Abstract
Introduction:
There is a clear bi-directional link between stressful events and depressive symptoms in adolescence, but the directionality of this link for anxiety symptoms remains underexamined. We critically evaluate the longitudinal relationship between stressors and anxiety among youth. Specifically, we examine whether stressors predict anxiety symptoms over a 1.5-year period (stress causation), and whether anxiety symptoms predict stressors over this period (stress generation). We examine potential influencing factors, including stressor type (independent vs. dependent) and emotion dysregulation (nonacceptance; goal-directed difficulty).
Methods:
Social, separation, and physical anxiety symptoms, and frequency and stressor type, were assessed every 3 months for 1.5 years among community youth (n=528, ages 8–17). Baseline emotion dysregulation was assessed. Time-lagged analyses evaluated the bi-directional relationship of stress and anxiety over time, controlling for previous anxiety and depression.
Results:
Interpersonal stressors predicted subsequent physical and social anxiety symptoms, but anxiety did not predict subsequent stressors. Both nonacceptance and goal-directed difficulties predicted subsequent anxiety symptoms and stressors, but did not moderate the relationship.
Conclusion:
The findings supported the stress causation model for youth anxiety, but not the stress generation model. Additionally, nonacceptance and goal-directed difficulty predicted both greater subsequent anxiety symptoms and stressors. We discuss implications for prevention and intervention.
Keywords: stress causation, stress generation, anxiety, emotion regulation, longitudinal
Exposure to stressful events increases from childhood into adolescence (Hankin et al., 2007). Considerable research demonstrates that stressful events predict depression (e.g., Aliri et al., 2019; Hammen, 2005; Vrshek-Schallhorn et al., 2015). Yet stressors also contribute to other mental health outcomes, including other internalizing and externalizing disorders (Grant & McMahon, 2005; March-Llanes et al., 2017; Moya-Higueras et al., 2018). Prior cross-sectional and retrospective work has established a link between stress and anxiety among youth (e.g., Asselmann & Beesdo-Baum, 2015; McMahon et al., 2003); however, the directionality of this relationship is unclear. Further, potential moderators of the relationship between stress and anxiety, such as emotion regulation strategies, remain underexplored. The current study used a multi-wave prospective design to examine whether stressors precede anxiety symptoms (stress causation model), whether anxiety symptoms precede subsequent stressors (stress generation model), and what role emotion dysregulation may play in this link.
Link between Stressors and Depression
We first examine the well-developed depression and stressors literature to inform our hypotheses regarding the temporal relationship between stressful events and anxiety symptoms. A large body of research has established a bi-directional link between stressful events and depressive symptoms (for a review, see Liu & Alloy, 2010). Stressors frequently precede depressive symptoms (e.g., Vrshek-Schallhorn et al., 2015). This process is referred to as stress causation, since stressful events are thought to contribute to later elevations of depressive symptoms. Research also implicates stress generation in the link between stressors and depression (Hammen, 1991; Hammen, 2005; March-Llanes et al., 2017). The stress generation model posits that individuals with depression contribute to, or generate, the stressful events they experience. Several studies support both the stress causation and stress generation models within the same sample for depression in youth (e.g., Cole et al., 2006; Hankin et al., 2007), and indicate that interpersonal stressors may play a particularly important role in both models (e.g., Hankin et al., 2007; Rudolph, 2008; Sheets & Craighead, 2014).
Link between Stressors and Anxiety
The findings regarding the temporal relationship between depression and stressful events inform hypotheses regarding the directionality of the link between anxiety and stressful events. Although a cross-sectional, retrospective link between stress and anxiety among youth has been established (e.g., Grant et al., 2014; McMahon et al., 2003), longitudinal studies examining the directionality of this relationship are more limited, and their findings are varied.
Stress Causation.
Numerous studies have found broad support for the stress causation model for anxiety in youth (e.g., Barrocas & Hankin, 2011; Hamilton et al., 2016; Hankin, 2008; Kopala-Sibley et al., 2015). A greater number of negative life events has been shown to occur in the year prior to the onset of anxiety disorders in children, compared to matched healthy controls (Allen et al., 2008). Another study found that stressors predicted anxiety symptoms over time in a sample of 350 adolescents ages 11–17 assessed every 5 weeks over a 5-month period (Hankin, 2008). In that same sample, increases in girls’ anxiety trajectories were linked to the presence of greater previous stressors, further supporting the stress causation model (Hankin, 2009).
