A2 | Kind Lab

Risk and Resilience Measures Related to Psychopathology in Youth

Lauren K White, Ran Barzilay, Tyler M Moore, Monica E Calkins, Jason D Jones, Megan M Himes, Jami F Young, Ruben C Gurm Raquel E Gur

The online version contains supplementary material available at https://doi.org/10.1007/s10578-021-01296-2

Keywords: Childhood adversity; Psychopathology; Resilience; Risk.

Abstract

Childhood adversity places youth at risk for multiple negative outcomes. The current study aimed to understand how a constellation of risk and resilience factors influenced mental health outcomes as a function of adversities: socioeconomic status (SES) and traumatic stressful events (TSEs). Specifically, we examined outcomes related to psychosis and mood disorders, as well as global clinical functioning. The current study is a longitudinal follow up of 140 participants from the Philadelphia Neurodevelopmental Cohort (PNC) assessed for adversities at Time 1 (Mean age: 14.11 years) and risk, resilience, and clinical outcomes at Time 2 (mean age: 21.54 years). In the context of TSE, a limited set of predictors emerged as important; a more diverse set of moderators emerged in the context of SES. Across adversities, social support was a unique predictor of psychosis spectrum diagnoses and global functioning; emotion dysregulation was an important predictor for mood diagnoses. The current findings underscore the importance of understanding effects of childhood adversity on maladaptive outcomes within a resilience framework.

Introduction

The link between early adversity and poor outcomes, such as psychopathology, is well documented [1]. Understanding the development of psychopathology gave rise to the study of the risk factors that increase an individual’s vulnerability to develop adverse outcomes. Complementary efforts have been intensifying to characterize resilience factors, which protect against the development of poor outcomes or promote positive developmental outcomes [2, 3]. In the fields of resilience science and developmental psychopathology, it is important to understand how multiple factors work in concert to promote positive outcomes or impede negative outcomes after adversity. This endeavor will help parse the heterogeneity observed in response to adversity and can guide intervention and prevention efforts [4–7]. Longitudinal approaches are crucial for understanding the interplay between early life adversity, maladaptive outcomes, and the factors that influence this link [4, 8].

Although resilience research has not used a consistent conceptualization of resilience, it broadly reflects the study of factors or processes that help an individual adapt successfully to life’s adversities [5, 7]; adversities that are often associated with the development of mental health disorders [9, 10]. Often when examining developmental outcomes (e.g., depression) resilience research focuses on a single individual level resilience factor (e.g., self-regulation) or resilience as a unidimensional construct or trait [4, 6]. However, as research increasingly highlights, the importance of understanding resilience from multiple levels – spanning across the individual child, family systems, and broader social and environmental domains – there has been a call for future resilience science to examine resilience factors across these multiple levels [3, 5, 7, 11, 12]. To fully understand the complexity of resilience and understand adaptation to mental health risk, longitudinal studies examining how these multiple levels of protective and risk factors converge to influence developmental outcomes are needed [5, 7, 13].

This endeavor, to assess multiple levels of resilience factors in youth, has proven difficult [7]. For instance, logistic barriers, such as time demands, age differences in a cohort, relying only on parent report, limits the ability to appropriately assess multisystem risk and resilience factors at a single timepoint in a single individual. To help advance resilience science and clinical targets, new methods for studying multilevel processes of risk and resilience are needed [6, 7]. To help address these barriers, a short self-report battery was developed (Risk and Resilience Battery [RRB]) to assess multisystem risk and resilience factors at the level of the individual child, family, and the greater social and neighborhood environments. Specifically, the RRB assess factors of self-reliance, emotion dysregulation, supportive and hostile close relationships, peer victimization, neighborhood danger, and recent personal and family-related stressors [14]. Prior efforts found that this rapid assessment of risk and resilience factors was valid in a wide age range and each of the factors related to concurrent assessments of psychopathology[14, 15]. Recent work has further validated the RRB in a cohort perinatal women [16], showing that during the postpartum period the multiple levels of resilience from the RRB prospectively predict postpartum depression and impaired infant-parent bonding [17]. Thus, brief RRB is likely to help advance resilience science to understand how multiple levels of resilience influence developmental trajectories in the face of adversity.

For moving resilience research and clinical work forward, identifying the specific resilience factors that promote or protect against maladaptive outcomes across different contexts (i.e., types of risk/adversity) is important [5]. That is, the specific outcome (adaptive benefit) of a resilience factor likely depends on the context of the adversity experienced by the individual [7, 18]. Thus, another important component in the study of resilience is the conceptualization and assessment of adversity. Studies of youth, such as the National Comorbidity Survey of Adolescents (ages 13-17), reported that childhood adversity was associated with a host of mental health disorders in adulthood [9, 10]. Indeed, adverse experience in childhood and adolescence are estimated to account for 30% of all mental health disorders [1]. While there are many types of adversity that warrant future examination in this context, the current study focuses on two commonly studied areas of early life stress in the field of developmental psychopathology [19–21]: low socioeconomic status (SES) and traumatic stressful events (TSE e.g., sexual abuse, violence). Low SES typically represents a chronic, often early occurring adversity [22, 23], which has been linked to increased risk for mental health diagnoses, poorer cognitive performance, and marked differences in brain development [19, 24, 25]. Strong links between childhood TSEs and increased psychopathology, poorer cognitive performance, and differences in brain function have also been established [26]. Assaultive or threat-related TSEs, such as sexual assault or a violent physical attack, and multiple childhood TSEs appear to be linked to an even greater risk for negative outcomes [19, 26, 27].

