Hostname: page-component-848d4c4894-75dct Total loading time: 0 Render date: 2024-06-01T23:49:52.232Z Has data issue: false hasContentIssue false

Stress system concordance as a predictor of longitudinal patterns of resilience in adolescence

Published online by Cambridge University Press:  12 July 2023

Andrea Wiglesworth*
Affiliation:
Psychology, University of Minnesota Twin Cities, Minneapolis, MN, USA
Jessica Butts
Affiliation:
Public Health, Division of Biostatistics, University of Minnesota Twin Cities, Minneapolis, MN, USA
Katherine A. Carosella
Affiliation:
Psychology, University of Minnesota Twin Cities, Minneapolis, MN, USA
Salahudeen Mirza
Affiliation:
Psychology, University of Minnesota Twin Cities, Minneapolis, MN, USA
Victoria Papke
Affiliation:
Psychology, University of Minnesota Twin Cities, Minneapolis, MN, USA
Jason José Bendezú
Affiliation:
Psychology, The Pennsylvania State University, University Park, PA, USA
Bonnie Klimes-Dougan
Affiliation:
Psychology, University of Minnesota Twin Cities, Minneapolis, MN, USA
Kathryn R. Cullen
Affiliation:
Psychiatry and Behavioral Sciences, University of Minnesota Twin Cities, Minneapolis, MN, USA
*
Corresponding author: A. Wiglesworth; Email: wigle017@umn.edu
Rights & Permissions [Opens in a new window]

Abstract

Resilience promotes positive adaptation to challenges and may facilitate recovery for adolescents experiencing psychopathology. This work examined concordance across the experience, expression, and physiological response to stress as a protective factor that may predict longitudinal patterns of psychopathology and well-being that mark resilience. Adolescents aged 14–17 at recruitment (oversampled for histories of non-suicidal self-injury; NSSI) were part of a three-wave (T1, T2, T3) longitudinal study. Multi-trajectory modeling produced four distinct profiles of stress experience, expression, and physiology at T1 (High-High-High, Low-Low-Low, High-Low-Moderate, and High-High-Low, respectively). Linear mixed-effect regressions modeled whether the profiles predicted depressive symptoms, suicide ideation, NSSI engagement, positive affect, satisfaction with life, and self-worth over time. Broadly, concordant stress response profiles (Low-Low-Low, High-High-High) were associated with resilient-like patterns of psychopathology and well-being over time. Adolescents with a concordant High-High-High stress response profile showed a trend of greater reduction in depressive symptoms (B = 0.71, p = 0.052), as well as increased global self-worth (B = –0.88, p = 0.055), from T2 to T3 compared to the discordant High-High-Low profile. Concordance across multi-level stress responses may be protective and promote future resilience, whereas blunted physiological responses in the presence of high perceived and expressed stress may indicate poorer outcomes over time.

Type
Special Issue Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2023. Published by Cambridge University Press

Introduction

Resilience is a dynamic process that promotes positive adaptation to threatening circumstances (Cicchetti, Reference Cicchetti2013; Curtis & Cicchetti, Reference Curtis and Cicchetti2003). Resilience mechanisms contribute to the variability in outcomes after exposure to a stressor and are implicated in protection against the development of psychopathology in particular (Cicchetti, Reference Cicchetti2010). While studies on resilience have largely focused on recovery from a traumatic event, and experiences of maltreatment in particular (Masten et al., Reference Masten, Best and Garmezy1990), we propose that resilience can also be considered as a process that promotes recovery from psychopathology to restore a thriving state. Adolescent-onset psychopathology is associated with antecedent adversities, is a developmental phenomenon that requires adaptation, and is prospectively linked to poorer mental health, educational, occupational, and social outcomes that persist into early adulthood (Davies et al., Reference Davies, Sullivan and Zammit2018). However, resilience processes may promote adaptive functioning and eventual recovery among a subset of individuals affected by psychopathology in adolescence. To summarize, though definitions of resilience typically call for identification of a discrete stressor, we argue that the multifactorial and cumulative stresses associated with adolescent-onset psychopathology can be situated in the resilience framework as “major assaults on the developmental process” (Luthar et al., Reference Luthar, Cicchetti and Becker2000). Therefore, highlighting protective mechanisms which promote healthy adaptation for psychopathology-affected adolescents is critical to ameliorating downstream sequelae and functional impairment.

Depressive symptoms and self-injurious thoughts and behaviors commonly emerge in adolescence and produce substantial distress for youth. Major depressive disorder has a past-year prevalence of 16% in adolescence (Daly, Reference Daly2022). Self-injurious thoughts and behaviors, while co-occurring with depression, are distinct transdiagnostic constructs and include suicidal thoughts and behaviors (e.g., suicide ideation and suicide attempt) as well as non-suicidal self-injury (NSSI). Suicide ideation and attempt have a lifetime prevalence of 12 and 4% respectively in adolescence (Nock et al., Reference Nock, Green, Hwang, McLaughlin, Sampson, Zaslavsky and Kessler2013). NSSI, which is the engagement in self-injury without intent to die, has an adolescent lifetime prevalence of 17% (Swannell, Martin, Page, Hasking, & St John, Reference Swannell, Martin, Page, Hasking and St John2014). NSSI is often conceptualized as a coping tool, as it typically occurs in the context of severe negative emotions, such as those associated with depressive disorders and suicide ideation, and serves to manage stressors and remove negative affect in the present (Başgöze, Wiglesworth, Carosella, Klimes-Dougan, & Cullen, Reference Başgöze, Mirza, Silamongkol, Hill, Falke, Thai, Westlund Schreiner, Parenteau, Roediger, Hendrickson, Mueller, Fiecas, Klimes-Dougan and Cullen2021; Bentley et al., Reference Bentley, Nock and Barlow2014; Lloyd-Richardson et al., Reference Lloyd-Richardson, Perrine, Dierker and Kelley2007). However, NSSI can be life-threatening and is associated with future suicide attempts (Castellví et al., Reference Castellví, Lucas-Romero, Miranda-Mendizábal, Parés-Badell, Almenara, Alonso, Blasco, Cebrià, Gabilondo, Gili, Lagares, Piqueras, Roca, Rodríguez-Marín, Rodríguez-Jimenez, Soto-Sanz and Alonso2017; Nock et al., Reference Nock, Borges, Bromet, Alonso, Angermeyer, Beautrais, Bruffaerts, Chiu, de Girolamo, Gluzman, de Graaf, Gureje, Haro, Huang, Karam, Kessler, Lepine, Levinson, Medina-Mora, Ono and Williams2008; Ribeiro et al., Reference Ribeiro, Franklin, Fox, Bentley, Kleiman, Chang and Nock2016). Thus, adolescents who engage in NSSI represent a group of young people who are experiencing significant distress and are experimenting with different ways to navigate these difficult experiences.

Despite the stressors associated with adolescent psychopathology, some youth appear to recover from these symptoms across the course of adolescence. Extant longitudinal research suggests that, while some adolescents with major depressive disorder continue experiencing recurrent depressive episodes and suicidal thoughts and behaviors well into adulthood, a substantial proportion recover (Curry et al., Reference Curry, Silva, Rohde, Ginsburg, Kratochvil, Simons, Kirchner, May, Kennard, Mayes, Feeny, Albano, Lavanier, Reinecke, Jacobs, Becker-Weidman, Weller, Emslie, Walkup, Kastelic and March2011; Johnson et al., Reference Johnson, Dupuis, Piche, Clayborne and Colman2018; Weissman et al., Reference Weissman, Wolk, Goldstein, Moreau, Adams, Greenwald, Klier, Ryan, Dahl and Wickramaratne1999). Though the literature is more limited on NSSI recovery, many adolescents who engage in NSSI appear to desist in late adolescence or early adulthood (Plener et al., Reference Plener, Schumacher, Munz and Groschwitz2015; Rajhvajn Bulat et al., Reference Rajhvajn Bulat, Sušac and Ajduković2023). These patterns of recovery implicate the presence of resilience processes and protective mechanisms, which, when understood, could be leveraged to restore adaptive functioning in youth suffering from psychopathology.

In addition to reductions in symptoms of psychopathology, evidence of increased well-being over time, especially in the wake of stressors and difficulties, is another important marker of resilience. Importantly, well-being and psychopathology do not represent separate ends of a single dimension; rather, they represent distinct– albeit often inversely related– domains that individuals can experience together in varying combinations (de Vries et al., Reference de Vries, Baselmans, Luykx, de Zeeuw, Minică, de Geus, Vinkers and Bartels2021). Well-being has been conceptualized as the combination of feeling good and functioning effectively, with the ability to manage negative emotions to a certain extent (Ruggeri et al., Reference Ruggeri, Garcia-Garzon, Maguire, Matz and Huppert2020). Thus, it is unsurprising that well-being and resilience are strongly linked (de Vries et al., Reference de Vries, Baselmans, Luykx, de Zeeuw, Minică, de Geus, Vinkers and Bartels2021; Klainin-Yobas et al., Reference Klainin-Yobas, Vongsirimas, Ramirez, Sarmiento and Fernandez2021; Li & Hasson, Reference Li and Hasson2020). Life satisfaction, self-worth and positive affect are critical domains of well-being that are linked with resilience (Cohn et al., Reference Cohn, Fredrickson, Brown, Mikels and Conway2009; Liu et al., Reference Liu, Wang and Li2012). However, well-being is not typically a focus of research on depression and self-injury in adolescence, which tends to narrowly measure “deficits” in functioning. In this regard, understanding the resilience processes underlying the positive modulation of these aspects of well-being, even when baseline circumstances may be distressing, could provide greater insight to adaptation in the context of stress.

Conceptual models hold that resilience processes involve transactions across systems and levels of analysis which move individuals towards healthy adaptation (Masten et al., Reference Masten, Lucke, Nelson and Stallworthy2021). In the current study, we focus on a multi-level illustration of the adolescent stress response as a potential resilience mechanism involved in developmental patterns towards adaptation and away from maladaptation. One of these levels-of-analysis involves the hypothalamic-pituitary adrenal (HPA) axis, which orchestrates an array of energy-mobilizing processes in response to stress (Herman et al., Reference Herman, Ostrander, Mueller and Figueiredo2005). One of the major end products of the HPA axis pathways is cortisol, an adrenocorticoid steroid hormone which modulates gene expression across tissues including the brain (Godoy et al., Reference Godoy, Rossignoli, Delfino-Pereira, Garcia-Cairasco and de Lima Umeoka2018). Heightened patterns of HPA functioning (both basal functioning and response to stressors as indexed by the steroid hormone cortisol) have been repeatedly demonstrated in those with depression (Iob et al., Reference Iob, Kirschbaum and Steptoe2020; Juruena et al., Reference Juruena, Bocharova, Agustini and Young2018; Klimes-Dougan et al., Reference Klimes-Dougan, Begnel, Almy, Thai, Schreiner and Cullen2019; Lopez-Duran et al., Reference Lopez-Duran, Kovacs and George2009), though these largely cross-sectional studies are unable to provide insights into the dynamic interaction between depressive symptom recovery and baseline HPA axis characteristics. In contrast to the patterns associated with depression, adolescents with NSSI demonstrate a flattening of the HPA response to social stressors (Kaess et al., Reference Kaess, Hille, Parzer, Maser-Gluth, Resch and Brunner2012; Klimes-Dougan et al., Reference Klimes-Dougan, Begnel, Almy, Thai, Schreiner and Cullen2019), a pattern which appears to be related to greater NSSI severity (Başgöze et al., Reference Başgöze, Mirza, Silamongkol, Hill, Falke, Thai, Westlund Schreiner, Parenteau, Roediger, Hendrickson, Mueller, Fiecas, Klimes-Dougan and Cullen2021). Evidence for both heightened and lowered HPA activity in depression and self-injurious thoughts and behavior may reflect developmental progression and adaptation to chronic stress at the biological level of analysis. That is, the HPA system is calibrated for negative feedback, such that consistently high levels of cortisol inhibit further HPA axis activation and subsequent hypercortisolemia. For example, in the context of chronic stress, allostatic changes can recalibrate the HPA axis to a hyporeactive physiological setpoint. While this adaptation is advantageous in that it mitigates potential negative sequelae of sustained glucocorticoid-mediated signaling pathways (e.g., neurotoxicity), it is also accompanied by certain trade-offs (Ellis & Giudice, Reference Ellis and Giudice2014; Rao & Androulakis, Reference Rao and Androulakis2019). Dampened responses to stressors may leave individuals unequipped to deal with future stressful situations (Leistner & Menke, Reference Leistner, Menke, Lanzenberger, Kranz and Savic2020; Taborsky et al., Reference Taborsky, English, Fawcett, Kuijper, Leimar, McNamara, Ruuskanen and Sandi2021). As these disparate patterns of HPA activity have been observed in adolescents with depression and NSSI, two highly comorbid clinical phenomena, it is plausible that there are multiple patterns of HPA activity linked to risk and resilience processes. These patterns may be better understood when following reports of symptoms prospectively over time. Indeed, capturing fluctuations in indicators of psychopathology and well-being across time, as opposed to at a single assessment, provides better insight to the nature and stability of patterns, as resilience is dynamic and may fluctuate to some degree (Luthar et al., Reference Luthar, Cicchetti and Becker2000; Wright & Masten, Reference Wright, Masten, Goldstein and Brooks2005).