Stress Generation.
In contrast, support for the stress generation model for anxiety in youth is equivocal. Several studies have found that symptoms of anxiety predict greater subsequent stressors, particularly dependent stressors (i.e., those stressors to which one directly contributes; Connolly et al., 2010; Shapero et al., 2013; Uliaszek et al., 2012). For example, anxiety symptoms predicted subsequent increases in self-reported achievement stressors and interpersonal dependent stressors among youth (Shapero et al., 2013). Anxiety symptoms similarly predicted subsequent stressors across one year in adolescents (Uliaszek et al., 2012). Other studies have not found evidence that anxiety symptoms precede increases in stressful events (e.g., Barrocas & Hankin, 2011). Given these conflicting results, it is important to clarify the directionality of the relationship between anxiety and stress.
Notably, most existing work has only examined one temporal direction consistent with either the stress causation model or stress generation model within a single sample. Examining this link bi-directionally within the same sample would permit a direct comparison of the stress causation and stress generation models, while holding sample variables constant, allowing for clearer conclusions about temporal directionality.
Stressor Type
It is also important to examine the role of stressor type within this link. Stressor type has emerged as an important consideration within the stress literature yet little research examines the role of stressor type in the relationship between stressors and anxiety. Studies on stress have distinguished between dependent and independent stressors, or those stressors to which one directly contributes, at least in part (e.g., relationship conflict), versus those “fateful” stressors that one has minimal, if any, influence over (e.g., a death in the family) (Hammen, 1991). In addition, stressors can be separated into interpersonal stressors (e.g., the end of a friendship) and achievement-oriented stressors (e.g., failing a test).
Although the majority of studies within the anxiety literature have examined the role of stressors broadly defined, several studies found that interpersonal dependent stressors were particularly associated with anxiety symptoms (e.g., Hamilton et al., 2016; Kopala-Sibley et al., 2015; Connolly et al., 2010). Interpersonal dependent stressors may, therefore, serve as the driving force in the putative link between anxiety and stressful events. That stated, given that anxiety symptoms are often linked to worries about uncontrollable or unpredictable events (e.g., worry about natural disasters, harm to parents; Chorpita & Barlow, 1998), it is also possible that independent stressors are more closely linked to increases in anxiety over time than dependent stressors. In support, one study found that independent stressors were more predictive of internalizing symptoms than dependent stressors (Moya-Higuera et al., 2018). Given that stressor types may be differentially related to symptoms of anxiety, there is a need for bi-directional studies that directly compare these two alternative hypotheses concerning the type of stressors most implicated in the link between stressors and anxiety.
Moderators of the Anxiety-Stress Relationship
Few studies have examined moderators of the relationship between stressful events and anxiety among youth. The limited existing work has found mixed evidence for gender (e.g., Brown et al., 2016; Shapero et al., 2013; Uliaszek et al., 2010) and evidence for self-esteem (Moksnes et al., 2010) as moderators of the association between stressful events and anxiety during adolescence. Given the limited nature and inconclusive findings of this collective work, additional research needs to explore additional potential moderators of the relationship between stressful events and anxiety, particularly in childhood and adolescence.
Emotion regulation strategies.
Mindfulness-and acceptance-based emotion regulation factors represent an unexplored area of potential moderating influence on the relationship between stressful events and anxiety. In particular, emotion nonacceptance and difficulty engaging in goal-directed behavior when upset represent two forms of emotion dysregulation that longitudinally predict increases in anxiety symptoms over time among youth (Schneider et al., 2016). Therefore, stressful events may be especially likely to induce anxiety among those who have difficulty continuing with day-to-day activities when upset (i.e., difficulty with goal-directed behavior) or those who have difficulty accepting their negative emotions (i.e., emotion nonacceptance). Similar logic may apply to stress generation as well, such that anxious youth who have difficulty accepting their emotions or engaging in goal-directed behavior when upset may be especially likely to create additional stressors for themselves. Therefore, it is important to explore whether these emotion dysregulation variables moderate the association between stressors and anxiety both within a stress causation and within a stress generation framework.