Evidence suggests that different types of childhood adversity are linked to different risk profiles [19, 27–29]; however, prior work often collapses across different types of adversity. Therefore, it is not fully understood how risk and resilience processes differentially influence developmental trajectories in the face of different types of adversity [30]. Given that different types of adversity are linked to different neurobiological and psychological outcomes [27, 29], the type and level of influence of different risk and resilience factors on outcomes is likely different across types of adversity. For instance, in the contexts of poverty, higher levels of self-reliance may be quite adaptive, whereas in the context of early life trauma, supportive close relationships may be most adaptive against mental health risk. Understanding this interplay is an important area of continued pursuit to advance resilience science as it will help advance our understanding of the parameters of risk and guide future intervention efforts.

In the current study, we continue our efforts using the RRB to examine how multisystem risk and resilience factors influence the link between two types of childhood adversity, TSE and low SES, and mental health outcomes, assessed on average seven years earlier. The study sample is a subsample of participants from the Philadelphia Neurodevelopmental Cohort (PNC) that were followed longitudinally and is enhanced for psychosis risk. Data from the PNC, a diverse community sample, allows examination of the relation between childhood adversities, such as low SES and TSEs [19], and a wide range of mental health measures. The current study focuses on the clinical outcomes of mood disorders and psychosis, the latter of which is largely underemphasized in resilience research. Given that resilience work has been criticized by relying too heavily on the presence or absence of clinical symptoms as “positive” or “negative” outcomes [8], the current study also examined measures of daily functioning (as assessed through several clinical functioning ratings). Prior resilience studies are often limited by focusing on a single domain of adversity, or collapsing across types of adversity, or by only examining a single or unidimensional resilience factor. Thus, by examining multisystem risk and resilience factors across different childhood adversities and looking at multiple mental health outcomes, we will be able to 1) help quantify the parameters of risk [5] 2) identify unique and transdiagnostic resilience factors that can help promote adaptive outcomes across different types of adversity [7, 31].

Materials and Methods

Participants

The current sample consists of sub-set of participants (N=140; 58 females) from the community-based PNC (N=9,498) that were recruited to the Lifespan Brain Institute (LiBI) of Penn Medicine (UPenn) and Children’s Hospital of Philadelphia (CHOP) for sequential research assessments as part of several ongoing studies, at which time they also completed the current study procedures. Most of these studies recruited based on clinical risk features, resulting in a sample enriched for psychosis spectrum and mood disorders. See Table 1 for Demographics. The mean time between visits was 7.43 years (SD=0.71), with a range of 5-9 years. The PNC, a community-based cohort, was recruited from the greater Philadelphia region based on a large pool of youth identified and genotyped at CHOP. The full PNC sample spans 8-21 years of age (mean age=14.2 years) with 51.7% females, 32.9% Black, and 55.8% White [32]. Written informed consent was obtained from participants age 18 and older; written assent and parental permission were obtained from children younger than 18 and their parents/legal guardian. Study procedures were approved by the UPenn and CHOP Institutional Review Boards.

Time 1 Assessment

Childhood Adversity (TSE and SES):

TSEs were assessed during the structured GOASSESS clinical interview [32], a computerized version of the Schedule for Affective Disorders and Schizophrenia for School-Age Children– Present and Lifetime Version (K-SADS-PL) [33]. Additional psychosis-related items from the PRIME Screen-Revised [34], and Scale of Prodromal Symptoms from the Structured Interview for Prodromal Syndromes (SIPS)[35], were included. Participants older than 11 years were interviewed individually by a trained clinical coordinator. Collateral reports were also obtained from interviews with caregivers for participants under 18 years of age. Interviews were conducted in the laboratory or at the participant’s home. For the TSE assessment, the interview included probes for lifetime exposure to situations in which the participant: 1) experienced a natural disaster; 2) experienced a bad accident; 3) thought that s/he or someone close to him/her was killed or hurt badly; 4) witnessed someone getting killed, badly beaten, or die; 5) saw a dead body; or was ever her/himself a victim of one of the following assaults: was 6) attacked or badly beaten; 7) threatened with a weapon; or 8) sexually assaulted. TSE assessment was missing for n=3. To quantify significant TSEs, we created a “high risk” group based on participants who endorsed an assaultive traumatic event or had at least two TSEs (n=42) compared to a “low risk” group, those who endorsed 0-1 non-assaultive TSE (n=95). This “high risk” group was created given that assaultive or threat-related TSEs, such as sexual assault or a violent physical attack, and the occurrence of multiple TSEs appear to be linked to an even greater risk for negative outcomes [19, 26, 27]. This grouping is consistent with previous work in the PNC [19].

