol indicators: (1) waking cortisol, (2) bedtime cortisol, (3) CAR, and (4) diurnal slope. The CAR was calculated by subtracting the cortisol level of each day’s Sample 2 (30 minutes post waking) from Sample 1 (waking). There were 62 cases in which sample 2 fell outside the 45-minute window after sample 1, and their CARs were coded as missing following the consensus guidelines (Stalder et al., Reference Stalder, Lupien, Kudielka, Adam, Pruessner, Wüst, Dockray, Smyth, Evans, Kirschbaum, Miller, Wetherell, Finke, Klucken and Clow2022). The daily diurnal slope was computed by taking the difference between waking cortisol level and bedtime cortisol level and dividing by the time between samples. Among the four cortisol indicators, only bedtime cortisol was positively skewed (7.50), and thus it was transformed using the natural log transformation.
Person-level measures
Depressive Symptoms. Depressive symptoms were measured at Waves 2 and 3. It was assessed by the widely used 20-item Center for Epidemiologic Studies of Depression Scale (Chen et al., Reference Chen, Jelsma, Hou, Benner and Kim2021; Radloff, Reference Radloff1977). Adolescents self-reported how often during the past week they had experienced depressive symptoms (a sample item: “Bothered by things usually not bothered by”) on a scale of 1 (rarely or none of the time) to 4 (most or all of the time). Higher mean scores indicate more depressive symptoms (αwave.2 = .87; αwave.3 = .88).
Anxiety. Anxiety was measured at Waves 2 and 3 by four items adapted from prior studies (Chen et al., Reference Chen, Jelsma, Hou, Benner and Kim2021; Reynolds & Richmond, Reference Reynolds and Richmond1978; Spitzer et al., Reference Spitzer, Kroenke, Williams and Löwe2006). On a scale of 1 (not at all) to 5 (nearly every day), adolescents self-reported how often they were bothered by the following problems over the last 2 weeks: (1) feeling nervous, (2) worrying about what is going to happen, (3) trouble relaxing, and (4) becoming easily annoyed or irritable. Higher mean scores suggest higher levels of anxiety (αwave.2 = .83; αwave.3 = .82).
Self-reported Sleep Quality. On a scale of 1 (poor) to 5 (excellent), adolescents self-reported their sleep quality with one item at Wave 2 and 3: “During the past month, how would you rate your sleep quality overall?” This was drawn from the Pittsburgh Sleep Quality Index (Buysse et al., Reference Buysse, Reynolds, Monk, Berman and Kupfer1989; Hou et al., Reference Hou, Kim and Benner2018).
Sleep Duration. Sleep duration was recorded by the Micro Motionlogger Watch during the 4-day diary. It was averaged across the 4 days of the daily diary to represent participants’ average sleep duration during the diary period.
Covariates
At the daily level, covariates included day of the study (i.e., 1–4) as well as a list of daily behaviors known to be associated with diurnal cortisol (Adam & Kumari, Reference Adam and Kumari2009): daily consumption of alcohol/nicotine/caffeine, daily exercise (1 = yes, 0 = no), waking time, oral disease or injury (1 = yes, 0 = no), and whether taken nonsteroid-based medication in the past 12 hours (1 = yes, 0 = no). For CAR, an additional daily-level covariate, time between cortisol sample 1 and 2, was added. At the person level, covariates included adolescent age, gender (0 = male, 1 = female), nativity (0 = Mexico-born, 1 = US-born), and family income, which were measured at Wave 1, to account for individual-level factors that have been found to relate to cortisol levels and discrimination experiences (Adam & Kumari, Reference Adam and Kumari2009; Chavous et al., Reference Chavous, Rivas-Drake, Smalls, Griffin and Cogburn2008). Parents reported their highest education level on a scale of 1 (no formal schooling) to 11 (finished graduate degree).