Stress perception and response selection are two additional levels of analysis germane to a more comprehensive understanding of adolescent stress response functioning and, thus, resilience mechanisms. In line with the multi-level resilience framework, considering multiple domains of stress responding may be critical to arriving at a more nuanced, comprehensive understanding of developmental patterns of psychopathology symptom recovery and well-being. Correspondence across the stress system is thought to aid in the efficacious management of stressors, through orienting to the stressor and mobilizing resources for responding to the stressor (Mauss & Robinson, Reference Mauss and Robinson2009). That is, patterns of concordance across domains of stress responses (e.g., perceived stress, behavior under stress conditions, cortisol response to stress) may indicate a well-orchestrated and flexible stress response, where the degree of hormonal response is marshaled to be commensurate with the perceived level of stress and demands to engage in goal-directed behavior in a given setting. Person-centered analytic approaches (Masten & Barnes, Reference Masten and Barnes2018; Von Eye & Bergman, Reference Von Eye and Bergman2003) are well-positioned for examining such patterns of the stress response and identifying profiles that may be protective.

Demonstrating the utility of a multi-level, person-centered approach, multi-trajectory modeling has been employed to characterize patterns of concordance across self-reported experience of stress, experimenter-rated expressions of stress, and physiological (cortisol) responses to stress (i.e., stress experience-expression-physiology concordance). As a person-centered data analytic approach, multi-trajectory modeling (Nagin et al., Reference Nagin, Jones, Passos and Tremblay2018) identifies subgroups of individuals based on shared profiles of similar intraindividual variation on growth (i.e., trajectories) across multiple indices of the phenomena of interest. In two studies, multi-trajectory modeling was applied to adolescent stress experience, expression, and physiology indices in response to a laboratory-based stressor paradigm: the Trier Social Stress Test (TSST; Kirschbaum et al., Reference Kirschbaum, Pirke and Hellhammer1993). Interestingly, these studies identified similar profiles and patterns of association with psychopathology in two different sample populations. The first study included adolescents with and without depression (Bendezú, Thai, et al., Reference Bendezú, Thai, Wiglesworth, Cullen and Klimes-Dougan2022). In this sample, three concordant profiles of stress experience, expression, and physiology were identified: High experience, High expression, and High physiology (H experi– H expres– H physio ), Moderate experience, Moderate expression, and Moderate physiology (M experi– M expres– M physio ), and Low experience, Low expression, and Low physiology (L experi– L expres– L physio ). One discordant profile was found: High experience, High Expression, and Low Physiology (H experi– H expres– L physio ). Compared to the discordant profile, the concordant L experi– L expres– L physio and M experi– M expres– M physio profiles had less severe depressive symptoms, suicide ideation, and NSSI engagement. While those with the H experi– H expres– H physio profile reported similarly severe depressive symptoms to those with the H experi– H expres– L physio profile, they reported lower severity suicide ideation and NSSI engagement, suggesting the potential protective nature of concordant stress responding.

The second study largely replicated these findings in a group of adolescents assigned female sex at birth that was oversampled for a history of NSSI (e.g., n = 36 with no NSSI history, n = 73 with NSSI history ranging from mild to severe; Carosella, Wiglesworth, et al., Reference Carosella, Wiglesworth, Bendezú, Brower, Mirza, Mueller and Klimes-DouganAccepted for Publication). In this study, three of the four previously identified profiles were found; L experi– L expres– L physio , H experi– H expres– H physio , and H experi– H expres– L physio . While the M experi– M expres– M physio profile was not found, a different fourth group was identified: High Experience–Low Expression–Moderate Physiology (H experi– L expres– M physio ). Again, when compared to those in the H experi– H expres– L physio profile, those in the L experi– L expres– L physio profile demonstrated less psychopathology including depressive symptoms and suicidal thoughts and behaviors. In this second study, however, there were not significant differences found between the H experi– H expres– L physio and H experi– H expres– H physio profiles in depressive symptoms or suicide ideation. Yet, those in the H experi– H expres– H physio profile were less likely than those in the H experi– H expres– L physio profile to report past suicide attempts. In this study, the reports of those with H experi– L expres– M physio profiles largely mirrored those with L experi– L expres– L physio profiles. Notably, no differences were found between profiles in the likelihood or severity of lifetime NSSI engagement. Though these cross-sectional studies provide greater insight into person-specific stress response profiles and the possible role of stress system concordance in resilience processes, an opportunity remains to examine the potential protective nature of stress system concordance in the context of longitudinal patterns of psychopathology and well-being.

Situated in a resilience framework, this paper expands on our prior cross-sectional work with adolescents who were oversampled for a history of NSSI (Carosella, Wiglesworth, et al., Reference Carosella, Wiglesworth, Bendezú, Brower, Mirza, Mueller and Klimes-DouganAccepted for Publication) by examining the association between the previously identified profiles of stress experience-expression-physiology concordance and developmental patterns of psychopathology and well-being across three timepoints. As there were no differences between the four identified profiles in the proportion of those with versus without a history NSSI, nor in the severity of NSSI history, a fundamental assumption of this study is that youth across all four profiles have experienced stressors requiring adaptation. For the purposes of this work, patterns of persistently low or decreasing psychopathology, including depressive symptoms, suicide ideation, and NSSI, and patterns of persistently high or increasing well-being, including positive affect, satisfaction with life, and self-worth, were viewed as being indicative of resilience (e.g., signaling adaptation to stressors associated with the adolescent transition and baseline/antecedent experiences related to psychopathology). Based on the extant theory of concordance and limited literature, we hypothesized that concordant baseline stress experience-expression-physiology profiles (L experi– L expres– L physio and H experi– H expres– H physio ) would be protective factors that confer resilience, as indicated by lesser psychopathology and greater wellbeing over time.

Method

Study procedures

All study procedures were approved by the Institutional Review Board at the University of Minnesota. Participants are from the BRain Imaging Development in Girls’ Emotion and Self (BRIDGES) Study, a longitudinal study that recruited adolescents assigned female sex at birth with varying histories of NSSI (Başgöze et al., Reference Başgöze, Mirza, Silamongkol, Hill, Falke, Thai, Westlund Schreiner, Parenteau, Roediger, Hendrickson, Mueller, Fiecas, Klimes-Dougan and Cullen2021). Data collection consisted of three waves (T1, T2, and T3). Three visits were conducted during each time point, which entailed clinical assessments administered via semi-structured interviews as well as self-report questionnaires (for T1, after completing the informed consent process with the parent or guardian and assent with the adolescent), a modified Trier Social Stress Test (TSST), and neuroimaging procedures. The primary goal of the BRIDGES Study was to use a Research Domain Criteria (RDoC) approach to examine sustained threat, cognitive control and self-processing in adolescents who were oversampled for engagement in NSSI.

Inclusion criteria for the BRIDGES Study were being assigned female sex at birth and between the ages of 12 and 17. The study sample was also oversampled for those with current or past engagement in NSSI, as we aimed to include adolescents with a range from no to severe NSSI engagement. Exclusion criteria for this study included lifetime substance use disorder (with an exception for tobacco use disorder) at T1, bipolar disorder diagnosis at T1, intellectual or developmental disability, major medical illness that would impact their neurobiology, and magnetic resonance imaging (MRI) contraindications at T1. Additional study information has been published elsewhere (Basgöze et al., Reference Başgöze, Mirza, Silamongkol, Hill, Falke, Thai, Westlund Schreiner, Parenteau, Roediger, Hendrickson, Mueller, Fiecas, Klimes-Dougan and Cullen2021).

Measures

Descriptives

Demographics. Parents and guardians reported on participant demographics, including race and ethnicity at T1 and age (coded with two decimal points following the year) and family income at all three time points. For this study, family income was dichotomized prior to inclusion as a covariate in the longitudinal models, though the disaggregated data are reported in Table 1. Income was coded categorically to indicate income < $60,000 versus ≥ $60,000 (the approximate median income for the surrounding area in 2019 at the midpoint of data collection).

Table 1. Descriptive statistics of the study sample

Note. L experi– L expres– L physio = Low experience, Low expression, and Low physiology, H experi– L expres– M physio = High experience, Low expression, and Moderate physiology, H experi– H expres– H physio = High experience, High expression, and High physiology, H experi– H expres– L physio = High experience, High expression, and Low physiology.

Clinical Diagnoses. The Kiddie Schedule of Affective Disorders and Schizophrenia Present and Lifetime Version for DSM-5 (KSADS-5; Kaufman et al., Reference Kaufman, Birmaher, Brent, Rao, Flynn, Moreci, Williamson and Ryan1997) is a semi-structured interview that was independently conducted with adolescents and the parent or guardian. This measure demonstrates excellent test-retest reliability and high interrater agreement. Adolescent and parent/guardian interviewers conducted a consensus process based on symptom reports to assign DSM-5 psychiatric diagnoses, which are used to characterize our sample herein.

Medications. Parents reported on child medication use at T1, which were included in the study as this time point coincides with the TSST administration and cortisol data collection. Medications were coded as a dichotomous variable to indicate whether the parent reported a medication that is suggested to impact the HPA axis and assessment of cortisol via saliva collection (Granger et al., Reference Granger, Hibel, Fortunato and Kapelewski2009).

Psychopathology outcomes of interest

Depressive Symptoms. The Beck Depression Inventory II (BDI-II; Beck et al., Reference Beck, Steer and Brown1996; Beck et al., Reference Beck, Steer, Ball and Ranieri1996) is a 21-item scale that measures depressive symptoms over the prior two week period. Participants responded to each item on a four-point scale, and all items were summed for a total score with higher scores indicating more severe symptoms. This measure shows good internal consistency in past research α = 0.91. The BDI-II also showed excellent internal consistency in this sample (Table S1).

Suicide Ideation. The Beck Scale for Suicidal Ideation (BSSI; Beck et al., Reference Beck, Kovacs and Weissman1979) is a 21-item scale that measures severity of suicidal thoughts, preparatory behaviors, and attempts. Participants indicate their responses on a three-point scale, and total scores were calculated as a sum of all items with higher scores indicating more severe suicide ideation. The BSSI computerized self-report version demonstrates excellent internal consistency (Cronbach α = 0.96). Past research has demonstrated the utility of sum scores of either the first five or first 19 items (de Beurs, Fokkema, de Groot, de Keijser, & Kerkhof, Reference de Beurs, Fokkema, de Groot, de Keijser and Kerkhof2015). For the present study, the sum score of the first five items was used as the outcome of interest, as items six through 19 were not administered to those who denied symptoms in items four and five. The internal consistency of the five-item BSSI in this sample was good in the current sample (Table S1).

Nonsuicidal Self-Injury. The Self-Injurious Thoughts and Behaviors Interview (SITBI; Nock et al., Reference Nock, Holmberg, Photos and Michel2007) is a semi-structured interview used to assess participant engagement in NSSI. For descriptive purposes in this study, we examined lifetime number of NSSI episodes and severity of injury from the T1 assessment to create four groups: No NSSI, Mild NSSI (fewer than four past episodes involving significant tissue damage, or unlimited NSSI episodes with no tissue damage), Moderate NSSI (four or more past NSSI episodes occurring less than once per month with significant tissue damage), and Severe NSSI (four or more past NSSI episodes occurring more than once per month with significant tissue damage). Prior work from our group has also used this approach, which resembles approaches taken by others (e.g., Esposito-Smythers et al., Reference Esposito-Smythers, Goldstein, Birmaher, Goldstein, Hunt, Ryan, Axelson, Strober, Gill, Hanley and Keller2010; Muehlenkamp et al., Reference Muehlenkamp, Brausch and Washburn2017), to identify group differences in mechanisms of sustained threat (Başgöze et al., Reference Başgöze, Mirza, Silamongkol, Hill, Falke, Thai, Westlund Schreiner, Parenteau, Roediger, Hendrickson, Mueller, Fiecas, Klimes-Dougan and Cullen2021). For our longitudinal analyses, we examined past year counts of NSSI engagement, which were log transformed to approximate a more normalized distribution (Table S1), as our outcome of interest.

Well-being outcomes of interest

Positive Affect. The Positive and Negative Affect Schedule - Expanded Form (PANAS-X; Clark & Watson, Reference Clark and Watson1994) is a 60-item self-report measure of 13 specific affective states. Participants indicate their responses on a five-point scale, and the subscale scores are calculated as a sum of constituent items. The Positive Affect subscale (10 items) score was retained, and higher scores indicate higher levels of positive affect. The Positive Affect subscale demonstrated good to excellent internal consistency (Table S1).

Satisfaction with Life. The Satisfaction with Life Scale (SLS; Diener et al., Reference Diener, Emmons, Larsen and Griffin1985) is a five-item self-report measure of an individual’s evaluative judgment of their overall satisfaction with life, independent of the affective component. Participants indicate their responses on a seven-point scale, and total scores are calculated as a sum of all items with higher scores indicating greater life satisfaction. The internal consistency of the SLS in this sample was excellent (Table S1).

Self-worth. The Self-Perception Profile for Adolescents (SPPA; Harter, Reference Harter1988) is a 45-item (nine subscales) self-report measure of adolescent self-concept. Within each domain, participants indicate their responses on a four-point scale of competence. The global self-worth subscale was retained for this study (higher scores indicate higher levels of self-worth) and showed good to excellent internal consistency in this sample (Table S1).