SES was measured using census-based geocoding of neighborhood-based variables (e.g., percent of residence in poverty, residents married, residence with at least a high school education, median family income) using participant addresses. Neighborhoods were blocked into census groups, generally containing 600-3,000 persons. A composite of the census variables (public in 2010) was created based on the neighborhood-based social-environmental variables [36]. This measure of SES, although related to measures such as parental education, extends beyond individual-level phenotypes of SES and uses multiple indexes of the participant’s surrounding neighborhood [36].

Time 1 baseline clinical symptoms:

Psychopathology (i.e., mood disorders, ADHD, anxiety disorders, OCD, PTSD) was assessed through the GOASSESS clinical interview detailed above [32]. Factor analyses conducted on all interview items [37] revealed several domains of psychopathology of which age-regressed mood and psychosis factors were used as covariates in the current selected analyses.

Time 2 Assessment

Risk and Resilience Assessment (Risk and Resilience Battery (RRB)):

The recently developed 47-item RRB [14] was used to assess risk and resilience factors across individual, family, and broader social and neighborhood environmental domains. This tool is comprised of items from well-established self-report measures (e.g., Network of Relationship Inventory [38]; Difficulties in Emotion Regulation Scale [39]; Multidimensional Peer Victimization Scale,[40]). The original set of scales were chosen by a team of experts (i.e., developmental psychologists, clinical psychologists, and adult and child and adolescent psychiatrists). For the full list of original scale and abbreviated battery items see [14]. From the full list of scales, a combination of Factor Analyses, Item Response Theory, and Computerized Adaptive Testing (CAT) were used to create the brief RRB. The seven risk and resilience factors measured are: self-reliance (3 items, α=.83; e.g., can usually find a way out of difficult situations), peer victimization (14 items, α=.82; e.g., called names or harassed online), emotion dysregulation (5 items, α=.87; e.g., difficulty concentrating or controlling behaviors when upset), supportive close relationships (4 items, α=.86; e.g., lasting relationship and level of care with primary and secondary caregivers), hostile close relationships (5 items, α=.89; e.g., level of arguing with primary and secondary caregivers), neighborhood danger (4 items, α=.84; e.g., lack of perceived level of trust and safety in neighborhood), and recent personal and family-related stressors (12 items, α=.40; e.g., parental divorce, move, family in trouble with the law). During pilot stages, the battery was modified according to participant age [14]; the youngest participants (<16 years) did not complete self-reliance items (n=5). Due to missing data, peer victimization and the two relationship factors were not calculated for n=1. Seven individual factor scores were created by summing all items within a factor and dividing by the total possible points for the participant on that factor. Linear and non-linear age effects were regressed out of each factor.

Clinical Symptoms and Diagnostic Assessment:

Clinical Diagnostic and Functioning Assessment:

At Time 2 participants underwent a similar computer programmed clinical interview as Time 1, but this interview was semi-structured. Modules were based on the KSADS and Structured Interview for Prodromal Symptoms (SIPS version 4.0). The K-SADS modules provided a standardized assessment of DSM-IV axis 1 mood disorders. The SIPS assessed psychosis spectrum symptoms. For participants under the age of 18, collateral clinical interviews were also conducted. After the interview, information was aggregated across proband and collateral reports and medical records (if available), and consensus diagnoses for each participant were made by at least two doctoral level clinicians. For the current analyses, we created Psychosis Spectrum Diagnosis Category (N=72; 51%) that included Schizophrenia (n=7), Psychosis Not Otherwise Specified (NOS, n=6), and those who met criteria for a “Prodromal” psychosis diagnosis [35, 41, 42] (n=59). We also created a Mood Disorder Category (n=41; 29%), including Major Depressive Disorder (MDD) (n=29), Bipolar Diagnoses (n=5), and Mood Disorder NOS (n=7). As part of the clinical consensus, participants also received three ratings to assess global functioning. These ratings were for SIPS Global Assessment of Functioning (GAF; scores rage: 0-100 [42] and the Global Function: Social (GF: Social) and Role (GF: Role) Scales, both with scores ranging from 1-10 [43]. For GAF, ratings reflected functioning over the past year, whereas GF: Social and Role reflected functioning over the past month. Correlations among function scores ranged from r= .44 to .57, ps<.0001. Three participants were missing clinical diagnoses and Global Function: Social/Role scores; six participants did not have GAF ratings.

Self-Report Clinical Symptoms:

At the Time 2 assessment, participants were also asked to complete several self-report clinical scales. Psychosis-related symptoms were assessed using the twelve item PRIME Screen-Revised [34]. The seven item PROMIS Depression Scale (PDS)[44] was used to assess depression symptoms. Psychosis Spectrum and Depression Symptom Scores were created from both scales by summing across scale items. Self-report scales were correlated at r=.37, p<.001.