Current Study
The current study examined the longitudinal relationship between stressful events and anxiety symptoms in a sample of community youth aged 8–17. We evaluated whether the number and frequency of stressful events predicted anxiety symptoms (stress causation), and whether anxiety symptoms predicted stressors (stress generation) over 1.5 years. We examined whether type of stressor (dependent [interpersonal and achievement] versus independent) affected the longitudinal stress-anxiety association. Finally, we investigated whether relevant forms of emotion dysregulation moderated longitudinal associations between anxiety and stress.
We hypothesized a bi-directional relationship between stressful events and anxiety symptoms consistent with stress causation and stress generation models. We also hypothesized that this bi-directional relationship would be more pronounced for those high in emotion nonacceptance and difficulties with goal-directed behavior. Given the link between interpersonal dependent stressors and depression (e.g., Hammen, 1991; Vrshek-Schallhorn et al., 2015), we hypothesized that dependent stressors may serve as the driving force in the putative link between anxiety and stressors. However, it is also possible that independent stressors, given their uncontrollable nature, are more closely linked to increases in anxiety over time. We thus evaluated these two alternative hypotheses.
Methods
Participants and Procedures
Participants were youth recruited from local area schools in Denver, CO and the central New Jersey area as part of a large, multiwave, longitudinal study approved by each site’s IRB (Gene, Environment, and Mood [GEM] Study; see Hankin et al., 2015, for details on sample characteristics, consent, and overall study design). Brief information letters were sent home directly to families in the participating school districts with a child in third, sixth, or ninth grade to recruit a representative sample of the general community. The sample for the current study consists of 528 youth (55.2% female) ranging in age from 8–17 years (mean age = 13.33 years, SD = 2.39). The current study focused on the assessments from the in-person assessment that occurred at 18 months post-baseline through the final 36 months follow-up. We focused on the data collected for these 7 time points spanning the 18- to 36-month follow-up period because the relevant measures assessing the putative moderating emotion variables were administered at these assessments. Therefore, in this paper “baseline” reflects the 18-month time point of the larger GEM study.
Measures
Multidimensional Anxiety Scale for Children (MASC; March et al.,1997).
The MASC is a 39-item self-report measure of anxiety symptoms in youth. The MASC was administered every 3 months for 18 months. We used the social anxiety, separation anxiety, and physical anxiety subscales, which map onto risk for social anxiety disorder, separation anxiety disorder, and panic disorder, respectively (Van Gastel & Ferdinand, 2008; Wei et al., 2014). The harm avoidance subscale was not used, as it has questionable discriminant validity and correlation with the other three subscales (Schneider et al., 2016). The MASC shows good reliability and validity (March et al., 1997). Internal consistency was adequate across time points (αs>.80 for physical anxiety, αs>.80 for social anxiety, and αs>.63 for separation anxiety).
Adolescent Life Events Questionnaire (ALEQ; Hankin & Abramson, 2002).
The ALEQ is a 37-item self-report measure that assesses a broad range of dependent and independent stressors typically occurring in adolescence, including problems with school, peers, and family. The ALEQ shows good test-retest reliability and validity (Hankin, 2008) and was administered every 3 months for 18 months. Youth rate how often a stressor has occurred in the past 3 months on a 5-point scale from “never” to “always.” These ratings are summed to create a total overall score, and a total for each stressor type. Items can be categorized as dependent-interpersonal (e.g. “Fighting with or problems with a friend”), dependent-achievement (e.g. “Got a bad grade on an exam, project, or paper in class”), or independent (e.g. “Death of a family member”). Items on the ALEQ have previously been categorized into these subscales with 100% agreement across independent raters (see Hankin et al., 2010). In total, 16 items were categorized as dependent-interpersonal, 2 items as dependent-achievement, and 14 items as independent. Five items were excluded as not belonging to any one category and were not analyzed further.
Difficulties in Emotion Regulation Scale–Youth Adapted (DERS; Gratz & Roemer, 2004).