Trier social stress test (Stress experience, expression, and physiology)

This study examined data from the T1 modified TSST. During the modified TSST (Kirschbaum et al., Reference Kirschbaum, Pirke and Hellhammer1993; Klimes-Dougan et al., Reference Klimes-Dougan, Eberly, Westlund Schreiner, Kurkiewicz, Houri, Schlesinger, Thomas, Mueller, Lim and Cullen2014; Yim et al., Reference Yim, Quas, Cahill and Hayakawa2010), participants were instructed to prepare a five minute speech that they might give if they were to introduce themselves to a new class. They then delivered their speech to two trained examiners. After the five minute speech, the adolescent performed verbal arithmetic calculations for five minutes. Throughout the speech and math tasks, the examiners maintained neutral expressions and refrained from providing reassurance or positive feedback. The correlations between the metrics of stress experience, expression, and physiology during TSST speech and math tasks can be found in Table S2.

Following the completion of the speech and math tasks, participants rated their experience of stress during the speech preparation, speech task, math task, and post-test period (e.g., “right now”) on a five-point scale, where 1 was the lowest stress and 5 was the highest.

During the TSST, each examiner independently rated their observations of the participants’ expression of stress during the speech task and math task on a six-point scale where 1 = “Not stressed at all,” 2 = “A little stressed,” 3 = “Moderately stressed,” 4 = “Very stressed,” 5 = “Considered discontinuing the procedure because they looked so stressed,” and 6 = “Discontinued the procedure because they were showing signs of distress.” Raters were instructed to conduct ratings based on signs of stress expression including fidgeting, blushing, freezing, abrupt pauses, and verbal references to stress. The independent ratings were averaged to produce a separate mean score for each task. Examiner ratings were moderately correlated for each task (r speech = 0.64, r math = 0.72, p’s < 0.001).

Salivary cortisol output in response to the TSST served as an index of participant physiological response to stress. This procedure included five saliva samples, collected pre-task (“CORT 1”), post-speech and math task at + 15 minutes from CORT1 (“CORT2”), +30 minutes (“CORT3”), +45 minutes (“CORT4”), and + 60 minutes (“CORT5”). Participants were instructed to push saliva through a straw into a vial. Vials were stored on site in a –25°C freezer and then shipped to Universitat Trier in Trier, Germany for assay with methods consistent to (Dressendörfer et al., Reference Dressendörfer, Kirschbaum, Rohde, Stahl and Strasburger1992). Cortisol values that were greater than three standard deviations of the mean were winsorized prior to analysis.

Data analysis

Sample Characteristics. Descriptive statistics including bivariate Pearson’s and Spearman’s correlations were performed on the demographic and clinical variables to characterize the sample and nature of the data. Outcomes of interest were all linear and were checked for normality (e.g., skewness and kurtosis). Past year NSSI engagement was the only variable log transformed prior to analysis. Correlations and descriptives are in Tables S1 and S2.

Multi-trajectory Modeling. Procedures used here are described in detail in our prior manuscript (Carosella, Wiglesworth, et al., Reference Carosella, Wiglesworth, Bendezú, Brower, Mirza, Mueller and Klimes-DouganAccepted for Publication). To identify profiles of within-person variation on stress experience, expression, and physiology, we employed multi-trajectory modeling (Nagin et al., Reference Nagin, Jones, Passos and Tremblay2018) in SAS version 9.4 using the PROC TRAJ procedure (SAS Institute, Reference Institute2010) with MULTGROUPS option. The PROC TRAJ procedure utilizes Full-Information Maximum Likelihood (FIML) to estimate model parameters in the presence of missing data, a method that is most appropriate when such data are assumed to be missing completely at random (MCAR). A nonsignificant (Little, Reference Little1988) MCAR test (X 2 = 72.167, p = 0.20) suggested that FIML was an appropriate missing data approach. For the current study, multi-trajectory modeling followed three steps: a) model specification, b) model adequacy evaluation, and c) trajectory distinction analysis. To specify the best fitting model, linear, quadratic, and cubic polynomial parameters were initially estimated for stress expression (two data points), experience (four data points), and physiology (five data points) trajectories, respectively. For each model specification step (e.g., one-profile solution, two-profile solution), non-significant polynomial parameters were removed until a solution containing only significant polynomial parameters was obtained. As recommended (Andruff et al., Reference Andruff, Carraro, Thompson, Gaudreau and Louvet2009; Helgeson et al., Reference Helgeson, Snyder and Seltman2004; Louvet et al., Reference Louvet, Gaudreau, Menaut, Genty and Deneuve2009), linear polynomial parameters were retained irrespective of statistical significance.

After this solution was obtained, the log Bayes factor approximation [2loge(B10)] was examined to assess model fit, where a log Bayes factor greater than 10 is said to be strong evidence for the superior fit of the more complex model (e.g., three-profile solution) relative to the less complex model (e.g., two-profile solution) (Jones & Nagin, Reference Jones and Nagin2007). Given our sample size, model specification was limited to four profiles. After the best fitting model was specified, we evaluated the adequacy of the solution with three metrics: average posterior probability (AvePPj), odds of correct classification (OCCj), and the ratio of the probability of profile assignment to the proportion of adolescents assigned to profiles (Probj/Propj) (Nagin et al., Reference Nagin, Jones, Passos and Tremblay2018). An AvePPj greater than 0.70, OCCj greater than 5.00, and Probj/Propj close to 1.00 provide evidence of multi-trajectory model adequacy. Following adequacy evaluation, trajectory distinction analyses were conducted. Specifically, a series of Wald tests compared intercept and polynomial parameter estimates for each identified profiles’ stress, experience, expression, and physiology trajectories. The results of these analyses help to show how baseline levels and reactivity patterns for each trajectory are relatively “higher” or “lower” across the identified profiles and, thus, how trajectories across profiles were distinct from one another.

Longitudinal Modeling. Sample sizes for each outcome measure at each time point are presented in Table S1. To better understand potential patterns of missing data that might inform our analyses or interpretations of our results, we conducted logistic regression analyses to determine whether T1 demographic or clinical characteristics predicted patterns of missing data across the entire sample and within each profile at T2 and T3 (see Tables S3–S8).

To model whether the baseline profiles of stress experience, expression, and physiology were associated with changes over time in outcomes of interest (i.e., depressive symptoms, suicide ideation, past year NSSI engagements, positive affect, global self-worth, satisfaction with life), we conducted a series of linear mixed-effect regression models using the lme4 package (Bates et al., Reference Bates, Mächler, Bolker and Walker2015) in R version 4.2.2 (R Core Team, 2022). These models account for the nested structure of the longitudinal data (e.g., observations nested within individuals). Each model included the fixed effect of profile and time point as main effects as well as their interaction. Profile was coded categorically, where each concordant stress response profile (L experi– L expres– L physio and H experi– H expres– H physio ) was entered as the reference group in separate analyses. This decision was based on our interest in understanding stress response concordance, as opposed to discordance, as a protective factor. Time point was also coded categorically and T2 was set as the reference time point, which allowed us to examine change in slope from T1 to T2 and T2 to T3. The profile by time point interaction allowed us to examine whether the slope (e.g., change in outcome) from T1 to T2 and from T2 to T3 differed between the profiles. Age (e.g., associated with size and structure of key regulatory glands; Linton & Dorshkind, Reference Linton and Dorshkind2004), income (e.g., associated with HPA hypoactivation; Evans et al., Reference Evans, Chen, Miller and Seeman2012; Lupien et al., Reference Lupien, King, Meaney and McEwen2000), and HPA acting medication use at T1 (Granger et al., Reference Granger, Hibel, Fortunato and Kapelewski2009) were all included as covariates given their theoretical links to stress responses and outcomes of interest. The covariate of age accounted for both age and time between assessments. In cases where income was missing at a given visit, the most recent income value was used instead. Given the limited variability in race/ethnicity in our sample and the limited correlation of minoritized race/ethnicity with outcomes of interest, this variable was not included as a covariate in our models. The emmeans R package (Lenth, Reference Lenth2023) was used to produce contrasts from the fitted linear mixed-effect models to test within group changes in indexes of psychopathology and well-being over time (e.g., whether slopes differed significantly) and between-group differences at each time point. Data visualizations were produced using the ggplot2 R package (Wickham, Reference Wickham2016). While we use the standard p = 0.05 threshold to determine whether results are statistically significant, we also report results where p ≤ 0.1 below given the possibility of limited statistical power to detect small to medium effects when comparing discrete groups.

Results

Overall sample characteristics

A total of 164 adolescents were enrolled in the BRIDGES Study. Of these youth, 113 had usable salivary cortisol data from the T1 TSST and were included in multi-trajectory modeling analyses. Finally, of those youth, four participants were missing income data across all timepoints, making our final sample of participants 109 adolescents ages 12.20–17.02 at T1. Descriptive statistics of age, race and ethnicity, and income can be found in Table 1. In the current sub-sample, based on the T1 SITBI, NSSI severity groups were as follows: No NSSI (n = 35); Mild NSSI (n = 10); Moderate NSSI (n = 41); Severe NSSI (n = 23). Those with a history of NSSI were more likely than those without to meet criteria for at least one clinical diagnosis (94.5% versus 47.1%). Current symptoms consistent with Major Depressive Disorder were the most common in our sample (53.2%) followed by Attention-deficit Hyperactivity Disorder (29.4%) and Generalized Anxiety Disorder (25.7%). Suicide ideation (62.4%), planning (39.4%), and attempts (31.2%) were also endorsed frequently by youth in our sample.

Multi-trajectory modeling

Results from multi-trajectory model analyses in this sample are reported in extensive detail elsewhere (Carosella et al., Reference Carosella, Wiglesworth, Bendezú, Brower, Mirza, Mueller and Klimes-DouganAccepted for Publication). Relevant parameter estimates, adequacy indices, and results of our trajectory distinction analyses can be found in the supplement (Table S9). Multi-trajectory model specification arrived at increasingly complex solutions: two- to one-group comparison [2loge(B10) = 263.66], three- to two-group comparison [2loge(B10) = 56.88], four- to three-group comparison [2loge(B10) = 10.46]. Following the methods outlined in (Nagin, Reference Nagin2005) for model specification and model adequacy evaluation, the four-group model was selected and found to fit the data well. To characterize the resultant profiles, we relied on significantly differing aspects of the trajectories (e.g., intercepts, response patterning) to illustrate how the profiles differed from one another.

As previously discussed, the four profiles identified were characterized as L experi– L expres– L physio (n = 41), H experi– H expres– L physio (n = 28), H experi– H expres– H physio (n = 19), and H experi– L expres– M physio (n = 25; Fig. 1). Four individuals included in our multi-trajectory models were missing gross income data, which was a covariate in our longitudinal analyses, resulting in smaller group sizes for the L experi– L expres– L physio (n = 39) and H experi– H expres– L physio (n = 26) profiles in this study. No significant group differences between the profiles based on age, proportion of racially or ethnically minoritized youth, or gross family income (<$60,000 versus ≥$60,000) were evident (all p’s > 0.05). Further, as previously described, there were no significant differences between profiles in NSSI history or severity at T1 (see Table S10).

Figure 1. Profiles indicated in the final four-group solution from the multi-trajectory model. raw scores for experience, expression, and physiology are indicated on the y-axis; a trier social stress test event markers for physiology are estimated based on the time-lag in salivary cortisol presentation; Lexperi–Lexpres–Lphysio = low experience, low expression, low physiology, Hexperi–Hexpres–Hphysio = high experience, high expression, high physiology, Hexperi–Lexpres–Mphysio = high experience, low expression, moderate physiology, Hexperi–Hexpres–Lphysio= high experience, high expression, low physiology.

Psychopathology predictive models

When comparing longitudinal patterns of psychopathology between the L experi– L expres– L physio profile and the two discordant stress profiles, results indicated that adolescents with the H experi– H expres– L physio profile demonstrated significantly higher depressive symptoms at T2 (B = 0.86, p < 0.001; Table S11). Post-hoc contrasts revealed that differences in depressive symptoms between these profiles were also significant at T1 and T3 (Table S12). However, the changes in depressive symptoms from Y1 to Y2 and from Y2 to Y3 were not statistically significant within each profile (Table S13), nor statistically different between the two profiles, indicating similarly flat patterns of depressive symptoms for the L experi– L expres– L physio and H experi– H expres– L physio profiles. The difference in suicide ideation severity between the L experi– L expres– L physio and H experi– H expres– L physio profiles was statistically significant at T1, though not statistically significant and moderately sized at T2 and T3 (Table S12). There were no significant or trend-level effects for NSSI engagement when comparing the two discordant profiles to the L experi– L expres– L physio profile.

Parameter estimates for longitudinal models of psychopathology where the H experi– H expres– H physio profile is the reference group can be found in Table 2. Adolescents with the concordant H experi– H expres– H physio profile demonstrated a trend of greater decreases in depressive symptoms from T2 to T3 as compared to those with the H experi– H expres– L physio profile (B = 0.71, p = 0.052). Similarly, a trend emerged when comparing the H experi– H expres– H physio profile to the H experi– L expres– M physio profile from T2 to T3 (B = 0.60, p = 0.085). Longitudinal patterns of suicide ideation severity were similar to the depressive symptom patterns from T2 to T3 (see Fig. 2), again with a trend in the data where those with the H experi– H expres– H physio profile reported a greater decrease in suicide ideation than did those with the H experi– H expres– L physio profile (p = 0.100), though this non-significant finding may be due to limited statistical power. Post-hoc contrasts indicated that, following these trends, those with the H experi– H expres– L physio profile reported significantly higher suicide ideation severity at T3 but not at T1 or T2 than those with the H experi– H expres– H physio profile (Table S13). Indeed, despite having similar suicidal ideation severity scores at T1, at T3 youth with the H experi– H expres– H physio profile reported very limited suicidal ideation, whereas the reported suicidal ideation of those with the H experi– H expres– L physio profile were, on average, unchanged from T1 to T3. No significant interaction or main effects for time point and profile were identified when examining longitudinal patterns of past year NSSI engagements when comparing discordant profiles to the concordant H experi– H expres– H physio profile (Table 2). Despite the lack of significant findings, the H experi– H expres– L physio profile group demonstrate an approximate 40% decrease in mean past year NSSI engagement between T1 and T3.