Analysis plan
First, to examine the daily association between discrimination and cortisol activities, we conducted multilevel analyses in R using the lme4 package (Bates et al., Reference Bates, Machler, Bolker and Walker2015). Daily cortisol parameters (i.e., waking cortisol, bedtime cortisol, CAR, and diurnal slope) were treated as dependent variables. At the WP level, we included daily discrimination as a predictor, and covariates were integrated at the between and within levels as appropriate. Daily discrimination was centered on each person’s mean to represent the WP effect. Daily-level covariates were centered by each participant’s personal mean, and person-level covariates were centered by the grand mean. Following the recommended statistical procedures for examining WP effects (Bolger & Laurenceau, Reference Bolger and Laurenceau2013), we also included the BP centered discrimination variable to adjust for possible differences between participants. The equations were as follows:
Level 1: Cortisolij = β 0i + β 1i (WP Daily Discrimination) + (β 2i,3i, 4i … Daily Covariates) + e ij
Level 2: β 0i = γ 00 + γ 01 (BP Discrimination) +(γ 02,3,4 … Person-Level Covariates) + u 0i,
β 1i = γ 10 + u 1i
At Level 1, the value β 0i was a regression intercept, reflecting the mean level of daily cortisol on days in which the predictor (i.e., WP daily discrimination) was at its mean and daily covariates were zero; β 1i was a regression slope (cortisol response to discrimination coefficient), representing the main effect of WP daily discrimination on diurnal cortisol patterns, holding all other variables constant; and the residual parameter (e ij ) indexed the day-to-day variability in cortisol values for each individual. At level 2, γ 00 represented the sample average level of cortisol, holding the person-level covariates constant; γ 10 represented the sample average effect of cortisol response to discrimination, holding the person-level covariates constant. In addition, u 0i and u 1i were variances reflecting individual differences from the sample average level of cortisol and cortisol response to discrimination estimates, respectively. Empirical Bayes (model-based) slopes were calculated for each individual by adding the sample average effect of daily discrimination on cortisol (γ 10 ) with individualized random slopes (u 1i , Singer & Willett, Reference Singer and Willett2003). Therefore, each person had unique regression parameters, representing their own relation between discrimination and cortisol. The estimated Empirical Bayes slopes were outputted and used for subsequent analyses. Models were estimated by means of restricted maximum likelihood. The estimates of the association between discrimination and each cortisol indicator represented a slope for each participant and were saved and used for subsequent analyses. For the same-day effect, only bedtime levels and diurnal slope were included as dependent variables given that the waking cortisol and the CAR would have occurred before the reports of discrimination. For the next-day effect, the lagged analyses were examined where daily discrimination predicted next-day cortisol (i.e., waking, bedtime, CAR, diurnal slope).
Second, to examine whether daily cortisol response to discrimination would influence mental and sleep health, we included the estimated slopes of discrimination and each cortisol indicator as predictors of Wave 3 depressive symptoms, anxiety, sleep quality, and sleep duration in a structural equation model using Mplus 8.3 (Muthén & Muthén, Reference Muthén and Muthén1998-2018). This model also integrated the person-level covariates. In order to avoid multicollinearity, each cortisol response to discrimination indicator was tested in a separate model due to high correlations among the estimated slopes. All health indicators (i.e., depressive symptoms, anxiety, and sleep quality) were included in the same model, thus adjusting for any covariance between the outcomes. Missing data were handled using full information maximum likelihood.
Alternative model
To provide more confidence in the proposed model for the present study, an alternative model was tested. As the biological reactivity to discriminatory treatment may be influenced by early experiences (Boyce & Ellis, Reference Boyce and Ellis2005), it is possible that individuals who had poor sleep quality might be more reactive biologically when facing discrimination. Indeed, research has shown that individuals with depressive symptoms are at greater risk for being the target of discrimination (Hou et al., Reference Hou, Kim, Wang, Shen and Orozco-Lapray2015) and that they tend to react more strongly to the negative treatment (Gotlib & Joormann, Reference Gotlib and Joormann2010). Therefore, the alternative model for research question 2 examined whether socioemotional and sleep health at an earlier wave (Wave 2) would influence adolescents’ diurnal cortisol response (i.e., waking cortisol, bedtime cortisol, CAR, diurnal cortisol slope) to discrimination. To avoid multicollinearity, each health indicator was tested in a separate model due to high correlations among them. All cortisol response to discrimination indicators were included in the same model.
Results
Descriptive statistics and correlations of the primary study variables averaged within each person are reported in Table 1. In total, 21.7% of participants reported experiencing discrimination on at least one of the 4 days. Participants who experienced more frequent discrimination tended to have steeper diurnal cortisol slopes (r = −.18, p < .01), worse sleep quality (r = −.13, p < .05), more anxiety (r = .35, p < .01), and more depressive symptoms (r = .36, p < .01).