The DERS-YA is an 11-item self-report measure adapted for the larger GEM study from the 36-item DERS, and was administered at the current study’s baseline. These 11 items represent the 2–3 items from each subscale (Awareness, Nonacceptance, Goals, and Impulse) with high loadings onto their respective factors, based on a psychometric study of the DERS in a large sample of youth (Neumann et al., 2010). The items were reworded as necessary to create a youth-friendly measure for the GEM study. Only the Nonacceptance and Goals subscales, each containing 3 items, were examined in the current study. We chose a priori to only focus on these 2 subscales in order to assess the effect of more active and potentially dysfunctional forms of emotion regulation with strong relevance for internalizing disorders (e.g., as opposed to impulsivity, which is more relevant for externalizing disorders; Olson et al., 1999). The Nonacceptance scale assessed lack of emotion acceptance (e.g., “When I’m upset, I feel bad for feeling that way”). The Goals scale assessed difficulty engaging in goal-directed behaviors when upset (e.g., “When I’m upset, I have a hard time doing things”). The DERS has good test-retest reliability, high internal consistency, and adequate construct and predictive validity (Gratz & Roemer, 2004). Internal consistency was good for both subscales (αs=.83).
Children’s Depressive Inventory (CDI; Kovacs, 1985).
The CDI is a 27-item self-report measure of depression symptoms in youth. Youth rate items on a scale from 0 to 2; higher scores indicate greater severity. The CDI has good reliability and validity (Kovacs, 1985). The CDI was used to control for depressive symptoms at this study’s baseline. Internal consistency at baseline was good (α=.87).
Statistical Approach
Hierarchical linear modeling (HLM 7.01; Raudenbush et al., 2013) was utilized to model time-lagged change over time. Although questions of bidirectional change over time are often studied using structural equation modeling approaches (e.g., the cross-lagged panel model), several studies have demonstrated that multilevel modeling and structural equation modeling provide equivalent or highly similar results (e.g., MacCallum et al., 1997). However, an advantage of multilevel models is their ability to separate within-person from between-person dynamic processes (Schuurman et al., 2016), which are aggregated in the cross-lagged panel model (Hamaker et al., 2015). As the current study examines within-person processes implicated in anxiety, stress causation, and stress generation, multilevel modeling was chosen as the most appropriate data analytic approach.
Repeated measures of anxiety and stress were group-mean centered on Level 1 to capture change within an individual over time. Each MASC anxiety subscale (physical, social, and separation) was examined separately across 3 models. Anxiety (physical, social, or separation) was coded as time (t) and stressors as time minus 1 (t-1) to test the stress causation hypothesis in each of these models, and vice-versa to test the stress generation hypothesis. Time was centered at 18 months, so the intercept represented the level of anxiety or stress that individuals ultimately developed in the study. All analyses included covariates of age, gender, and initial depression on Level 2. Random slopes and intercepts were included for all effects.
Stress causation model.
To examine whether stressors predicted later anxiety symptoms, we conducted time-lagged analyses. First, we entered repeated physical, social, or separation anxiety subscale scores from months 3– 18 (6 time points, coded as t) as the outcome measure on Level 1 of the multi-level model. Repeated stress scores and anxiety subscale scores at the previous time points (months baseline [BL] to 15, coded as t-1) were also captured on Level 1. This facilitated assessment of how stressors at t-1 predicted anxiety at t, controlling for anxiety at t-1, within individuals across the 1.5-year period. We covaried age, gender, and initial depression symptoms on Level 2; age and initial depression symptoms were grand-mean centered and gender was dummy-coded (female = 0, male = 1).
Stress generation model.
To examine how earlier anxiety predicted later overall stressors, on Level 1 total stress scores from 3–18 months served as the outcome measure (coded as t), and anxiety subscale scores and stress scores from BL-15 months served as the predictors (coded as t-1). This allowed us to assess how anxiety at t-1 predicted stressors at t, controlling for stressors at t-1, within individuals across the 1.5-year period. We included the same Level 2 covariates as above.
Dependent vs. independent stressors.
To examine the role of stressor type, we conducted the analyses described above, but replaced the overall stress score with either the independent or dependent stress subscale score. If the dependent stress subscale was significant, we re-ran the model with the dependent-interpersonal and dependent-achievement subscales to identify the specific type of dependent stressor driving the relationship.