Figure 2. Longitudinal patterns of reporting of symptoms of psychopathology disaggregated by profile. Lexperi–Lexpres–Lphysio = low experience, low expression, low physiology, Hexperi–Hexpres–Hphysio = high experience, high expression, high physiology, Hexperi–Lexpres–Mphysio = high experience, low expression, moderate physiology, Hexperi–Hexpres–Lphysio= high experience, high expression, low physiology.

Table 2. Linear mixed effect regression model results demonstrating group differences in indices of psychopathology over time based on stress experience, expression, and physiology profile

** p < 0.01.

* p < 0.05.

x p ≤ 0.10.

Well-being predictive models

When comparing longitudinal patterns of well-being between the L experi– L expres– L physio profile and the two discordant stress profiles, there were several significant main effects (Table S14). Compared to adolescents with the L experi– L expres– L physio profile, those with the H experi– H expres– L physio profile demonstrated significantly lower global self-worth (B = -0.68, p = 0.018) and satisfaction with life (B = -0.66, p = 0.017) at T2; global self-worth was also significantly lower for those with the H experi– H expres– L physio profile at T1 and T3 (Table S15). At T2, those with the H experi– L expres– M physio profile also showed significantly lower global self-worth compared to those with the L experi– L expres– L physio profile (B = -0.70, p = 0.015). Similar to the psychopathology models, the changes in well-being indexes from T1 to T2 and from T2 to T3 within each discordant profile were not significantly different from the changes over time in the L experi– L expres– L physio profile, indicating similar longitudinal patterns of well-being across these groups (Table S14, S16).

Parameter estimates for longitudinal models of well-being where the H experi– H expres– H physio profile is set as the reference group are found in Table 3. Results from longitudinal models of positive affect revealed a significant profile by time point interaction where those with the concordant H experi– H expres– H physio profile demonstrated greater increases in positive affect from T1 to T2 as compared to the discordant H experi– L expres– M physio profile (B = 1.00, p = 0.009), which demonstrated a decreasing trend in positive affect over this time period. Although differences in positive affect from T1 to T2 and T2 to T3 between the between the H experi– H expres– H physio and H experi– H expres– L physio profiles were not statistically significant (p = 0.096 and p = 0.191 respectively), visual inspection illustrates that those in the H experi– H expres– H physio profile showed an upward trend in positive affect over time that was not observed for those in other profiles (Figure 3). Similarly, when examining global self-worth, the concordant H experi– H expres– H physio profile demonstrated a marginally greater increase in self-worth from T2 to T3 as compared to the H experi– H expres– L physio profile (B = -0.88, p = 0.055). The H experi– H expres– L physio profile demonstrated a trend of decreasing self-worth over that same time period, resulting in significantly different reports of self-worth at T3 (Table S15). There were no notable trends when comparing profiles on satisfaction with life, as these scores largely did not change over time (Figure 3).

Figure 3. Longitudinal patterns of reporting of symptoms of well-being disaggregated by profile. Lexperi–Lexpres–Lphysio = low experience, low expression, low physiology, Hexperi–Hexpres–Hphysio = high experience, high expression, high physiology, Hexperi–Lexpres–Mphysio = high experience, low expression, moderate physiology, Hexperi–Hexpres–Lphysio= high experience, high expression, low physiology.

Table 3. Linear mixed effect regression model results demonstrating group differences in indices of well-being over time based on stress experience, expression, and physiology profile

*** p < 0.001.

** p < 0.01.

* p < 0.05.

x p ≤ 0.10.

Discussion

This study examined whether profiles of multimodal stress response, defined using a person-centered approach, predicted patterns of resilience (as indicated by trends in psychopathology and well-being over time) in adolescents with a range of histories of NSSI engagement. Key strengths of the study included the longitudinal design, the focus on a high-risk sample of adolescents with high rates of NSSI and psychopathology histories, and the person-centered, multidimensional characterization of stress response trajectories that did not rely solely on self-report data. While three of the profiles demonstrated broadly resilient patterns indicating improvement in well-being and diminishing psychopathological symptoms over time, the H experi– H expres– L physio profile showed the poorest outcomes with respect to longitudinal trends, including tempered patterns of recovery in depression and suicide ideation severity, and blunted positive affect and self-worth over time. However, it is notable that this profile showed the greatest reduction of NSSI engagement over time, though these slopes were non-significant.

Much of the adolescent resilience literature focuses on vulnerability to developing psychopathology using high-risk samples, where “high-risk” is often as defined by varied stress exposures and “resilience” is manifested as an absence of psychopathology. However, in keeping with recent calls for research that move away from an absence-of-deficit approach to defining resilience (Fergus & Zimmerman, Reference Fergus and Zimmerman2005; Masten et al., Reference Masten, Lucke, Nelson and Stallworthy2021; Yates et al., Reference Yates, Tyrell and Masten2015), the current work posits that resilience mechanisms can continue to operate even after the development of symptoms of psychopathology. In other words, those who experience psychopathology and/or struggle with suicidal thoughts and behaviors and NSSI may still draw upon protective mechanisms, which in turn promote recovery. Thus, we examined a sample of adolescents who, at the onset of the study, were already experiencing high rates of psychopathology including suicidal thoughts and behaviors and self-injury. By examining whether concordant stress response profiles predicted longitudinal patterns of psychopathology and well-being indicative of resilience, we uncovered patterns that generate hypotheses about potential protective mechanisms facilitating recovery in adolescence.

Here we focused on the stress response system as a potential key protective factor in influencing resilience patterns in adolescence. In particular, based on theory of stress correspondence and our own past work (Carosella et al., Reference Carosella, Wiglesworth, Bendezú, Brower, Mirza, Mueller and Klimes-DouganAccepted for Publication), we hypothesized that concordance across the experience, expression, and physiological levels of the stress response would confer resilience. As hypothesized, adolescents with the two concordant profiles showed the most consistent patterns of resilience, with H experi– H expres– H physio profiles showing largely consistent trends of recovery (e.g., an approximate 50% decrease in depressive symptoms on average from T1 to T3), and adolescents with L experi– L expres– L physio profiles showing maintained patterns of relatively low psychopathology and high well-being. In contrast, adolescents in the H experi– L expres– M physio profile showed less consistent outcomes, with relatively low psychopathology but some fluctuations (though not statistically significant) in positive affect over time. Noteably, similar to previous cross-sectional findings (Bendezú et al., Reference Bendezú, Thai, Wiglesworth, Cullen and Klimes-Dougan2022), those in the H experi– H expres– L physio profile demonstrated a non-resilient pattern in which depressive symptoms and suicide ideation severity largely persisted over time, and indexes of well-being remained low or trended downward.

Comparing longitudinal patterns of symptomatology between those with the H experi– H expres– L physio and the H experi– H expres– H physio profiles provides compelling evidence that, although the allostatic change of recalibration of the HPA system in the context of chronic stress may be adaptive in the short-term, this pattern may also have long-term consequences. That is, this pattern where the perception of threat is high but is not accompanied by a commensurately high physiological response may reflect a broader process of allostatic loading which incurs risk for delayed or dampened recovery from psychopathology and impaired growth in well-being. Similarly, another clear contrast emerges when comparing the L experi– L expres– L physio and H experi– H expres– L physio profiles, both of which are characterized by low physiological response to stress. However, adolescents with the L experi– L expres– L physio profile demonstrated a stress response that was commensurately low with their low perceived and exhibited stress levels, which may reflect that they did not experience the TSST to be particularly stressful, rather than reflecting allostatic change. Accordingly, differences between these profiles in symptoms of psychopathology and well-being between were striking; results indicated that the longitudinal patterns of psychopathology and well-being were similarly flat (e.g., limited change year-to-year) for these two groups, but that for the L experi– L expres– L physio profile this flat patterning indicated persistently lower psychopathology and higher well-being, and for the H experi– H expres– L physio profile this flat patterning indicated persistently high psychopathology and low well-being. For example, the H experi– H expres– L physio demonstrated depressive symptoms severity suicidal ideation severity scores about two times higher than those of the L experi– L expres– L physio profile across the three timepoints. Thus, these results suggest that a low physiological response to stress may only confer resilience when it is consistent with a lower experience and expression of threat. Importantly, those with L experi– L expres– L physio response profiles showed arguably the most optimal longitudinal outcomes, though adolescents with this profile were not less likely than their peers to have some history of NSSI (e.g., participants recruited based on having no history of NSSI at T1 were not overrepresented in the group characterized by this profile of stress response). Thus, this concordant profile was a predictor of favorable trends in psychopathology and well-being despite the likely presence of stressors associated with adolescent NSSI engagement.

The patterns that emerged when comparing adolescents with the L experi– L expres– L physio and H experi– H expres– L physio profiles on longitudinal trends of suicide ideation and NSSI are consistent with findings from a prior person-centered, multi-level study that similarly investigated a sample of female adolescents at risk for self-injurious thoughts and behaviors (Bendezú, Calhoun, Patterson, et al., Reference Bendezú, Calhoun, Patterson, Findley, Rudolph, Hastings, Nock and Prinstein2022). In that study, youth who experienced more pronounced negative affect responses and less pronounced salivary cortisol responses to the TSST were more likely to report suicide ideation, suicide attempt, and NSSI at three-to-nine-month follow-up relative to their counterparts who experienced less pronounced negative affect and less pronounced cortisol responses. Notably, female youth in the high negative affect/low salivary cortisol and low negative affect/low salivary cortisol profiles also exhibited the lowest and highest basal positive affect levels, respectively, in the sample, which is consistent with conceptualizations of adolescent risk for self-injurious thoughts and behaviors as involving disturbance of positive valence systems (Craske et al., Reference Craske, Meuret, Ritz, Treanor and Dour2016; Hankin, Reference Hankin2012). Though these studies differ with respect to indices utilized to extract the profiles (e.g., emotion experiences vs. stress expression and experience), this partial overlap in profiles identified as well as similar longitudinal links to self-injurious thoughts and behaviors suggests that person-centered, multi-level approaches (such as multi-trajectory modeling) to the study of adolescent stress response functioning may reliably identify meaningful subgroups of female adolescents with varied levels of risk for psychopathology.

While the current study is built upon the hypothesis that concordance in stress processing may facilitate resilience processes and predict positive future outcomes, it is important to acknowledge the nuances unveiled in these results. For example, examining satisfaction with life results suggested that, while those in the H experi– H expres– L physio profile showed the lowest and flattest pattern over time with scores that reflect slight dissatisfaction with life (e.g., score < 20; Diener et al., Reference Diener, Emmons, Larsen and Griffin1985), those in the H experi– H expres– H physio profile also showed relatively lower satisfaction with life compared to their peers. It is possible that in this context, while the concordance at the physiological level may be protective, the high psychological burden of stress still interferes with life satisfaction. To highlight further nuances, while the H experi– H expres– H physio profile showed consistent improvements generally in symptoms of psychopathology and well-being, adolescents in the discordant H experi– L expres– M physio profile maintained relatively low psychopathology across the three time points, suggesting that this profile may also be protective, despite its discordant nature. The only exception to this interpretation, however, is the somewhat unstable reports of positive affect which vary (non-significantly) over time. Thus, the ways that stress response patterns map on to resilience may not simply be explained by the bifurcation of concordance and discordance.

Moreover, the potential protective nature of stress response concordance does not seem to apply indiscriminately to all outcomes, even highly related outcomes, of psychopathology and well-being. Indeed, depression, suicide ideation, and NSSI are highly related constructs that have some unique correlates and possibly unique mechanisms (Başgöze, Wiglesworth, et al., Reference Başgöze, Wiglesworth, Carosella, Klimes-Dougan and Cullen2021). While the patterns of depression and suicide ideation were similar when comparing groups, those for NSSI engagement were less similar (e.g., adolescents in all profiles except for L experi– L expres– L physio reporting largely similar past year NSSI engagement at T3). Further, while adolescents in the H experi– H expres– H physio profile demonstrated somewhat consistent trends in decreasing psychopathology over time and increasing positive affect and self-worth, this patterning did not map on qualitatively to their patterns of satisfaction with life, which trended downward from T1 to T2 and upward from T2 to T3, resulting in no significant change over the course of the study. Ultimately, it is our hope that this study acts as a catalyst for hypothesis generation regarding mechanisms that facilitate resilience among those already experiencing psychopathology during adolescence.