Table 1. Bivariate correlations and descriptive statistics of study variables averaged within each person
All cortisol levels represent nmol/L. Total possible N = 282.
*p < .05.
**p < .01.
***p < .001.
Association between daily discrimination and diurnal cortisol patterns
Unconditional means models were first conducted for each of the diurnal cortisol parameters to calculate the intraclass correlations. The results suggested that 25%–44% of the variance in cortisol parameters were attributable to BP variation, whereas 56%–75% was attributable to WP variation (Table 1), which supported the use of two-level models. In addition, inclusion of the random slope for daily discrimination resulted in lower AIC and BIC values and higher R-square value compared to the models without the random slope term, which suggests the importance of including random slope in modeling.
At the WP level, daily experiences of discrimination were not significantly associated with any of the diurnal cortisol indicators (Table 2). However, at the BP level, average discrimination across the study was significantly associated with same-day diurnal slope (b = −.18, p = .02) and next-day evening cortisol (b = −.63, p < .001) after accounting for all time invariant and time-varying covariates. That is, individuals who reported higher levels of discrimination across the study period had steeper diurnal slopes and lower evening cortisol.
Table 2. Multi-level models for associations between daily discrimination and adolescents’ diurnal cortisol patterns
AUC = area under the curve; BP = between-person; CAR = cortisol awakening response; WP = within-person.
Significant results were in bold.
All cortisol levels represent nm/h.
Linking the discrimination-cortisol response association to adolescents’ mental and sleep health
As shown in Table 3, the daily association between discrimination and the same-day diurnal slope was observed to be negatively related to sleep duration (b = −1.70, p = .02), indicating that stronger negative associations between discrimination and diurnal cortisol slope (i.e., steeper diurnal cortisol slope) were related to longer sleep duration. This effect was also significant for the relation between next-day diurnal slope response to daily discrimination and sleep duration (b = −1.19, p = .01). In addition, the daily association between discrimination and same-day bedtime cortisol was positively associated with sleep quality (b = 1.44, p = .02). That is, stronger negative associations between discrimination and bedtime cortisol were, in turn, related to lower levels of sleep quality. This effect was also significant for next-day bedtime cortisol response to daily discrimination (b = .33, p = .01). Finally, the daily association between discrimination and next-day diurnal slope was observed to be negatively related to both depressive symptoms (b = −.39, p = .02) and anxiety (b = −.46, p = .02). That is, stronger negative associations between discrimination and diurnal cortisol slope (i.e., steeper diurnal cortisol slope) were associated with higher levels of depressive symptoms and anxiety.
Table 3. Regression coefficients between cortisol response to discrimination and mental and sleep health outcomes
Discrim = discrimination.
Significant findings are in bold.
a 0 = Mexico−born, 1 = US − born.
Alternative model
The alternative model for research question 2 examined whether socioemotional and sleep health at an earlier wave (Wave 2) influenced adolescents’ diurnal cortisol response (i.e., waking cortisol, bedtime cortisol, CAR, diurnal cortisol slope) to discrimination. As shown in Table 4, depressive symptoms, anxiety, and sleep quality at Wave 2 were not significantly associated with any of the cortisol response to daily discrimination indicators, and, as such, the data supported the conclusion that diurnal cortisol response to discrimination precedes subsequent health-related problems.
Table 4. Regression coefficients predicting cortisol response to discrimination from w2 well−being
CAR = cortisol awakening response; Discrim = discrimination.
Discussion
Discrimination is prevalent in the lives of Mexican-origin adolescents, one of the largest and fastest-growing ethnic minority groups in the U.S. (Noe-Bustamante et al., Reference Noe-Bustamante, Flores and Shah2019). Discrimination also takes a toll on individuals’ biological stress responses, including the HPA axis functioning (Doane & Zeiders, Reference Doane and Zeiders2014; Huynh et al., Reference Huynh, Guan, Almeida, McCreath and Fuligni2016; Zeiders et al., Reference Zeiders, Doane and Roosa2012). Adolescents are particularly vulnerable to such harmful effects given that their HPA axis is reactive to stressors and that they are sensitive to social-evaluative stressors (Dickerson & Kemeny, Reference Dickerson and Kemeny2004; Romeo, Reference Romeo2013). The current study integrated both daily diary and post-diary survey data to examine whether daily diurnal cortisol responses to discrimination influenced adolescents’ mental and sleep health. The results indicated that daily experiences of discrimination were associated with steeper diurnal cortisol slope and lower evening cortisol; however, such physiological responses tended to come with a toll on mental and sleep health. Given the recent political attention on immigration in the U.S. and the associated backlash against immigrants, the findings of the current study appear to be especially pertinent at a time when Mexican Americans are facing increasing discriminatory and unfair treatment (Lopez et al., Reference Lopez, Gonzalez-Barrera and Krogstad2018).