Data Analysis

Separate hierarchical logistic and linear regressions were conducted to examine 1) the predictive relations between exposure (TSE/SES) and outcome; 2) the joint associations between the risk and resilience factors and outcome; and 3) interactions between adversity and risk and resilience factors on outcomes. Separate models were used for TSE (high vs. low risk groups) and SES (continuous measure). The outcomes of focus were: clinical diagnoses (Psychosis Spectrum and Mood disorders) and global functioning. Results using self-report clinical symptoms are presented in Supplemental Materials. For the regression models, Step 1 included TSE (or SES), sex, age, time between visits, and race (for TSE analyses only). Step 2 entered all seven age regressed risk-resilience factors to examine the unique associations of each risk-resilience measures on outcomes, while controlling for the other risk-resilience factors. Step 3 entered the seven interactions terms between TSE (or SES) and the risk-resilience factors. This step examined how the risk-resilience factors moderate the link between adversity and outcomes. For clinical outcome analyses, baseline symptom covariates were entered in Step 2. Logistic regressions were used to examine the diagnostic outcomes. Linear regressions were used for all continuous outcomes. Although increase in R2 is reported for each regression step, emphasis was given to the individual coefficients for each item and interaction term. Parallel analyses using a continuous measure of TSEs are reported in the supplemental material. To probe significant interactions, main effects models were performed within each adversity groups. For SES, the low SES group were those in the lowest tertile (Low SES) compared to participants in the middle and high SES tertiles (benign SES). SES and race were highly confounded in the current sample (94% of the lowest tertile were Black participants). As such, White participants (n=53) were excluded from the SES analyses, resulting in n= 44 in the low SES group and n=43 in the benign SES group. For illustrative purposes, significant moderations are plotted by adversity groups at high and low levels of the risk-resilience factor (based on median split). All continuous independent variables were standardized (z-scores). Outlying values (±3.5 SD) were rescaled (Winsorized) to the nearest, non-outlying value.

Results

Demographics for participants in the full sample and SES analyses (sub-sample) are presented in Table 1.

Predicting Psychosis Spectrum Disorders

TSE:

Participants in the high-risk TSE group Time 1 were at an increased likelihood of having a psychosis spectrum disorder at Time 2 (Table 2). Self-reliance and supportive close relationship factors were unique predictors of disorder; higher scores on either factor were associated with a decreased probability of having a psychosis spectrum disorder. A significant interaction between TSE group and supportive close relationships emerged; supportive close relationship quality was a protective factor of having a disorder only in the low-risk group (OR=0.4, Wald=7.96, p=.005; high-risk: OR=1.50, Wald=.50, p=.48; Figure 1).

Fig. 1. Joint effects of Traumatic Stressful Events (TSE) and Risk and Resilience Factors on Clinical and Functioning Outcomes

SES:

There was no direct association between Time 1 SES and likelihood of having a psychosis spectrum disorder at Time 2 (Table 2). The only significant unique risk and resilience predictor was supportive close relationships; participants endorsing higher levels of supportive close relationships were less likely to have a psychosis spectrum disorder. No significant interactions emerged.

Predicting Mood Disorders

TSE:

Participants in the high-risk TSE group were not significantly more likely to have a mood disorder at Time 2 (Table 2). Although, as reported in the supplemental materials, a link between TSEs and self-reported depression symptoms at T2 was found. Higher emotion dysregulation was uniquely associated with a higher likelihood of having a mood diagnosis. No significant interactions emerged.

SES:

SES at Time 1 predicted a greater likelihood of having a mood disorder at T2 at trend level (Table 2). Higher emotion dysregulation scores were associated with a greater probability of having a mood disorder. Significant interactions emerged between SES and peer victimization as well as SES and neighborhood danger (Figure 2). Higher peer victimization was related to a higher probability of having a mood disorder in the low SES group (OR=10.01, Wald=3.90, p=.05), but not in the benign SES group, Wald=<1. Neighborhood danger had the larger effect on mood disorder likelihood in the low SES group, at trend (low SES: OR=0.20, Wald=2.78, p=.13; benign SES: Wald<1).

Fig. 2. Joint effects of Socioeconomic Status (SES) and Risk and Resilience Factors on Clinical and Functioning Outcomes

Predicting Global Functioning

TSE:

The High-risk TSE group had significantly lower functioning at Time 2, as measured by the GAF and GF:Role (Table 3). Higher levels of supportive close relationships significantly predicted higher levels of functioning across all three functioning measures. Higher emotion dysregulation was also associated with lower GAF scores; higher neighborhood danger predicted lower social functioning (GF:Social). Significant interactions with TSE and supportive close relationships emerged across all three measures of functioning (interaction predicting GAF is illustrated in Figure 1). Supportive relationships only predicted higher functioning for the low-risk group for GAF(β=.40, t=4.161, p<.001; high-risk: β=−.03, t=−0.21, p=.84), social functioning (β=.48 t=4.78, p<.001; high-risk: β=−.04, t=−0.20, p=.85), and role functioning (β=.28 t=2.46, p=.02; high-risk, β=−.19, t=−1.11, p=.28).