Limitations and future directions

While our study had many strengths in the design, there are limitations that are important to acknowledge and be pursued as future directions for this line of work. First, considerations of internal validity must be noted when interpreting the present results. While it is critical to capture the perceptions of adolescents to assess outcomes over time, and the measures used to evaluate psychopathology and well-being were primarily well-validated indexes, the observed longitudinal patterns of psychopathology and well-being are all subject to biases associated with self-report measures, and in some cases limited to count variables as an index of severity (e.g., NSSI) or have restricted variability (e.g., BSSI, SLS). Further, while diminished or low levels of psychopathology and evidence of high or increased well-being are reasonable ways to capture resilience processes, future work may benefit from conceptualizing resilience as adaptation or competence across a larger variety of domains (Luthar et al., Reference Luthar, Cicchetti and Becker2000).

Our study posits that the stress response profiles may serve as a protective factor that promotes coping or successful adaptation to stressful events. A recent meta-analysis concluded that the TSST is a useful tool for eliciting a stress response in youth (Seddon et al., Reference Seddon, Rodriguez, Provencher, Raftery-Helmer, Hersh, Labelle and Thomassin2020). Yet, there remains uncertainty around the interpretation of a low cortisol response to a moderate stressor such as the TSST. Indeed, it is possible that for some who have high distress tolerance, this paradigm was not sufficient to elicit a stress response (e.g., possibly explaining the L experi– L expres– L physio response profile). For others, the TSST may have elicited responses across other biological systems (e.g., neural stress response, sympathomedullary pathway [SAM]) which may be asymmetrical to HPA axis responding (e.g., Başgöze, Mirza, et al., Reference Başgöze, Mirza, Silamongkol, Hill, Falke, Thai, Westlund Schreiner, Parenteau, Roediger, Hendrickson, Mueller, Fiecas, Klimes-Dougan and Cullen2021; Pham et al., Reference Pham, Bendezú and Wadsworth2023); such patterns could be related to allostatic overload and HPA recalibration (e.g., a hypothetical explanatory factor for the H experi– H expres– L physio response profile). Additionally, however, the assessment of cortisol can be impacted by many factors such as time of day, phase of the menstrual cycle, and diet. While the study methods took care to consider these factors and, when possible, control for them, there is still room for potential error. In addition to more thoroughly accounting for these factors in the identification of stress response profiles, a future line of inquiry would be to evaluate how stress response profiles themselves may change over time as a potentially dynamic indicator of resilience. While this kind of fine-grained longitudinal data on dynamic change in stress response together with dynamic change of psychopathology will be difficult to collect, it will be necessary to resolve ongoing questions in the field.

Second, this study has limitations regarding external validity and generalizability related to the sample size and make up. Our sample size at baseline was modest. However, participant retention was a challenge (particularly given that data collection occurred amid the COVID-19 pandemic), and the available data at follow-ups was more limited. The diminished sample size at T2 and T3 presents the possibility that our study was underpowered to detect small to medium effect sizes when comparing discrete groups (e.g., profiles) in the analyses undertaken. Additionally, the study sample had limited diversity with respect to race, ethnicity, sex, gender, and socioeconomic status, limiting generalizability. Interpersonal and systemic racism, homophobia, and transphobia as well as culturally and community bound protective factors are inextricable from stress processes, psychopathology, and resilience. However, our small sample sizes of participants from different minoritized racial or ethnic backgrounds as well as gender identities and sexual orientations meant it was not possible to disaggregate these identities at the broadest levels in our models. Moreover, we did not directly measure experiences of discrimination (e.g., interpersonal and systemic) that may contribute to group differences between minoritized youth and majority youth with regard to stress profiles or resilience patterns. Given these limitations in measurement and sampling, race/ethnicity, gender identity, and sexual orientation were not included as covariates in our models. Future work will benefit from examining these questions about stress responding and resilience in racially and ethnically diverse samples or in within-group research designs and moving beyond using race as a proxy for racism and discrimination, instead measuring these factors directly (Azibo, Reference Azibo1988; Lett, Asabor, Beltrán, Cannon, & Arah, Reference Lett, Asabor, Beltrán, Cannon and Arah2022; Needham et al., Reference Needham, Ali, Allgood, Ro, Hirschtick and Fleischer2022). Similarly, in recruiting diverse samples with respect to gender identity and sexual orientation, the field can continue to understand how minority stress that permeates lesbian, gay, bisexual, transgender, queer, and expansive (LGBTQ+) experiences relate to these processes of interest (e.g., Hatzenbuehler & McLaughlin, Reference Hatzenbuehler and McLaughlin2014). Notably, all participants in this sample were assigned female sex at birth. Given that internalizing pathology and stress responses typically differ between female and male adolescents, it is likely that these results do not fully generalize to male adolescents (Mazurka et al., Reference Mazurka, Wynne-Edwards and Harkness2018). Finally, the proportion of the population especially likely to enroll in this research study may have involved families who, due to various forms of privilege (e.g., resources available to find a provider), were more likely to be attentive to their adolescents’ psychopathology or receptive to intervention. Thus, there may be factors related to the sample that self-select for such a study that also relate to resilience patterns.

In the current study, we posited that the adolescent period is accompanied by a number of stressors, both indexed and perpetuated by adolescent-onset psychopathology. Thus, by oversampling for those with NSSI, we expect that our study sample was enriched for adolescent stress experiences. However, there are several external factors not addressed in this study, including de facto measures of stress exposure prior to baseline and across follow-up, each of which frame interesting and worthwhile threads for future research on stress responses and resilience. Antecedent experiences of severe maltreatment, as one example, likely contribute to physiological and neurobiological stress response systems and psychopathology over time (Hosseini-Kamkar et al., Reference Hosseini-Kamkar, Lowe and Morton2021; McLaughlin et al., Reference McLaughlin, Weissman and Bitrán2019; Schär, Mürner-Lavanchy, Schmidt, Koenig, & Kaess, Reference Schär, Mürner-Lavanchy, Schmidt, Koenig and Kaess2022). Moreover, adaptive physiological stress processing is bolstered by experiences within parent-child dyads, families, and neighborhood contexts (Gunnar & Donzella, Reference Gunnar and Donzella2002). For example, parental emotion socialization may be an antecedent experience that can bolster advantageous responses to stress, potentially leading to less severe symptomatology and more positive adaptation over time (White et al., Reference White, Silamongkol, Wiglesworth, Labella, Goetz, Cullen and Klimes-Dougan2021). In this naturalistic longitudinal study, involvement in treatment was not included as an exclusion criterion as part of our study procedures. Instead, our study team frequently provided recommendations for treatment when adolescents were identified to need a higher level of care than was presently being provided to them. However, it is unknown whether involvement in treatment, due to or separate from recommendations provided by our team clinicians, impacted youths’ psychopathology and well-being. As a final example, the potential stressors incurred from the COVID-19 pandemic, and their impacts on longitudinal patterns of psychopathology and well-being, were not addressed in this study. While all participants in our study sample completed the T1 assessment including the TSST prior to the onset of the pandemic, many adolescents completed the T2 and T3 time points at various stages during the pandemic. Thus, adolescents’ stress response profiles may have had an effect on how youth navigated pandemic related stressors, and pandemic related stressors may have played an important role in modulating adolescents’ patterns of psychopathology and well-being over time (Carosella et al., Reference Carosella, Wiglesworth, Silamongkol, Tavares, Falke, Fiecas, Cullen and Klimes-Dougan2021; Carosella, Mirza, Başgöze, Cullen, & Klimes-Dougan, Reference Carosella, Mirza, Başgöze, Cullen and Klimes-Dougan2023).

Continued work investigating multi-level markers of adaptive processes is critical to advancing a process-informed understanding of resilience (Cicchetti, Reference Cicchetti2010). In particular, expanding the repertoire of physiological and neurobiological measures will continue to aid in uncovering the multifactorial processes underlying protection and recovery. This work focused specifically on the HPA axis as a key biological component of the response to threat, although other mechanisms including those which may be assessed by neuroimaging are promising next steps. Recent work by our group suggested that stress system concordance was associated with greater positive frontolimbic connectivity (Bendezú, Thai, Wiglesworth, Cullen, & Klimes-Dougan, Reference Bendezú, Thai, Wiglesworth, Cullen and Klimes-Dougan2022). Further, we hope that the present study can spur future hypotheses about the role of concordance across the levels of the stress response system as it pertains to resilience. Linking molecular markers (e.g., peripheral hormones, oxidative stress mechanisms, gene expression, epigenetic modification) to brain and behavior could clarify coactions and bidirectional influences.

Implications for intervention

Our study contributes to a burgeoning literature highlighting the promise of multiple-levels-of-analysis, person-centered approaches in characterizing risk and resilience processes (Cicchetti & Dawson, Reference Cicchetti and Dawson2002; Cicchetti & Rogosch, Reference Cicchetti and Rogosch1996), with implications for the evaluation of interventions designed to move biological rhythms implicated as resilience mechanisms. For example, awareness of the utility of multi-system approaches (e.g., analyzing HPA function in tandem with peripheral systems) in distinguishing well-regulated versus dysregulated biological function has begun to increase (Bendezú, Calhoun, et al., Reference Bendezú, Calhoun, Vinograd, Patterson, Rudolph, Giletta, Hastings, Nock, Slavich and Prinstein2022; Buss et al., Reference Buss, Jaffee, Wadsworth and Kliewer2018; Chen et al., Reference Chen, Raine and Granger2020). However, biologically potent interventions have overwhelmingly focused on a single biomarker, notably cortisol (Dozier, Roben, Caron, Hoye, & Bernard, Reference Dozier, Roben, Caron, Hoye and Bernard2018; Fisher et al., Reference Fisher, Beauchamp, Roos, Noll, Flannery and Delker2016). Restricting analysis to single systems (e.g., HPA axis in isolation) may therefore limit understanding unique patterns of comprehensive psychobiological dysregulation (e.g., experience-expression-physiology stress response concordance) that may be more or less amenable to change resulting from intervention, precluding full understanding of an intervention’s efficacy. Indeed, while some studies show evidence of HPA recalibration in the context of recovery from depressive symptoms (Aihara et al., Reference Aihara, Ida, Yuuki, Oshima, Kumano, Takahashi, Fukuda, Oriuchi, Endo, Matsuda and Mikuni2007; Yuuki et al., Reference Yuuki, Ida, Oshima, Kumano, Takahashi, Fukuda, Oriuchi, Endo, Matsuda and Mikuni2005), others suggest that HPA responses remain altered even following symptom remission (Bhagwagar et al., Reference Bhagwagar, Hafizi and Cowen2003; Lange et al., Reference Lange, Zschucke, Ising, Uhr, Bermpohl and Adli2013). Given that studies now show experience-expression-physiology stress response concordance to be linked to psychopathology and well-being both concurrently and longitudinally, one logical next step might be to utilize person-centered, multi-level approaches to study whether youth with discordant stress response profiles (e.g., H experi– H expres– L physio ) recalibrate towards stress response concordance (e.g., L experi– L expres– L physio ) following intervention for depression and related sequelae (e.g., suicide ideation, NSSI) and whether such recalibration is linked to symptom reduction at follow-up. Indeed, a recent study employing a multiple-levels-of-analysis, person-centered design showed recalibration of dysregulated HPA-SAM stress response asymmetry (e.g., High HPA-Low SAM) towards well-regulated HPA-SAM stress response symmetry (e.g., Low HPA-Low SAM) as a function of assignment to psychosocial intervention (Bendezú & Wadsworth, Reference Bendezú and Wadsworth2023). Whether similar findings emerge when experience-expression-physiology stress response concordance is the mechanism of risk and therapeutic action remains to be seen.

Moreover, the present findings provide an opportunity to identify individuals which may be in the most need of treatment due to their risk for persistent patterns of psychopathology and low well-being. The results suggest that individuals demonstrating a discordant response to threat (e.g., H experi– H expres– L physio ) may be in particular need of targeted intervention. For example, adolescents reporting high experience of stress may benefit from dialectic behavioral therapy and other approaches involving mindfulness to help them regulate and navigate intensely stressful situations. And yet, a multi-systemic approach may be necessary, including both individual-focused interventions (e.g., evidence-based, skills-focused psychotherapy) and parent-focused interventions designed to support parents and provide training in how they relate to their adolescents who are suffering from depression (e.g., Reigstad et al., Reference Reigstad, Gunlicks-Stoessel, Westervelt and Gearity2022). Interestingly, some alterations in the environment have been accompanied by enhanced responsiveness of the HPA axis (Fisher et al., Reference Fisher, Van Ryzin and Gunnar2011). In a randomized control intervention with teachers, students in the banking time condition (which aimed to improve a teachers interaction quality with a particular child) showed a steeper diurnal slope at the end of the school year than children in the low and no treatment conditions (Hatfield & Williford, Reference Hatfield and Williford2017). Similarly, institutionalized children who are adopted generally show increased diurnal cortisol, however those who show flatter diurnal patterns post-adoption demonstrate more behavioral and emotional problems (Kroupina et al., Reference Kroupina, Fuglestad, Iverson, Himes, Mason, Gunnar, Miller, Petryk and Johnson2012). While the potential therapeutic nature of restoring stress response concordance is speculative at this juncture, process-focused approaches which facilitate the adaptive functioning of the basic stress response system do present as a compelling route to intervention (Yates et al., Reference Yates, Tyrell and Masten2015).