The first goal of the study was to examine the association between the daily experience of discrimination and diurnal cortisol responses among Mexican-origin adolescents. Surprisingly, at the WP level, daily experience of discrimination did not predict any of the diurnal cortisol activity indicators after taking into account the control variables. That is, on days when adolescents reported greater discrimination compared to their average levels, their same-day and next-day diurnal cortisol patterns did not change significantly. It is possible that there may not be enough variability in the reported discrimination experiences to capture the significance of the WP effect, as only 21% of the study sample reported at least one incidence of discrimination across the 4-day study period. Future studies with longer study periods (e.g., 14-day diary) or weekly reports of discrimination could better capture the variation in discrimination experiences and provide further insights into the with-person relations. It is also possible that there may be potential moderators that are contributing to these null findings. For example, Seaton and Zeiders’s recent work (2020) found that, for individuals with high levels of racial centrality, increased daily racial discrimination was associated with higher waking cortisol; however, this association was not significant for those with low racial centrality. Future studies should continue exploring culturally salient moderators such as acculturation (White et al., Reference White, Zeiders, Knight, Roosa and Tein2014), ethnic identity (Umaña-Taylor et al., Reference Umaña-Taylor, Gonzales-Backen and Guimond2009), and family racial socialization (Umaña-Taylor & Hill, Reference Umaña-Taylor and Hill2020) in order to gain a more comprehensive understanding of the WP associations between daily discrimination and diurnal cortisol patterns among Mexican-origin adolescents.
In contrast, at the BP level, discrimination was associated with steeper diurnal slopes and lower evening cortisol. This suggests that individuals who reported experiencing more discrimination tended to have steeper diurnal slopes, which is possibly driven by lower evening cortisol levels. Together, these findings lend support to the activation hypothesis (Dickerson & Kemeny, Reference Dickerson and Kemeny2004; Selye, 1956), suggesting that when Mexican-origin adolescents encounter the stressful experience of discrimination in the short term, they tend to activate their cortisol response to prepare to mobilize physiological resources and adjust to the threatening situation by demonstrating activated physiological functioning. Our results shed light on the mixed findings of the effects of discrimination on diurnal cortisol responses by highlighting the importance of taking into account the timing between discrimination and diurnal cortisol responses (Busse et al., Reference Busse, Yim, Campos and Marshburn2017). Specifically, unlike most of the prior studies that measured discrimination over the past 12 months or lifetime, our study asked participants to report discrimination experiences on a daily basis, and thus our findings are more in line with the cortisol release patterns that are due to acute stress (Miller et al., Reference Miller, Chen and Zhou2007). Indeed, a recent study that examined daily experiences of racial discrimination also found that increased racial discrimination experiences on a day-to-day basis were associated with steeper diurnal cortisol slopes and less pronounced CAR for Black adults who endorsed high levels of racial centrality (Seaton & Zeiders, Reference Seaton and Zeiders2020). It is possible that when Mexican-origin adolescents experience discrimination as an acute event, they might be well-equipped to deal with the discriminatory experience by relying on positive aspects of their racial/ethnic identity or by mobilizing their cognitive resources and attributing the cause of negative treatment to external factors rather than to themselves (Graham & Chen, Reference Graham and Chen2020). Thus, they respond to discrimination incidents with activated physiological responses. The significant finding of discrimination and activated physiological responses profiles at the BP level also implies that when discrimination occurs in a broader environment that permits such mistreatment, it might be linked with short-term activated physiological reactions, which is consistent with theoretical perspectives suggesting that daily awareness of racism can act as a protective mechanism for negatively stigmatized minority groups (Fuller-Rowell et al., Reference Fuller-Rowell, Doan and Eccles2012; Helms, Reference Helms1995; Sellers et al., Reference Sellers, Smith, Shelton, Rowley and Chavous1998). Future studies should further consider the timing between discrimination exposure and cortisol response to examine the potential cognitive and psychological mechanisms through which daily experiences of discrimination influence diurnal cortisol responses.