Nonacceptance and goal-directed behavior.
To test whether nonacceptance and difficulties with goal-directed behavior at BL predicted level of overall stressors or specific anxiety symptoms over time (each Level 1 variables), and whether they moderated the relationship between the two, we added BL nonacceptance or BL difficulties with goal-directed behavior to Level 2 in all analyses that examined total stressors. This enabled us to simultaneously test for prediction (i.e., whether these Level 2 variables predicted the Level 1 variables) and moderation (i.e., whether these Level 2 variables moderated the relationship between the Level 1 variables) within the same model.
Results
Descriptives
Descriptive statistics for participant variables, MASC subscale scores, and ALEQ stress ratings are displayed in Table 1. Mean levels of anxiety as measured by MASC subscale scores were comparable to scores found in similar samples (e.g., Baldwin & Dadds, 2007). The majority of youth (98.3% at baseline) reported experiencing at least one stressor during the previous three months. Correlations among predictor variables are provided in Table 1.
Table 1.
Correlations and Descriptive Statistics Among L2 and L1 Predictor Variables (n = 528)
| Descriptives | L2 Correlations at BL | ||||||||
| M (SD) Range | CDI | DERS Nonacceptance | DERS Goals | Age | Gender | ||||
| CDI | 5.31 (5.94) 0–37 |
1 | |||||||
| DERS Nonacceptance | 8.23 (4.37) 3–21 |
.29** | 1 | ||||||
| DERS Goals | 10.98 (4.99) 3–21 |
.35** | .53** | 1 | |||||
| Age | 13.33 (2.39) 8.5–17.52 |
.25** | .13** | .14** | 1 | ||||
| Gender | 0.42 (0.49) 0–1 |
.07 | .03 | .05 | .01 | 1 | |||
| Descriptives | L1 Correlations at BL and 18mo | ||||||||
|
M (SD) Range |
MASC PH | MASC SA | MASC SP | ALEQ Total | ALEQ Dependent | ALEQ Independent | ALEQ Interpersonal | ALEQ Achievement | |
| MASC PH | 6.84 (5.94) 0–33 |
1 | |||||||
| MASC SA | 7.95 (5.66) 0–27 |
.50** .52** |
1 | ||||||
| MASC SP | 5.62 (4.21) 0–26 |
.38** .35** |
.37** .36** |
1 | |||||
| ALEQ Total | 56.05 (14.27) 37–144 |
.46** .56** |
.42** .53** |
.05 .17** |
1 | ||||
| ALEQ Dependent | 30.04 (8.62) 19–76 |
.45** .54** |
.44** .53** |
.05 .15** |
.96** .96** |
1 | |||
| ALEQ Independent | 19.36 (4.68) 14–48 |
.39** .47** |
.27** .38** |
.06 .18** |
.85** .83** |
.69** .66** |
1 | ||
| ALEQ Interpersonal | 25.23 (7.53) 15–65 |
.47** .54** |
.45** .53** |
.06 .15** |
.94** .94** |
.98** .98** |
.67** .65** |
1 | |
| ALEQ Achievement | 3.62 (1.57) 2–10 |
.23** .31** |
.27** .35** |
−.02 .10* |
.60** .59** |
.63** .62** |
.45** .40** |
.49** .48** |
1 |
Note. Descriptive statistics and L2 predictors measured at baseline. Top numbers among L1 correlations represent correlations between L1 predictors at BL and bottom numbers represent correlations between L1 predictors at 18mo; L2 = Level 2, BL = baseline, M = mean, SD = standard deviation, CDI = Children’s Depressive Inventory, DERS = Difficulties in Emotion Regulation Scale – Youth Adapted, L1 = Level 1, 18mo = 18 months, MASC = Multidimensional Anxiety Scale for Children, PH = physical anxiety, SA = social anxiety, SP = separation anxiety, ALEQ = Adolescent Life Events Questionnaire.
p < .05, two-tailed
p < .01, two-tailed
Stress Causation Model: Prediction of Anxiety Symptoms
Stressful events.