Conclusion

In addition to promoting positive adaptation that may stave off the development of psychopathology, resilience processes may aid in the recovery from psychopathology and cultivation of well-being in adolescence. This study showed that using a person-centered approach to characterize multimodal stress response profiles in at-risk adolescents was useful in understanding how concordant stress responses act as a protective factor in adolescence to predict resilient patterns of depressive symptoms and self-worth over time. This work extends past research by suggesting that these experience, expression, and physiological response profiles not only highlight concurrent risks, but are also likely to have implications for resilience by predicting changes over time in both psychopathology and well-being. Taken together, our findings help to identify possible pathways toward developing personalized interventions directed at fostering resilience and ultimately reducing the burden of distress and despair associated with adolescent-onset psychopathology. We hope that the current work will invite future research examining multiple dimensions of resilience and how these dimensions unfold over time even after the onset of adolescent psychopathology.

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/S0954579423000731.

Acknowledgements

We honor the Očhéthi Šakówiŋ and Anishinabewaki ᐊᓂᔑᓈᐯᐗᑭ, the Original Peoples of the land on which we live and work. This acknowledgement is a small piece of the larger goal to disrupt colonialism and Indigenous erasure and restore Indigenous Peoples to their lands. The authors also sincerely thank the adolescents and family members who participated in the study to make this research possible.

Funding statement

This study was funded by the National Institute of Mental Health (KRC, grant number R01MH107394). Author AW is funded by the National Science Foundation Graduate Research Fellowship Program.

Competing interests

None.