When linking the discrimination-cortisol association to adolescent mental and sleep health, the results showed that next-day steeper diurnal slopes following discrimination were associated with higher levels of both depressive symptoms and anxiety at Wave 3. In terms of sleep health, our results showed that the same/next-day lower bedtime cortisol following discrimination was significantly associated with lower levels of sleep quality and the same/next-day steeper diurnal slopes following discrimination were associated with longer sleep duration. Together, these findings suggest that the short-term activation of cortisol responses in the face of discrimination may take a toll on Mexican-origin adolescents’ mental and sleep health. It is possible that, even though the marginalized groups are well-equipped to deal with discrimination by showing resilience in adapting their physiological responses, their cognitive resources might be depleted, and the cognitive strain might pile up to undermine their mental health and sleep quality. Our findings echoed the literature on the under-the-skin wear and tear of stressors (McEwen, Reference McEwen2003), which argues that there is an adaptation cost associated with the activation of physiological systems. This can lead to wear and tear in the body, contributing to allostatic load. These findings further added to the Latina/x/o health disparities literature by demonstrating that the short-term altered physiological response might be one potential pathway through which discrimination undermines mental and sleep health among Mexican-origin youth. Future studies examining potential moderators in these processes are needed to identify potential avenues for interventions. For example, as family connectedness is highly valued in Mexican culture (Stein et al., Reference Stein, Cupito, Mendez, Prandoni, Huq and Westerberg2014), future studies could explore whether Mexican-origin adolescents with supportive parents could be shielded from the adaption cost in the face of discrimination.
As a final note, the findings of the short-term activation of cortisol responses following discrimination linked to longer sleep duration were somewhat surprising. This finding suggests that sleep might serve as a restorative process, and engaging in longer sleep could potentially assist individuals in recovering from the physiological adaptations in response to stress (e.g., discrimination). This is consistent with research on the escape-to-sleep coping strategy, which has found that individuals tend to use sleep as a coping mechanism to counter adverse social conditions (Goosby et al., Reference Goosby, Cheadle, Strong-Bak, Roth and Nelson2018; Yip, Reference Yip2015). Given that sleep and cortisol often work together in response to stress (e.g., discrimination; Goosby et al., Reference Goosby, Straley and Cheadle2017) and the fact that both sleep duration and cortisol responses were measured daily in the current study, it is also possible that longer sleep duration following a daily discrimination incident could potentially result in improved memory encoding, and consequently leading to greater cortisol responses. Future studies with longer intervals between cortisol responses to stress and sleep measurements are encouraged to delve deeper into the stress response pathways following discrimination experiences.
Strengths, limitations, and future directions
The current study contributed to the extant discrimination literature in several key ways. First, using daily diary data, the current study was able to capture adolescents’ experiences of discrimination closer in time to the event occurring and identified BP effects of daily discrimination on diurnal cortisol. The results suggested a different short-term diurnal cortisol pattern in response to discrimination (i.e., steeper diurnal slope, less pronounced CAR, and lower bedtime cortisol) as compared to prior studies that primarily measure chronic discrimination experienced over longer time spans documenting flatter slopes, higher bedtime cortisol, and more pronounced CAR (Doane & Zeiders, Reference Doane and Zeiders2014; Huynh et al., Reference Huynh, Guan, Almeida, McCreath and Fuligni2016; Zeiders et al., Reference Zeiders, Doane and Roosa2012). Our findings highlight the benefits of investigating day-to-day experiences of discrimination using the daily diary approach. Second, by integrating daily diary and Wave 3 post-diary survey data, the current study linked the day-to-day cortisol responses to discrimination to overall mental and sleep health. Findings of the current study highlight that daily experiences in adolescents’ lives can accumulate to influence their developmental outcomes. Recognizing the bidirectional associations between stress responses and well-being (Boyce & Ellis, Reference Boyce and Ellis2005), we also tested alternative models using health indicators at Wave 2 to predict diurnal cortisol responses to discrimination. The alternative models were not supported, providing additional evidence that stress response to discrimination is more likely to precede health-related problems.