Greater number and frequency of stressful events at the previous time points significantly predicted greater levels of physical anxiety and social anxiety symptoms at the following time points, controlling for previous levels of these anxiety symptoms and baseline depression, evidencing a significant main effect of stressful events on subsequent physical and social anxiety (see Table 2). Stressors did not, however, predict separation anxiety symptoms.
Table 2.
Earlier Stressors (t-1) and Emotion Dysregulation Variables (BL) Predicting Later Anxiety Types (t) (n = 512)
| Fixed Effect | Unstandardized Effect | Standard Error | t-ratio | p-value |
|---|---|---|---|---|
| Physical Anxiety (t) Predicted by Stressors (t-1) | ||||
| Nonacceptance | ||||
| Nonacceptance Intercept (18mo) | 0.26 | 0.07 | 3.95 | <.001** |
| Stress Slope (t-1) | 0.04 | 0.02 | 2.51 | .01* |
| Nonacceptance × Stress Slope | 0.002 | 0.003 | 0.86 | .39 |
| Goal-directed Behavior | ||||
| Goals Intercept (18mo) | 0.20 | 0.05 | 3.69 | <.001** |
| Stress Slope (t-1) | 0.04 | 0.02 | 2.46 | .01* |
| Goals × Stress Slope | −0.0002 | 0.003 | −0.07 | .94 |
| Social Anxiety (t) Predicted by Stressors (t-1) | ||||
| Nonacceptance | ||||
| Nonacceptance Intercept (18mo) | 0.27 | 0.06 | 4.18 | <.001** |
| Stress Slope (t-1) | 0.03 | 0.02 | 2.17 | .03* |
| Nonacceptance × Stress Slope | 0.002 | 0.003 | 0.65 | .51 |
| Goal-directed Behavior | ||||
| Goals Intercept (18mo) | 0.23 | 0.05 | 4.49 | <.001** |
| Stress Slope (t-1) | 0.04 | 0.02 | 2.19 | .03* |
| Goals × Stress Slope | 0.0001 | 0.003 | 0.04 | .97 |
| Separation Anxiety (t) Predicted by Stressors (t-1) | ||||
| Nonacceptance | ||||
| Nonacceptance Intercept (18mo) | 0.13 | 0.04 | 3.24 | 0.001** |
| Stress Slope (t-1) | 0.006 | 0.01 | 0.51 | 0.61 |
| Nonacceptance × Stress Slope | −0.0002 | 0.002 | −0.09 | 0.93 |
| Goal-directed Behavior | ||||
| Goals Intercept (18mo) | 0.13 | 0.03 | 3.70 | <.001** |
| Stress Slope (t-1) | 0.005 | 0.01 | 0.44 | .66 |
| Goals × Stress Slope | 0.00006 | 0.002 | 0.03 | .97 |
Note. Repeated stress and anxiety scores were entered on Level 1 of the model and BL emotion dysregulation variables were entered on Level 2. Age, gender, and initial depression were included as covariates on Level 2. All interactions represent cross-level interactions; t-1 = time minus 1, i.e., BL thru 15 months; t = time, i.e., months 3 thru 18; BL = baseline
p < .05
p < .01
Stressor type.
Independent stressors did not predict any type of anxiety symptoms (see Table 3). Dependent stressors, by contrast, predicted greater subsequent physical anxiety and social anxiety symptoms, but not separation anxiety symptoms (see Table 3). This effect was driven by dependent interpersonal stressors rather than dependent achievement stressors.
Table 3.