References

Aihara, M., Ida, I., Yuuki, N., Oshima, A., Kumano, H., Takahashi, K., Fukuda, M., Oriuchi, N., Endo, K., Matsuda, H., & Mikuni, M. (2007). HPA axis dysfunction in unmedicated major depressive disorder and its normalization by pharmacotherapy correlates with alteration of neural activity in prefrontal cortex and limbic/paralimbic regions. Psychiatry Research, 155(3), 245256. https://doi.org/10.1016/j.pscychresns.2006.11.002 CrossRefGoogle ScholarPubMed
Andruff, H., Carraro, N., Thompson, A., Gaudreau, P., & Louvet, B. (2009). Latent class growth modelling: A tutorial. Tutorials in Quantitative Methods for Psychology, 5(1), 1124. https://doi.org/10.20982/tqmp.05.1.p011 CrossRefGoogle Scholar
Azibo, D. A. (1988). Understanding the proper and improper usage of the comparative research framework. Journal of Black Psychology, 15(1), 8191. https://doi.org/10.1177/00957984880151010 CrossRefGoogle Scholar
Başgöze, Z., Mirza, S. A., Silamongkol, T., Hill, D., Falke, C., Thai, M., Westlund Schreiner, M., Parenteau, A. M., Roediger, D. J., Hendrickson, T. J., Mueller, B. A., Fiecas, M. B., Klimes-Dougan, B., & Cullen, K. R. (2021). Multimodal assessment of sustained threat in adolescents with nonsuicidal self-injury. Development and Psychopathology, 33(5), 17741792. https://doi.org/10.1017/S0954579421000754 CrossRefGoogle ScholarPubMed
Başgöze, Z., Wiglesworth, A., Carosella, K. A., Klimes-Dougan, B., & Cullen, K. R. (2021). Depression, non-suicidal self-injury, and suicidality in adolescents: Common and distinct precursors, correlates, and outcomes. Journal of Psychiatry and Brain Science, 6(5), e210018. https://doi.org/10.20900/jpbs.20210018 Google ScholarPubMed
Bates, D., Mächler, M., Bolker, B., & Walker, S. (2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67(1), 148. https://doi.org/10.18637/jss.v067.i01 CrossRefGoogle Scholar
Beck, A. T., Kovacs, M., & Weissman, A. (1979). Assessment of suicidal intention: The scale for suicide ideation. Journal of Consulting and Clinical Psychology, 47(2), 343352. https://doi.org/10.1037/0022-006X.47.2.343 CrossRefGoogle ScholarPubMed
Beck, A. T., Steer, R. A., Ball, R., & Ranieri, W. F. (1996). Comparison of beck depression inventories-ia and-ii in psychiatric outpatients. Journal of Personality Assessment, 67(3), 588597. https://doi.org/10.1207/s15327752jpa6703_13 CrossRefGoogle ScholarPubMed
Beck, A. T., Steer, R. A., & Brown, G. (1996). Beck depression inventory-II. Psychological Assessment. https://doi.org/10.1037/t00742-000 Google Scholar
Bendezú, J. J., Calhoun, C. D., Patterson, M. W., Findley, A., Rudolph, K. D., Hastings, P., Nock, M. K., & Prinstein, M. J. (2022). Adolescent girls’ stress responses as prospective predictors of self-injurious thoughts and behaviors: A person-centered, multilevel study. Development and Psychopathology, 34(4), 14471467. https://doi.org/10.1017/S0954579420002229 CrossRefGoogle ScholarPubMed
Bendezú, J. J., Calhoun, C. D., Vinograd, M., Patterson, M. W., Rudolph, K. D., Giletta, M., Hastings, P., Nock, M. K., Slavich, G. M., & Prinstein, M. J. (2022). Exploring joint HPA-inflammatory stress response profiles in adolescent girls: Implications for developmental models of neuroendocrine dysregulation. Developmental Psychobiology, 64(3), e22247. https://doi.org/10.1002/dev.22247 CrossRefGoogle ScholarPubMed
Bendezú, J. J., Thai, M., Wiglesworth, A., Cullen, K. R., & Klimes-Dougan, B. (2022). Adolescent stress experience-expression–physiology correspondence: Links to depression, self-injurious thoughts and behaviors, and frontolimbic neural circuity. Journal of Affective Disorders, 300, 269279. https://doi.org/10.1016/j.jad.2021.12.098 CrossRefGoogle ScholarPubMed
Bendezú, J. J., & Wadsworth, M. E. (2023). Intervening to optimize HPA-sAM co-activation among youth exposed to poverty-related stress: A person-centered, multilevel approach. Biological Psychiatry, 93(9), S61. https://doi.org/10.1016/j.biopsych.2023.02.167 CrossRefGoogle Scholar
Bentley, K. H., Nock, M. K., & Barlow, D. H. (2014). The four-function model of nonsuicidal self-injury: Key directions for future research. Clinical Psychological Science, 2(5), 638656. https://doi.org/10.1177/2167702613514563 CrossRefGoogle Scholar
Bhagwagar, Z., Hafizi, S., & Cowen, P. J. (2003). Increase in concentration of waking salivary cortisol in recovered patients with depression. American Journal of Psychiatry, 160(10), 18901891. https://doi.org/10.1176/appi.ajp.160.10.1890 CrossRefGoogle ScholarPubMed
Buss, K. A., Jaffee, S., Wadsworth, M. E., & Kliewer, W. (2018). Impact of psychophysiological stress-response systems on psychological development: Moving beyond the single biomarker approach. Developmental Psychology, 54(9), 16011605. https://doi.org/10.1037/dev0000596 CrossRefGoogle ScholarPubMed
Carosella, K. A., Mirza, S., Başgöze, Z., Cullen, K. R., & Klimes-Dougan, B. (2023). Adolescent non-suicidal self-injury during the COVID-19 pandemic: A prospective longitudinal study of biological predictors of maladaptive emotion regulation. Psychoneuroendocrinology, 151, 106056. https://doi.org/10.1016/j.psyneuen.2023.106056 CrossRefGoogle ScholarPubMed
Carosella, K. A., Wiglesworth, A., Bendezú, J. J., Brower, R., Mirza, S. A., Mueller, B. A., & Klimes-Dougan, B. (Accepted for Publication). Patterns of experience, expression, and physiology of stress relate to depressive symptoms and self-injurious thoughts and behaviors in adolescents: A person-centered approach. Psychological Medicine.Google Scholar
Carosella, K. A., Wiglesworth, A., Silamongkol, T., Tavares, N., Falke, C. A., Fiecas, M. B., Cullen, K. R., & Klimes-Dougan, B. (2021). Non-suicidal self-injury in the context of COVID-19: The importance of psychosocial factors for female adolescents. Journal of Affective Disorders Reports, 4, 100137. https://doi.org/10.1016/j.jadr.2021.100137 CrossRefGoogle Scholar
Castellví, P., Lucas-Romero, E., Miranda-Mendizábal, A., Parés-Badell, O., Almenara, J., Alonso, I., Blasco, M. J., Cebrià, A., Gabilondo, A., Gili, M., Lagares, C., Piqueras, J. A., Roca, M., Rodríguez-Marín, J., Rodríguez-Jimenez, T., Soto-Sanz, V., & Alonso, J. (2017). Longitudinal association between self-injurious thoughts and behaviors and suicidal behavior in adolescents and young adults: A systematic review with meta-analysis. Journal of Affective Disorders, 215, 3748. https://doi.org/10.1016/j.jad.2017.03.035 CrossRefGoogle Scholar
Chen, F. R., Raine, A., & Granger, D. A. (2020). The within-person coordination of HPA and ANS activity in stress response: Relation with behavior problems. Psychoneuroendocrinology, 121, 104805. https://doi.org/10.1016/j.psyneuen.2020.104805 CrossRefGoogle ScholarPubMed
Cicchetti, Dante (2010). Resilience under conditions of extreme stress: A multilevel perspective. World Psychiatry, 9(3), 145154 CrossRefGoogle Scholar
Cicchetti, Dante (2013). Annual research review: Resilient functioning in maltreated children--past, present, and future perspectives. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 54(4), 402422. https://doi.org/10.1111/j.1469-7610.2012.02608.x CrossRefGoogle ScholarPubMed
Cicchetti, Dante, & Dawson, G. (2002). Editorial: Multiple levels of analysis. Development and Psychopathology, 14(3), 417420. https://doi.org/10.1017/S0954579402003012 CrossRefGoogle ScholarPubMed
Cicchetti, Dante, & Rogosch, F. A. (1996). Equifinality and multifinality in developmental psychopathology. Development and Psychopathology, 8(4), 597600. https://doi.org/10.1017/S0954579400007318 CrossRefGoogle Scholar
Clark, L. A., & Watson, D. (1994). The PANAS-x: Manual for the positive and negative affect schedule - expanded form. University of Iowa. https://doi.org/10.17077/48vt-m4t2 Google Scholar
Cohn, M. A., Fredrickson, B. L., Brown, S. L., Mikels, J. A., & Conway, A. M. (2009). Happiness unpacked: Positive emotions increase life satisfaction by building resilience. Emotion, 9(3), 361368. https://doi.org/10.1037/a0015952 CrossRefGoogle ScholarPubMed
Craske, M. G., Meuret, A. E., Ritz, T., Treanor, M., & Dour, H. J. (2016). Treatment for anhedonia: A neuroscience driven approach. Depression and Anxiety, 33(10), 927938. https://doi.org/10.1002/da.22490 CrossRefGoogle ScholarPubMed
Curry, J., Silva, S., Rohde, P., Ginsburg, G., Kratochvil, C., Simons, A., Kirchner, J., May, D., Kennard, B., Mayes, T., Feeny, N., Albano, A. M., Lavanier, S., Reinecke, M., Jacobs, R., Becker-Weidman, E., Weller, E., Emslie, G., Walkup, J., Kastelic, E., … March, J. (2011). Recovery and recurrence following treatment for adolescent major depression. Archives of General Psychiatry, 68(3), 263269. https://doi.org/10.1001/archgenpsychiatry.2010.150 CrossRefGoogle ScholarPubMed
Curtis, W. J., & Cicchetti, D. (2003). Moving research on resilience into the 21st century: Theoretical and methodological considerations in examining the biological contributors to resilience. Development and Psychopathology, 15(3), 773810. https://doi.org/10.1017/s0954579403000373 CrossRefGoogle ScholarPubMed
Daly, M. (2022). Prevalence of depression among adolescents in the U.S. from 2009 to 2019: Analysis of trends by sex, race/ethnicity, and income. Journal of Adolescent Health, 70(3), 496499. https://doi.org/10.1016/j.jadohealth.2021.08.026 CrossRefGoogle Scholar
Davies, J., Sullivan, S., & Zammit, S. (2018). Adverse life outcomes associated with adolescent psychotic experiences and depressive symptoms. Social Psychiatry and Psychiatric Epidemiology, 53(5), 497507. https://doi.org/10.1007/s00127-018-1496-z CrossRefGoogle ScholarPubMed
de Beurs, D. P., Fokkema, M., de Groot, M. H., de Keijser, J., & Kerkhof, A. J. F. M. (2015). Longitudinal measurement invariance of the beck scale for suicide ideation. Psychiatry Research, 225(3), 368373. https://doi.org/10.1016/j.psychres.2014.11.075 CrossRefGoogle ScholarPubMed
de Vries, L. P., Baselmans, B. M. L., Luykx, J. J., de Zeeuw, E. L., Minică, C. C., de Geus, E. J. C., Vinkers, C. H., & Bartels, M. (2021). Genetic evidence for a large overlap and potential bidirectional causal effects between resilience and well-being. Neurobiology of Stress, 14, 100315. https://doi.org/10.1016/j.ynstr.2021.100315 CrossRefGoogle ScholarPubMed
Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The satisfaction with life scale. Journal of Personality Assessment, 49(1), 7175. https://doi.org/10.1207/s15327752jpa4901_13 CrossRefGoogle ScholarPubMed
Dozier, M., Roben, C. K. P., Caron, E., Hoye, J., & Bernard, K. (2018). Attachment and biobehavioral catch-up: An evidence-based intervention for vulnerable infants and their families. Psychotherapy Research: Journal of the Society for Psychotherapy Research, 28(1), 1829. https://doi.org/10.1080/10503307.2016.1229873 CrossRefGoogle ScholarPubMed
Dressendörfer, R. A., Kirschbaum, C., Rohde, W., Stahl, F., & Strasburger, C. J. (1992). Synthesis of a cortisol-biotin conjugate and evaluation as a tracer in an immunoassay for salivary cortisol measurement. The Journal of Steroid Biochemistry and Molecular Biology, 43(7), 683692. https://doi.org/10.1016/0960-0760(92)90294-s CrossRefGoogle Scholar
Ellis, B. J., & Giudice, M. D. (2014). Beyond allostatic load: Rethinking the role of stress in regulating human development. Development and Psychopathology, 26(1), 120. https://doi.org/10.1017/S0954579413000849 CrossRefGoogle ScholarPubMed
Esposito-Smythers, C., Goldstein, T., Birmaher, B., Goldstein, B., Hunt, J., Ryan, N., Axelson, D., Strober, M., Gill, M. K., Hanley, A., & Keller, M. (2010). Clinical and psychosocial correlates of non-suicidal self-injury within a sample of children and adolescents with bipolar disorder. Journal of Affective Disorders, 125(1), 8997. https://doi.org/10.1016/j.jad.2009.12.029 CrossRefGoogle ScholarPubMed
Evans, G., Chen, E., Miller, G. E., & Seeman, T. E. (2012). How poverty gets under the skin: A life course perspective. The Oxford Handbook of Poverty and Child Development, 1, 1336. https://doi.org/10.1093/oxfordhb/9780199769100.013.0001 CrossRefGoogle Scholar
Fergus, S., & Zimmerman, M. A. (2005). Adolescent resilience: A framework for understanding healthy development in the face of risk. Annu. Rev. Public Health, 26(1), 399419.CrossRefGoogle ScholarPubMed
Fisher, P. A., Beauchamp, K. G., Roos, L. E., Noll, L. K., Flannery, J., & Delker, B. C. (2016). The neurobiology of intervention and prevention in early adversity. Annual Review of Clinical Psychology, 12(1), 331357. https://doi.org/10.1146/annurev-clinpsy-032814-112855 CrossRefGoogle ScholarPubMed
Fisher, P. A., Van Ryzin, M. J., & Gunnar, M. R. (2011). Mitigating HPA axis dysregulation associated with placement changes in foster care. Psychoneuroendocrinology, 36(4), 531539. https://doi.org/10.1016/j.psyneuen.2010.08.007 CrossRefGoogle ScholarPubMed
Godoy, L. D., Rossignoli, M. T., Delfino-Pereira, P., Garcia-Cairasco, N., & de Lima Umeoka, E. H. (2018). A comprehensive overview on stress neurobiology: Basic concepts and clinical implications. Frontiers in Behavioral Neuroscience, 12, 127.CrossRefGoogle ScholarPubMed
Granger, D. A., Hibel, L. C., Fortunato, C. K., & Kapelewski, C. H. (2009). Medication effects on salivary cortisol: Tactics and strategy to minimize impact in behavioral and developmental science. Psychoneuroendocrinology, 34(10), 14371448. https://doi.org/10.1016/j.psyneuen.2009.06.017 CrossRefGoogle ScholarPubMed
Gunnar, M. R., & Donzella, B. (2002). Social regulation of the cortisol levels in early human development. Psychoneuroendocrinology, 27(1), 199220. https://doi.org/10.1016/S0306-4530(01)00045-2 CrossRefGoogle ScholarPubMed
Hankin, B. L. (2012). Future directions in vulnerability to depression among youth: Integrating risk factors and processes across multiple levels of analysis. Journal of Clinical Child and Adolescent Psychology: The Official Journal for the Society of Clinical Child and Adolescent Psychology, American Psychological Association, Division, 41(5), 695718. https://doi.org/10.1080/15374416.2012.711708 CrossRefGoogle ScholarPubMed
Harter, S. (1988). Self-perception profile for adolescents. Gifted Child Quarterly.Google Scholar
Hatfield, B. E., & Williford, A. P. (2017). Cortisol patterns for young children displaying disruptive behavior: Links to a teacher-child, relationship-focused intervention. Prevention Science, 18(1), 4049. https://doi.org/10.1007/s11121-016-0693-9 CrossRefGoogle ScholarPubMed
Hatzenbuehler, M. L., & McLaughlin, K. A. (2014). Structural stigma and hypothalamic-pituitary-adrenocortical axis reactivity in lesbian, gay, and bisexual young adults. Annals of Behavioral Medicine : A Publication of the Society of Behavioral Medicine, 47(1), 3947. https://doi.org/10.1007/s12160-013-9556-9 CrossRefGoogle ScholarPubMed
Helgeson, V. S., Snyder, P., & Seltman, H. (2004). Psychological and physical adjustment to breast cancer over 4 years: Identifying distinct trajectories of change. Health Psychology: Official Journal of the Division of Health Psychology, American Psychological Association, 23(1), 315. https://doi.org/10.1037/0278-6133.23.1.3 CrossRefGoogle ScholarPubMed
Herman, J. P., Ostrander, M. M., Mueller, N. K., & Figueiredo, H. (2005). Limbic system mechanisms of stress regulation: Hypothalamo-pituitary-adrenocortical axis. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 29(8), 12011213. https://doi.org/10.1016/j.pnpbp.2005.08.006 CrossRefGoogle ScholarPubMed
Hosseini-Kamkar, N., Lowe, C., & Morton, J. B. (2021). The differential calibration of the HPA axis as a function of trauma versus adversity: A systematic review and p-curve meta-analyses. Neuroscience & Biobehavioral Reviews, 127, 54135. https://doi.org/10.1016/j.neubiorev.2021.04.006 CrossRefGoogle ScholarPubMed
Institute, S. A. S. (2010). STAT-sAS, version 9.4. SAS Institute Cary, North Carolina.Google Scholar
Iob, E., Kirschbaum, C., & Steptoe, A. (2020). Persistent depressive symptoms, HPA-axis hyperactivity, and inflammation: The role of cognitive-affective and somatic symptoms. Molecular Psychiatry, 25(5), 11301140. https://doi.org/10.1038/s41380-019-0501-6 CrossRefGoogle ScholarPubMed
Johnson, D., Dupuis, G., Piche, J., Clayborne, Z., & Colman, I. (2018). Adult mental health outcomes of adolescent depression: A systematic review. Depression and Anxiety, 35(8), 700716. https://doi.org/10.1002/da.22777 CrossRefGoogle ScholarPubMed
Jones, B. L., & Nagin, D. S. (2007). Advances in group-based trajectory modeling and an SAS procedure for estimating them. Sociological Methods & Research, 35(4), 542571. https://doi.org/10.1177/0049124106292364 CrossRefGoogle Scholar
Juruena, M. F., Bocharova, M., Agustini, B., & Young, A. H. (2018). Atypical depression and non-atypical depression: Is HPA axis function a biomarker? A systematic review. Journal of Affective Disorders, 233, 4567. https://doi.org/10.1016/j.jad.2017.09.052 CrossRefGoogle ScholarPubMed
Kaess, M., Hille, M., Parzer, P., Maser-Gluth, C., Resch, F., & Brunner, R. (2012). Alterations in the neuroendocrinological stress response to acute psychosocial stress in adolescents engaging in nonsuicidal self-injury. Psychoneuroendocrinology, 37(1), 157161. https://doi.org/10.1016/j.psyneuen.2011.05.009 CrossRefGoogle ScholarPubMed
Kaufman, J., Birmaher, B., Brent, D., Rao, U., Flynn, C., Moreci, P., Williamson, D., & Ryan, N. (1997). Schedule for affective disorders and schizophrenia for school-age children-present and lifetime version (K-sADS-pL): Initial reliability and validity data. Journal of the American Academy of Child & Adolescent Psychiatry, 36(7), 980988. https://doi.org/10.1097/00004583-199707000-00021 CrossRefGoogle ScholarPubMed
Kirschbaum, C., Pirke, K.-M., & Hellhammer, D. H. (1993). The trier social stress test – a tool for investigating psychobiological stress responses in a laboratory setting. Neuropsychobiology, 28(1-2), 7681. https://doi.org/10.1159/000119004 CrossRefGoogle Scholar
Klainin-Yobas, P., Vongsirimas, N., Ramirez, D. Q., Sarmiento, J., & Fernandez, Z. (2021). Evaluating the relationships among stress, resilience and psychological well-being among young adults: A structural equation modelling approach. BMC Nursing, 20(1), 119. https://doi.org/10.1186/s12912-021-00645-9 CrossRefGoogle ScholarPubMed
Klimes-Dougan, B., Begnel, E., Almy, B., Thai, M., Schreiner, M. W., & Cullen, K. R. (2019). Hypothalamic-pituitary-adrenal axis dysregulation in depressed adolescents with non-suicidal self-injury. Psychoneuroendocrinology, 102, 216224. https://doi.org/10.1016/j.psyneuen.2018.11.004 CrossRefGoogle ScholarPubMed
Klimes-Dougan, B., Eberly, L. E., Westlund Schreiner, M., Kurkiewicz, P., Houri, A., Schlesinger, A., Thomas, K. M., Mueller, B. A., Lim, K. O., & Cullen, K. R. (2014). Multilevel assessment of the neurobiological threat system in depressed adolescents: Interplay between the limbic system and hypothalamic-pituitary–adrenal axis. Development and Psychopathology, 26(4pt2), 13211335. https://doi.org/10.1017/S0954579414001059 CrossRefGoogle ScholarPubMed
Kroupina, M. G., Fuglestad, A. J., Iverson, S. L., Himes, J. H., Mason, P. W., Gunnar, M. R., Miller, B. S., Petryk, A., & Johnson, D. E. (2012). Adoption as an intervention for institutionally reared children: HPA functioning and developmental status. Infant Behavior and Development, 35(4), 829837. https://doi.org/10.1016/j.infbeh.2012.07.011 CrossRefGoogle ScholarPubMed
Lange, C., Zschucke, E., Ising, M., Uhr, M., Bermpohl, F., & Adli, M. (2013). Evidence for a normal HPA axis response to psychosocial stress in patients remitted from depression. Psychoneuroendocrinology, 38(11), 27292736. https://doi.org/10.1016/j.psyneuen.2013.06.033 CrossRefGoogle ScholarPubMed
Leistner, C., & Menke, A. (2020). Chapter 4 - hypothalamic-pituitary–adrenal axis and stress. In Lanzenberger, R., Kranz, G. S., & Savic, I. (Eds.), Handbook of clinical neurology (vol. vol. 175, p. 5564). Elsevier, Retrieved 20 November 2022 from (https://doi.org/10.1016/B978-0-444-64123-6.00004-7 Google Scholar
Lenth, R. (2023). Emmeans: Estimated marginal means, aka least-squares means, R package version 1.8.4-1. https://CRAN.R-project.org/package=emmeans Google Scholar
Lett, E., Asabor, E., Beltrán, S., Cannon, A. M., & Arah, O. A. (2022). Conceptualizing, contextualizing, and operationalizing race in quantitative health sciences research. The Annals of Family Medicine, 20(2), 157163. https://doi.org/10.1370/afm.2792 CrossRefGoogle ScholarPubMed
Li, Z.-S., & Hasson, F. (2020). Resilience, stress, and psychological well-being in nursing students: A systematic review. Nurse Education Today, 90, 104440. https://doi.org/10.1016/j.nedt.2020.104440 CrossRefGoogle ScholarPubMed
Linton, P. J., & Dorshkind, K. (2004). Age-related changes in lymphocyte development and function. Nature Immunology, 5(2), 133139. https://doi.org/10.1038/ni1033 CrossRefGoogle ScholarPubMed
Little, R. J. A. (1988). A test of missing completely at random for multivariate data with missing values. Journal of the American Statistical Association, 83(404), 11981202. https://doi.org/10.1080/01621459.1988.10478722 CrossRefGoogle Scholar
Liu, Y., Wang, Z.-H., & Li, Z.-G. (2012). Affective mediators of the influence of neuroticism and resilience on life satisfaction. Personality and Individual Differences, 52(7), 833838. https://doi.org/10.1016/j.paid.2012.01.017 CrossRefGoogle Scholar
Lloyd-Richardson, E. E., Perrine, N., Dierker, L., & Kelley, M. L. (2007). Characteristics and functions of non-suicidal self-injury in a community sample of adolescents. Psychological Medicine, 37(8), 11831192. https://doi.org/10.1017/S003329170700027X CrossRefGoogle Scholar
Lopez-Duran, N. L., Kovacs, M., & George, C. J. (2009). Hypothalamic-pituitary-adrenal axis dysregulation in depressed children and adolescents: A meta-analysis. Psychoneuroendocrinology, 34(9), 12721283. https://doi.org/10.1016/j.psyneuen.2009.03.016 CrossRefGoogle ScholarPubMed
Louvet, B., Gaudreau, P., Menaut, A., Genty, J., & Deneuve, P. (2009). Revisiting the changing and stable properties of coping utilization using latent class growth analysis: A longitudinal investigation with soccer referees. Psychology of Sport and Exercise, 10(1), 124135. https://doi.org/10.1016/j.psychsport.2008.02.002 CrossRefGoogle Scholar
Lupien, S. J., King, S., Meaney, M. J., & McEwen, B. S. (2000). Child’s stress hormone levels correlate with mother’s socioeconomic status and depressive state. Biological Psychiatry, 48(10), 976980. https://doi.org/10.1016/s0006-3223(00)00965-3 CrossRefGoogle ScholarPubMed
Luthar, S. S., Cicchetti, D., & Becker, B. (2000). The construct of resilience: A critical evaluation and guidelines for future work. Child Development, 71(3), 543562. https://doi.org/10.1111/1467-8624.00164 CrossRefGoogle ScholarPubMed
Masten, A. S., & Barnes, A. J. (2018). Resilience in children: Developmental perspectives. Children, 5(7), 98. https://doi.org/10.3390/children5070098 CrossRefGoogle ScholarPubMed
Masten, A. S., Best, K. M., & Garmezy, N. (1990). Resilience and development: Contributions from the study of children who overcome adversity. Development and Psychopathology, 2(4), 425444. https://doi.org/10.1017/S0954579400005812 CrossRefGoogle Scholar
Masten, A. S., Lucke, C. M., Nelson, K. M., & Stallworthy, I. C. (2021). Resilience in development and psychopathology: Multisystem perspectives. Annual Review of Clinical Psychology, 17(1), 521549. https://doi.org/10.1146/annurev-clinpsy-081219-120307 CrossRefGoogle ScholarPubMed
Mauss, I. B., & Robinson, M. D. (2009). Measures of emotion: A review. Cognition and Emotion, 23(2), 209237. https://doi.org/10.1080/02699930802204677 CrossRefGoogle ScholarPubMed
Mazurka, R., Wynne-Edwards, K. E., & Harkness, K. L. (2018). Sex differences in the cortisol response to the trier social stress test in depressed and nondepressed adolescents. Clinical Psychological Science, 6(3), 301314. https://doi.org/10.1177/2167702617739973 CrossRefGoogle Scholar
McLaughlin, K. A., Weissman, D., & Bitrán, D. (2019). Childhood adversity and neural development: A systematic review. Annual Review of Developmental Psychology, 1(1), 277312. https://doi.org/10.1146/annurev-devpsych-121318-084950 CrossRefGoogle ScholarPubMed
Muehlenkamp, J. J., Brausch, A. M., & Washburn, J. J. (2017). How much is enough? Examining frequency criteria for NSSI disorder in adolescent inpatients. Journal of Consulting and Clinical Psychology, 85(6), 611619. https://doi.org/10.1037/ccp0000209 CrossRefGoogle ScholarPubMed
Nagin, D. S. (2005). Group-based modeling of development. Harvard University Press.CrossRefGoogle Scholar
Nagin, D. S., Jones, B. L., Passos, V. L., & Tremblay, R. E. (2018). Group-based multi-trajectory modeling. Statistical Methods in Medical Research, 27(7), 20152023. https://doi.org/10.1177/0962280216673085 CrossRefGoogle ScholarPubMed
Needham, B. L., Ali, T., Allgood, K. L., Ro, A., Hirschtick, J. L., & Fleischer, N. L. (2022). Institutional racism and health: A framework for conceptualization, measurement, and analysis. Journal of Racial and Ethnic Health Disparities. Advance online publication. https://doi.org/10.1007/s40615-022-01381-9 Google Scholar
Nock, M. K., Borges, G., Bromet, E. J., Alonso, J., Angermeyer, M., Beautrais, A., Bruffaerts, R., Chiu, W. T., de Girolamo, G., Gluzman, S., de Graaf, R., Gureje, O., Haro, J. M., Huang, Y., Karam, E., Kessler, R. C., Lepine, J. P., Levinson, D., Medina-Mora, M. E., Ono, Y., … Williams, D. (2008). Cross-national prevalence and risk factors for suicidal ideation, plans and attempts. The British Journal of Psychiatry, 192(2), 98105. https://doi.org/10.1192/bjp.bp.107.040113 CrossRefGoogle ScholarPubMed
Nock, M. K., Green, J. G., Hwang, I., McLaughlin, K. A., Sampson, N. A., Zaslavsky, A. M., & Kessler, R. C. (2013). Prevalence, correlates, and treatment of lifetime suicidal behavior among adolescents: Results from the national comorbidity survey replication adolescent supplement. JAMA Psychiatry, 70(3), 300310. https://doi.org/10.1001/2013.jamapsychiatry.55 CrossRefGoogle ScholarPubMed
Nock, M. K., Holmberg, E. B., Photos, V. I., & Michel, B. D. (2007). Self-injurious thoughts and behaviors interview: Development, reliability, and validity in an adolescent sample. Psychological Assessment, 19(3), 309317. https://doi.org/10.1037/1040-3590.19.3.309 CrossRefGoogle Scholar
Pham, H. T., Bendezú, J. J., & Wadsworth, M. E. (2023). HPA-sAM co-activation among racially diverse, economically disadvantaged early adolescents: Secondary analysis with a preliminary test of a multisystem, person-centered approach. Biological Psychology, 179, 108546. https://doi.org/10.1016/j.biopsycho.2023.108546 CrossRefGoogle ScholarPubMed
Plener, P. L., Schumacher, T. S., Munz, L. M., & Groschwitz, R. C. (2015). The longitudinal course of non-suicidal self-injury and deliberate self-harm: A systematic review of the literature. Borderline Personality Disorder and Emotion Dysregulation, 2(1), 2. https://doi.org/10.1186/s40479-014-0024-3 CrossRefGoogle ScholarPubMed
R Core Team (2022). R: A language and environment for statistical computing, https://www.r-project.org/ Google Scholar
Rajhvajn Bulat, L., Sušac, N., & Ajduković, M. (2023). Predicting prolonged non-suicidal self-injury behaviour and suicidal ideations in adolescence – the role of personal and environmental factors. Current Psychology, 112. Advance online publication. https://doi.org/10.1007/s12144-023-04404-8 Google ScholarPubMed
Rao, R., & Androulakis, I. P. (2019). Allostatic adaptation and personalized physiological trade-offs in the circadian regulation of the HPA axis: A mathematical modeling approach. Scientific Reports, 9(1), 11212. https://doi.org/10.1038/s41598-019-47605-7 CrossRefGoogle Scholar
Reigstad, K. M., Gunlicks-Stoessel, M. L., Westervelt, A., & Gearity, A. R. (2022). Healthy emotions and relationships with teens—a guide for parents: An intervention development study of a parent-based intervention for adolescents with depression. Clinical Child Psychology and Psychiatry, 27(3), 586597. https://doi.org/10.1177/13591045221078416 CrossRefGoogle ScholarPubMed
Ribeiro, J. D., Franklin, J. C., Fox, K. R., Bentley, K. H., Kleiman, E. M., Chang, B. P., & Nock, M. K. (2016). Self-injurious thoughts and behaviors as risk factors for future suicide ideation, attempts, and death: A meta-analysis of longitudinal studies. Psychological Medicine, 46(2), 225236. https://doi.org/10.1017/S0033291715001804 CrossRefGoogle Scholar
Ruggeri, K., Garcia-Garzon, E., Maguire, Á., Matz, S., & Huppert, F. A. (2020). Well-being is more than happiness and life satisfaction: A multidimensional analysis of 21 countries. Health and Quality of Life Outcomes, 18(1), 192. https://doi.org/10.1186/s12955-020-01423-y CrossRefGoogle ScholarPubMed
Schär, S., Mürner-Lavanchy, I., Schmidt, S. J., Koenig, J., & Kaess, M. (2022). Child maltreatment and hypothalamic-pituitary-adrenal axis functioning: A systematic review and meta-analysis. Frontiers in Neuroendocrinology, 66, 100987. https://doi.org/10.1016/j.yfrne.2022.100987 CrossRefGoogle ScholarPubMed
Seddon, J. A., Rodriguez, V. J., Provencher, Y., Raftery-Helmer, J., Hersh, J., Labelle, P. R., & Thomassin, K. (2020). Meta-analysis of the effectiveness of the trier social stress test in eliciting physiological stress responses in children and adolescents. Psychoneuroendocrinology, 116, 104582. https://doi.org/10.1016/j.psyneuen.2020.104582 CrossRefGoogle ScholarPubMed
Swannell, S. V., Martin, G. E., Page, A., Hasking, P., & St John, N. J. (2014). Prevalence of nonsuicidal self-injury in nonclinical samples: Systematic review, meta-analysis and meta-regression. Suicide and Life-Threatening Behavior, 44(3), 273303. https://doi.org/10.1111/sltb.12070 CrossRefGoogle ScholarPubMed
Taborsky, B., English, S., Fawcett, T. W., Kuijper, B., Leimar, O., McNamara, J. M., Ruuskanen, S., & Sandi, C. (2021). Towards an evolutionary theory of stress responses. Trends in Ecology & Evolution, 36(1), 3948. https://doi.org/10.1016/j.tree.2020.09.003 CrossRefGoogle ScholarPubMed
Von Eye, A., & Bergman, L. R. (2003). Research strategies in developmental psychopathology: Dimensional identity and the person-oriented approach. Development and Psychopathology, 15(3), 553580. https://doi.org/10.1017/S0954579403000294 CrossRefGoogle ScholarPubMed
Weissman, M. M., Wolk, S., Goldstein, R. B., Moreau, D., Adams, P., Greenwald, S., Klier, C. M., Ryan, N. D., Dahl, R. E., & Wickramaratne, P. (1999). Depressed adolescents grown up. JAMA, 281(18), 17071713. https://doi.org/10.1001/jama.281.18.1707 CrossRefGoogle ScholarPubMed
White, H. V., Silamongkol, T., Wiglesworth, A., Labella, M. H., Goetz, E. R., Cullen, K. R., & Klimes-Dougan, B. (2021). Maternal emotion socialization of adolescent girls engaging in non-suicidal self-injury. Research on Child and Adolescent Psychopathology, 49(5), 683695. https://doi.org/10.1007/s10802-020-00758-w CrossRefGoogle ScholarPubMed
Wickham, H. (2016). ggplot2: Elegant graphics for data analysis. Springer. https://link.springer.com/book/10.1007/978-0-387-98141-3 CrossRefGoogle Scholar
Wright, M. O., & Masten, A. S. (2005). Resilience processes in development. In Goldstein, S., & Brooks, R. B. (Eds.), Handbook of resilience in children (pp. 1737). Boston, MA: Springer US, https://doi.org/10.1007/0-306-48572-9_2 CrossRefGoogle Scholar
Yates, T. M., Tyrell, F. A., & Masten, A. S. (2015). Resilience theory and the practice of positive psychology from individuals to societies. In Positive psychology in practice (pp. 773788). John Wiley & Sons, Ltd, https://doi.org/10.1002/9781118996874.ch44 CrossRefGoogle Scholar
Yim, I. S., Quas, J. A., Cahill, L., & Hayakawa, C. M. (2010). Children’s and adults’ salivary cortisol responses to an identical psychosocial laboratory stressor. Psychoneuroendocrinology, 35(2), 241248. https://doi.org/10.1016/j.psyneuen.2009.06.014 CrossRefGoogle Scholar
Yuuki, N., Ida, I., Oshima, A., Kumano, H., Takahashi, K., Fukuda, M., Oriuchi, N., Endo, K., Matsuda, H., & Mikuni, M. (2005). HPA axis normalization, estimated by DEX/CRH test, but less alteration on cerebral glucose metabolism in depressed patients receiving ECT after medication treatment failures. Acta Psychiatrica Scandinavica, 112(4), 257265. https://doi.org/10.1111/j.1600-0447.2005.00625.x CrossRefGoogle ScholarPubMed
Figure 0