Although the contributions of this study are clear, limitations must be acknowledged. First, the daily diary component of the study was restricted to four consecutive school days. Although daily dairy designs provide a snapshot of individuals’ day-to-day lives, the four-day design may not be sufficient for capturing incidents of discrimination given how infrequent reports of everyday discrimination are (Goosby et al., Reference Goosby, Cheadle and Mitchell2018; Huynh & Fuligni, Reference Huynh and Fuligni2010). Future studies with longer diary periods (e.g., 14-day or 21-day diary) or weekly reports of discrimination over a month are strongly encouraged to examine whether the study findings could be replicated over longer time periods. In a similar vein, the four-day design also may not be sufficient to capture the complexities of the physiological processes underlying adolescents’ daily experiences given that diurnal cortisol activity is not highly stable across days (Ross et al., Reference Ross, Murphy, Adam, Chen and Miller2014). Future studies with a greater number of days and across a greater span of developmental periods are needed to gain a more comprehensive understanding of the physiological underpinnings of discriminatory experiences. Second, although the current study integrated the daily diary data with post-diary survey data, the design is a short-term longitudinal design with the mental and sleep health indicators collected shortly after the daily surveys were concluded. Future studies with longer-term longitudinal designs are needed to better capture the accumulative effects of daily experiences on long-term health (Barker, Reference Barker1968; Gallimore et al., Reference Gallimore, Goldenberg and Weisner1993). Thirdly, the generalizability of the findings is limited given that the sample of the current study was collected in the southern U.S. in areas with a large Mexican American population. Research has shown that ethnic enclaves tend to protect ethnic minorities from discrimination (White et al., Reference White, Zeiders, Knight, Roosa and Tein2014). Therefore, future studies with samples from settings that vary in diversity and same-ethnic representation are needed to replicate the findings of the current study. Lastly, current CAR guidelines recommend accessing CAR using at least three samples in the morning (Stalder et al., Reference Stalder, Lupien, Kudielka, Adam, Pruessner, Wüst, Dockray, Smyth, Evans, Kirschbaum, Miller, Wetherell, Finke, Klucken and Clow2022). Here, only two saliva samples were collected to calculate CAR. Therefore, future studies should replicate the analyses in the current study using cortisol data adhere to these new cortisol data collection guidelines (Stalder et al., Reference Stalder, Lupien, Kudielka, Adam, Pruessner, Wüst, Dockray, Smyth, Evans, Kirschbaum, Miller, Wetherell, Finke, Klucken and Clow2022).
Conclusions
The current study took a novel approach in integrating daily diary and post-diary survey data to delve into the physiological mechanisms underlying perceptions of discrimination and larger health disparities. The current study provided important insights into adolescents’ daily experiences and physiological responses to discrimination as well as the implications of these associations for adolescents’ mental and sleep health. Our findings suggest that there might be adaptation costs associated with the short-term cortisol activation in the face of discrimination. The current study makes an important contribution to understanding the biological process underlying Mexican-origin adolescents’ experiences of discrimination and points to the importance of considering discrimination experiences for Mexican-origin youth’s stress biology in unpacking health disparities among this population.
Funding statement
Support for this research was provided through awards to Su Yeong Kim from (1) National Science Foundation, Division of Behavioral and Cognitive Sciences, 1651128 and 0956123, (2) National Institute on Minority Health and Health Disparities 1R21MD012706-01A1 and 3R21MD-012706-02S1, (3) Eunice Kennedy Shriver National Institute of Child Health and Human Development 5R03HD060045-02, (4) Russell Sage Foundation, 2699, (5) Spencer Foundation, 10023427, (6) Hogg Foundation for Mental Health JRG-102, (7) Office of the Vice President for Research and Creative Grant and Special Research Grant from the University of Texas at Austin, (8) College of Natural Sciences Catalyst Grant from the University of Texas at Austin, and (9) Eunice Kennedy Shriver National Institute of Child Health and Human Development 2P2CHD042849-19 grant awarded to the Population Research Center at The University of Texas at Austin. These funding sources had no role other than financial support.
Competing interests
None.