SES:

SES at Time 1 was not related to functioning at Time 2 (see Table 3). Supportive close relationships predicted higher functioning across GAF and GF:Social. Interestingly, higher self-reliance was related to lower social functioning scores. No other main effects of risk and resilience measures emerged. A significant interaction emerged for SES and hostile close relationships for all functioning outcomes, but was only statistically significant for GAF (Figure 2). Hostile close relationships was negatively related to GAF score at trend in the benign SES group, but showed no effect in the low SES group (β=−.27 t=−1.73, p=.1; low SES: β=.05 t=0.32, p=.75). For analyses predicting social functioning, SES x self-reliance and SES x neighborhood danger were significant (Figure 2). Self-reliance was negatively related to GF:Social for the benign SES group, but showed no effect in the low SES group (β=−.40 t=−2.41, p=.02; low SES: β=.02 t=0.11, p=.91). Higher neighborhood danger scores were related to lower GF:Social score for the low SES group, (β=−.46 t=−3.03, p=.005), but not for the benign SES group (β=.01 t=0.06, p=.95).

Conclusions

Using a resilience framework to understand developmental trajectories associated with childhood adversity offers insight into the heterogeneity of outcomes associated with early risk. The current study presents data from an initial longitudinal effort to study risk and resilience using the brief RRB. Several notable sets of findings emerged. First, the multiple levels of risk and resilience factors significantly predicted clinical diagnoses and functioning, even after controlling for baseline clinical symptoms and adversity. This bolsters support that the RRB is a promising tool for resilience research. Second, within the context of different childhood adversities (SES and TSE), a different set of factors emerged as more influential depending on type of adversity and outcome. Moreover, some risk and resilience factors emerged as more influential on outcomes (e.g., supportive close relationships) than others (i.e., recent life stressors).

The set of multiple levels of risk and resilience factors yielded from the 47-item RRB showed important influence on clinical status and functioning, even after controlling for important covariates. The current study conservatively examined the set of factors as joint predictors, controlling for shared variance among the factors, allowing for the examination of strong, unique contributions from each factor. No resilience factor examined in the current study emerged as transdiagnostic, protecting against all outcomes across TSE and SES. The findings did reveal that supportive close relationships was an important unique resilience factor for psychosis spectrum diagnoses and global functioning; whereas, emotion regulation was a particularly important resilience factor when predicting mood diagnoses. Notably, factors related to broader environmental domains (e.g., neighborhood danger and recent life stressors) were not strong, unique predictors of clinical diagnosis. Although, neighborhood danger was a significant moderator of early adversity and, as reported in Supplemental Materials, related to self-reported depression symptoms. It is likely that examination of direct relations between the risk and resilience factors and outcome measures, not controlling for the entire set of factors, would yield stronger direct relations [14]. However, the present study shows how this constellation of factors influence risk and resilience trajectories in concert, an important step for resilience research [4]. The consistency of results with prior work supports the use of the RRB in resilience research. For instance, supportive relationships as a resilience factor is well documented [45, 46]; the current findings illustrate that this type of resilience can be adequately captured in a five-item scale. It will be important for future work to understand how each of the factors interact with one another, particularly across development, to influence an individual’s risk and resilience trajectories [4, 47].

The current study found several notable similarities and differences in how the intrapersonal (individual level), interpersonal (family and other relationships), and broader environmental domain risk-resilience factors modulate developmental trajectories across the two distinct environmental adversities. Within the context of SES, multiple different risk and resilience factors across domains emerged as significant moderators. On the other hand, fewer factors were significant moderators in the context of TSE and did not prove to be protective for those with a history of TSE. Taken together, these findings align with prior developmental work that delineated the similar and divergent risk and resilience pathways linked to adversity and outcome [19, 27–29]. Our results provide further insight into possible targets for prevention, intervention, and treatment endeavors.

Prior longitudinal studies have shown that different resilience factors may be important for protecting against specific types of psychopathology [29]. The current set of findings support this prior work using the RRB. Supportive close relationships was the only significant modifier of the link between adversity and psychosis spectrum disorders. While predicting psychosis-related disorders is often underemphasized in longitudinal resilience research [48], the importance of positive interpersonal relationships in the course and treatment of psychosis related disorders is well established [49]. In contrast, several protective and risk factors were found to be significant moderators of mood disorders, spanning the different levels of resilience (i.e., emotion dysregulation, peer victimization, and neighborhood danger). The current report only focused on two clinical outcomes, but differential patterns would likely emerge for other psychopathology outcomes. Understanding how this set of risk and resilience factors influence the development of trauma-related clinical disorders is an important next step.