Table 1. Descriptive statistics of the study sample

Figure 1

Figure 1. Profiles indicated in the final four-group solution from the multi-trajectory model. raw scores for experience, expression, and physiology are indicated on the y-axis; a trier social stress test event markers for physiology are estimated based on the time-lag in salivary cortisol presentation; Lexperi–Lexpres–Lphysio = low experience, low expression, low physiology, Hexperi–Hexpres–Hphysio = high experience, high expression, high physiology, Hexperi–Lexpres–Mphysio = high experience, low expression, moderate physiology, Hexperi–Hexpres–Lphysio= high experience, high expression, low physiology.

Figure 2

Figure 2. Longitudinal patterns of reporting of symptoms of psychopathology disaggregated by profile. Lexperi–Lexpres–Lphysio = low experience, low expression, low physiology, Hexperi–Hexpres–Hphysio = high experience, high expression, high physiology, Hexperi–Lexpres–Mphysio = high experience, low expression, moderate physiology, Hexperi–Hexpres–Lphysio= high experience, high expression, low physiology.

Figure 3

Table 2. Linear mixed effect regression model results demonstrating group differences in indices of psychopathology over time based on stress experience, expression, and physiology profile

Figure 4

Figure 3. Longitudinal patterns of reporting of symptoms of well-being disaggregated by profile. Lexperi–Lexpres–Lphysio = low experience, low expression, low physiology, Hexperi–Hexpres–Hphysio = high experience, high expression, high physiology, Hexperi–Lexpres–Mphysio = high experience, low expression, moderate physiology, Hexperi–Hexpres–Lphysio= high experience, high expression, low physiology.

Figure 5

Table 3. Linear mixed effect regression model results demonstrating group differences in indices of well-being over time based on stress experience, expression, and physiology profile

Supplementary material: File

Wiglesworth et al. supplementary material

Tables S1-S16

Download Wiglesworth et al. supplementary material(File)
File 85.1 KB