The factors from the RRB showed significant effects on an individual’s outcome, even in the absence of known or reported childhood adversity. It is likely these factors have important influence over how individuals deal and cope with everyday stressors or less severe adversity [12, 50]. In the current set of findings, the ability to regulate emotions and the presence of positive close relationships in youth’s lives tended to promote broad positive outcomes irrespective of early adversity. Self-reliance, while protective against psychosis symptoms, was related to poor social functioning in the SES sub-sample. Some recent work examines how certain resilience factors may be adaptive in one context, but maladaptive in other contexts [51]. Thus, it may be that for Black youth, increased self-reliance and independence, while adaptive in certain domains, may hinder proficiency in the social environment. The surprising lack of influence of some resilience factors in the at-risk groups (high TSE, low SES) could be due to several reasons. First, the assessed resilience factors may not have been powerful enough to overcome significant early risk. Prior work often relies on retrospective reports of adversity or a single resilience factor, which tends to inflate the effects of resilience [52]; the current use of prospective reports and cumulative effects of resilience, while study strengths, may have resulted in more subtle influences. The lack of information about timing and chronicity of TSEs onset of disorders may have influenced the current pattern of results; developmental timing likely plays an important role in risk and resilience pathways [53]. Resilience was assessed concurrently to outcomes; which likely obscured the prospective relations between adversity, resilience, and outcomes. Understanding the development of resilience factors, especially in the context of early life adversities, will also be important to fully understand the pathways between adversity, resilience, and mental health outcomes. Lastly, the current study only assessed psychological outcomes; resilience may be detected in the at-risk samples in cognitive or other health domains[6].

The current results should be interpreted in light of several limitations. Although the RRB and outcomes in the current study were assessed concurrently and do not indicate whether risk-resilience measures can predict future outcomes, this is the first step in highlighting the significant relationship between the RRB and developmental outcomes in youth. Moreover, many of these risk-resilience factors are theorized to be stable or trait like (e.g., self-reliance, supportive close relationships), thus the current results likely reflect information on longitudinal patterns. It is also important to note that the current sample is a convenience sample, as participants were recruited for several ongoing studies, often based on Time 1 clinical status (e.g., the presence or absence of psychosis spectrum symptoms). As such, the cohort is enriched for psychosis-related disorders; the current pattern of results may differ in a larger sample not enriched for these symptoms. On the other hand, this enrichment increases our ability to examine relations among adversity, risk-resilience, and psychosis spectrum disorders. Additionally, the current study is underpowered to detect small, subtle effects. Another important consideration surrounds the complexity of childhood adversities; the current study separated effects of SES and TSE, within each of these domains there are numerous factors that on their own have significant influence on developmental outcomes [10]. Future studies using a polyenvironmental (exposome) approach are warranted [54].

Summary

References 1. Kessler RC, McLaughlin KA, Green JG, et al. (2010) Childhood adversities and adult psychopathology in the WHO World Mental Health Surveys. Br J Psychiatry 197:378–385. 10.1192/bjp.bp.110.080499 2. Southwick SM, Bonanno GA, Masten AS, et al. (2014) Resilience definitions, theory, and challenges: interdisciplinary perspectives. Eur J Psychotraumatol 5:. 10.3402/ejpt.v5.25338 3. Masten AS, Barnes A (2018) Resilience in Children: Developmental Perspectives. Children 5:98. 10.3390/children5070098 4. Zimmerman MA, Stoddard SA, Eisman AB, et al. (2013) Adolescent Resilience: Promotive Factors That Inform Prevention. Child Dev Perspect 7:215–220. 10.1111/cdep.12042 5. Ungar M, Theron L (2020) Resilience and mental health: how multisystemic processes contribute to positive outcomes. The Lancet Psychiatry 7:441–448. 10.1016/S2215-0366(19)30434-1 6. Denckla CA, Cicchetti D, Kubzansky LD, et al. (2020) Psychological resilience: an update on definitions, a critical appraisal, and research recommendations. Eur J Psychotraumatol 11:1822064. 10.1080/20008198.2020.1822064 7. Masten AS, Lucke CM, Nelson KM, Stallworthy IC (2021) Resilience in Development and Psychopathology: Multisystem Perspectives. Annu. Rev. Clin. Psychol 17:521–549 8. Feldman R (2020) What is resilience: an affiliative neuroscience approach. World Psychiatry 19:132–150. 10.1002/wps.20729 9. Kessler RC, Ph D, Avenevoli S, et al. (2009) The National Comorbidity Survey Adolescent Supplement (NCS-A): II. Overview and Design. Acad Child Adolesc Psychiatry 48:380–385. 10.1097/CHI.0b013e3181999705 10. Mclaughlin KA, Green JG, Gruber MJ, et al. (2012) Childhood adversities and first onset of psychiatric disorders in a national sample of adolescents. Arch Gen Psychiatry 69:1151–1160. 10.1001/archgenpsychiatry.2011.2277 11. Kalisch R, Cramer AOJ, Binder H, et al. (2019) Deconstructing and Reconstructing Resilience: A Dynamic Network Approach. Perspect Psychol Sci 14:765–777. 10.1177/1745691619855637 12. Sameroff AJ, Rosenblum KL (2006) Psychosocial constraints on the development of resilience. Ann N Y Acad Sci 1094:116–124. 10.1196/annals.1376.010 13. Kalisch R, Baker DG, Basten U, et al. (2017) The resilience framework as a strategy to combat stress-related disorders. Nat Hum Behav 1:784–790. 10.1038/s41562-017-0200-8 14. Moore TM, White LK, Barzilay R, et al. (2020) Development of a scale battery for rapid assessment of risk and resilience. Psychiatry Res 288:112996. 10.1016/j.psychres.2020.112996 15. Gur RE, White LK, Shani S, et al. (2021) A binational study assessing risk and resilience factors in 22q11.2 deletion syndrome. J Psychiatr Res 138:319–325. 10.1016/j.jpsychires.2021.03.058 16. Gur RE, White LK, Waller R, et al. (2020) The disproportionate burden of the COVID-19 pandemic among pregnant black women. Psychiatry Res 293:113475. 10.1016/j.psychres.2020.113475 17. Kornfield SL, White LK, Waller R, et al. (2021) Risk And Resilience Factors Influencing Postpartum Depression And Mother-Infant Bonding During COVID-19. Health Aff 40:1566–1574. 10.1377/hlthaff.2021.00803 18. Ungar M (2019) Designing resilience research: Using multiple methods to investigate risk exposure, promotive and protective processes, and contextually relevant outcomes for children and youth. Child Abuse Negl 96:104098. 10.1016/j.chiabu.2019.104098 19. Gur RE, Moore TM, Rosen AFGG, et al. (2019) Burden of Environmental Adversity Associated With Psychopathology, Maturation, and Brain Behavior Parameters in Youths. JAMA Psychiatry 76:966–975. 10.1001/jamapsychiatry.2019.0943 20. Lähdepuro A, Savolainen K, Lahti-Pulkkinen M, et al. (2019) The Impact of Early Life Stress on Anxiety Symptoms in Late Adulthood. Sci Rep 9:1–14. 10.1038/s41598-019-40698-0 21. Hanson JL, Nacewicz BM, Sutterer MJ, et al. (2015) Behavioral Problems After Early Life Stress: Contributions of the Hippocampus and Amygdala. Biol Psychiatry 77:314–323. 10.1016/j.biopsych.2014.04.020 22. Betancourt LM, Avants B, Farah MJ, et al. (2016) Effect of socioeconomic status (SES) disparity on neural development in female African-American infants at age 1 month. Dev Sci 19:947–956. 10.1111/desc.12344 23. Hanson JL, Hair N, Shen DG, et al. (2013) Family Poverty Affects the Rate of Human Infant Brain Growth. PLoS One 8:e80954. 10.1371/journal.pone.0080954 24. Ursache A, Noble KG (2016) Socioeconomic status, white matter, and executive function in children. Brain Behav 6:1–13. 10.1002/brb3.531 25. Reiss F (2013) Socioeconomic inequalities and mental health problems in children and adolescents: A systematic review. Soc Sci Med 90:24–31. 10.1016/j.socscimed.2013.04.026 26. Barzilay R, Calkins ME, Moore TM, et al. (2019) Association between traumatic stress load, psychopathology, and cognition in the Philadelphia Neurodevelopmental Cohort. Psychol Med 49:325–334. 10.1017/S0033291718000880 27. Sumner JA, Colich NL, Uddin M, et al. (2019) Early Experiences of Threat, but Not Deprivation, Are Associated With Accelerated Biological Aging in Children and Adolescents. Biol Psychiatry 85:268–278. 10.1016/j.biopsych.2018.09.008 28. LeMoult J, Humphreys KL, Tracy A, et al. (2020) Meta-analysis: Exposure to Early Life Stress and Risk for Depression in Childhood and Adolescence. J Am Acad Child Adolesc Psychiatry 59:842–855. 10.1016/j.jaac.2019.10.011 29. Hostinar CE, Nusslock R, Miller GE (2018) Future Directions in the Study of Early-Life Stress and Physical and Emotional Health: Implications of the Neuroimmune Network Hypothesis. J Clin Child Adolesc Psychol 47:142–156. 10.1080/15374416.2016.1266647 30. Masten AS, Cicchetti D (2016) Resilience in Development: Progress and Transformation. Dev Psychopathol 4:1–63. 10.1300/J385v04n01 31. McLaughlin KA, Colich NL, Rodman AM, Weissman DG (2020) Mechanisms linking childhood trauma exposure and psychopathology: A transdiagnostic model of risk and resilience. BMC Med 18:1–11. 10.1186/s12916-020-01561-6 32. Calkins ME, Merikangas KR, Moore TM, et al. (2015) The Philadelphia Neurodevelopmental Cohort: constructing a deep phenotyping collaborative. J Child Psychol Psychiatry 56:1356–1369. 10.1111/jcpp.12416 33. Merikangas K, Avenevoli S, Costello J, et al. (2009) National comorbidity survey replication adolescent supplement (NCS-A): I. Background and measures. J Am Acad Child Adolesc Psychiatry 48:367–9. 10.1097/CHI.0b013e31819996f1 34. Kobayashi H, Nemoto T, Koshikawa H, et al. (2008) A self-reported instrument for prodromal symptoms of psychosis: Testing the clinical validity of the PRIME Screen-Revised (PS-R) in a Japanese population. Schizophr Res 106:356–362. 10.1016/j.schres.2008.08.018 35. Miller TJ, McGlashan TH, Rosen JL, et al. (2003) Prodromal assessment with the Structured Interview for Prodromal Syndromes and the Scale of Prodromal Symptoms: Predictive validity, interrater reliability, and training to reliability. Schizophr Bull 29:703–715. 10.1093/oxfordjournals.schbul.a007040 36. Moore TM, Martin IK, Gur OM, et al. (2016) Characterizing social environment’s association with neurocognition using census and crime data linked to the Philadelphia Neurodevelopmental Cohort. Psychol Med 46:599–610. 10.1017/S0033291715002111 37. Moore TM, Calkins ME, Satterthwaite TD, et al. (2019) Development of a computerized adaptive screening tool for overall psychopathology (“p”). J Psychiatr Res 116:26–33. 10.1016/j.jpsychires.2019.05.028 38. Furman W, Buhrmester D (1985) Children’s perceptions of the personal relationships in their social networks. Dev Psychol 21:1016–1024. 10.1037//0012-1649.21.6.1016 39. Kaufman EA, Xia M, Fosco G, et al. (2016) The Difficulties in Emotion Regulation Scale Short Form (DERS-SF): Validation and Replication in Adolescent and Adult Samples. J Psychopathol Behav Assess 38:443–455. 10.1007/s10862-015-9529-3 40. Betts LR, Houston JE, Steer OL (2015) Development of the Multidimensional Peer Victimization Scale–Revised (MPVS-R) and the Multidimensional Peer Bullying Scale (MPVS-RB). J Genet Psychol 176:93–109. 10.1080/00221325.2015.1007915 41. Calkins ME, Moore TM, Satterthwaite TD, et al. (2017) Persistence of psychosis spectrum symptoms in the Philadelphia Neurodevelopmental Cohort: a prospective two-year follow-up. World Psychiatry 16:62–76. 10.1002/wps.20386 42. McGlashan T, Miller TJ, Woods SW, et al. (2003) Structured Interview for Prodromal Syndromes, version 4. PRIME Research Clinic, Yale School of Medicine., New Haven, CT 43. Cornblatt BA, Auther AM, Niendam T, et al. (2007) Preliminary findings for two new measures of social and role functioning in the prodromal phase of schizophrenia. Schizophr Bull 33:688–702. 10.1093/schbul/sbm029 44. Pilkonis PA, Choi SW, Reise SP, et al. (2011) Item banks for measuring emotional distress from the PROMIS: Depression, anxiety and anger. Assessment 18:263–283. 10.1177/1073191111411667 45. McLaughlin KA, Lambert HK (2017) Child trauma exposure and psychopathology: mechanisms of risk and resilience. Curr Opin Psychol 14:29–34. 10.1016/j.copsyc.2016.10.004 46. Trickey D, Siddaway AP, Meiser-Stedman R, et al. (2012) A meta-analysis of risk factors for post-traumatic stress disorder in children and adolescents. Clin Psychol Rev 32:122–138. 10.1016/j.cpr.2011.12.001 47. Masten AS, Coatsworth JD (1998) The Development of Competence in Favorable and Unfavorable Environments: Lessons from Research on Successful Children. Am Psychol 53:205–220. 10.1037/0003-066X.53.2.205 48. Pan PM, Gadelha A, Argolo FC, et al. (2019) Childhood trauma and adolescent psychotic experiences in a community-based cohort: The potential role of positive attributes as a protective factor. Schizophr Res 205:23–29. 10.1016/j.schres.2018.06.044 49. Galderisi S, Rossi A, Rocca P, et al. (2014) The influence of illness-related variables, personal resources and context-related factors on real-life functioning of people with schizophrenia. World Psychiatry 13:275–287. 10.1002/wps.20167 50. Seery MD, Quinton WJ (2016) Understanding Resilience. In: Advances in Experimental Social Psychology, 1st ed. Elsevier Inc., Buffalo, US, pp 181–245 51. Mahdiani H, Ungar M (2021) The Dark Side of Resilience. Advers Resil Sci 2:147–155. 10.1007/s42844-021-00031-z 52. Green JG, McLaughlin KA, Berglund PA, et al. (2010) Childhood Adversities and Adult Psychiatric Disorders in the National Comorbidity Survey Replication I. Arch Gen Psychiatry 67:113. 10.1001/archgenpsychiatry.2009.186 53. Chaby LE, Zhang L, Liberzon I (2017) The effects of stress in early life and adolescence on posttraumatic stress disorder, depression, and anxiety symptomatology in adulthood. Curr Opin Behav Sci 14:86–93. 10.1016/j.cobeha.2017.01.001 54. Guloksuz S, van Os J, Rutten BPF (2018) The Exposome Paradigm and the Complexities of Environmental Research in Psychiatry. JAMA Psychiatry 75:985. 10.1001/jamapsychiatry.2018.1211 55. Masten AS (2019) Resilience from a developmental systems perspective. World Psychiatry 18:101–102. 10.1002/wps.20591

Taken together, the current results find that factors across multiple levels of resilience (i.e., child/individual, family systems, and social and neighborhood environmental domains), assessed via the brief self-report RRB, both protect against and promote certain psychological outcomes. The effect of resilience factors depends on the type of adversity and outcome under examination, confirming multiple prior reports that pathways of resilience are complex [5, 7, 30, 55]. Importantly, the current study shows that a diverse set of risk and protective factors that contribute to these complex pathways can be assessed via a brief and easily administered self-report battery.