Earlier Stressor Types (t-1) Predicting Later Anxiety Types (t) (n = 526)
| Fixed Effect | Unstandardized Effect | Standard Error | t-ratio | p-value |
|---|---|---|---|---|
| Independent Stressors | ||||
| Physical Anxiety | ||||
| Intercept (18mo) | 6.84 | 0.36 | 18.87 | <.001** |
| Physical Anxiety Slope (t-1) | −0.17 | 0.03 | −5.82 | <.001** |
| Independent Stress Slope (t-1) | 0.05 | 0.04 | 1.26 | .21 |
| Time Slope (t-1) | −0.10 | 0.07 | −1.47 | .14 |
| Social Anxiety | ||||
| Intercept (18mo) | 8.24 | 0.34 | 24.33 | <.001** |
| Social Anxiety Slope (t-1) | −0.15 | 0.03 | −5.15 | <.001** |
| Independent Stress Slope (t-1) | 0.06 | 0.04 | 1.28 | .20 |
| Time Slope (t-1) | −0.11 | 0.07 | −1.63 | .10 |
| Separation Anxiety | ||||
| Intercept (18mo) | 5.18 | 0.23 | 22.65 | <.001** |
| Separation Anxiety Slope (t-1) | −0.13 | 0.03 | −5.11 | <.001** |
| Independent Stress Slope(t-1) | 0.003 | 0.03 | 0.10 | .92 |
| Time Slope (t-1) | −0.26 | 0.05 | −5.63 | <.001** |
| Dependent Stressors | ||||
| Physical Anxiety | ||||
| Intercept (18mo) | 6.88 | 0.36 | 18.95 | <.001** |
| Physical Anxiety Slope (t-1) | −0.18 | 0.03 | −6.10 | <.001** |
| Dependent Stress Slope (t-1) | 0.07 | 0.02 | 2.82 | .005** |
| Interpersonal Stress Slope | 0.07 | 0.03 | 2.72 | .007** |
| Achievement Stress Slope | 0.14 | 0.10 | 1.33 | .18 |
| Time Slope (t-1) | −0.08 | 0.07 | −1.17 | .24 |
| Social Anxiety | ||||
| Intercept (18mo) | 8.27 | 0.34 | 24.41 | <.001** |
| Social Anxiety Slope (t-1) | −0.16 | 0.03 | −0.65 | <.001** |
| Dependent Stress Slope (t-1) | 0.05 | 0.02 | 2.23 | .03* |
| Interpersonal Stress Slope | 0.06 | 0.03 | 2.34 | .02* |
| Achievement Stress Slope | 0.10 | 0.10 | 1.03 | .31 |
| Time Slope (t-1) | −0.10 | 0.07 | −1.50 | .13 |
| Separation Anxiety | ||||
| Intercept (18mo) | 5.20 | 0.23 | 22.62 | <.001** |
| Separation Anxiety Slope (t-1) | −0.14 | 0.03 | −5.40 | <.001** |
| Dependent Stress Slope (t-1) | 0.02 | 0.02 | 0.95 | .34 |
| Time Slope (t-1) | −0.26 | 0.05 | −5.33 | <.001** |
Note. Repeated stress and anxiety scores were entered on Level 1 of the model. Age, gender, and initial depression were included as covariates on Level 2. The anxiety slopes (i.e., test-retest paths) reflect the association between baseline and change over time. Negative anxiety slopes indicate that those who reported higher symptom levels at baseline have lower change across time; t-1 = time minus 1, i.e., BL thru 15 months; t = time, i.e., 3 thru 18 months; BL = baseline.
p < .05
p < .01
Nonacceptance.
Greater nonacceptance of emotions predicted greater levels of physical, social, and separation anxiety at the following time points, controlling for anxiety symptoms at the previous time points and baseline depression (see Table 2). Nonacceptance did not moderate the relationship between number and frequency of stressors and any types of anxiety symptoms.
Goal-directed behavior.
Greater difficulty with goal-directed behavior at baseline significantly predicted greater levels of physical, social, and separation anxiety symptoms at the subsequent time points, controlling for anxiety symptoms at the previous time points and baseline depression (see Table 2). Similarly to nonacceptance, goal-directed behavior did not moderate the relationship between stressful events overall and any anxiety symptoms.
Stress Generation: Prediction of Stressors
Anxiety.
Greater level of physical, social, and separation anxiety symptoms at previous time points did not predict greater number and frequency of stressors at subsequent time points, controlling for previous stressors and baseline depression (see Table 4). We examined stressor type to ensure that there was not an effect of physical, social, or separation anxiety on a specific type of stressor that was obscured in the omnibus test. However, anxiety symptoms also did not predict independent or dependent stressors (see Table 5).
Table 4.
Earlier Anxiety Types (t-1) and Emotion Dysregulation Variables (BL) Predicting Later Stressors (t) (n = 512)
| Fixed Effect | Unstandardized Effect | Standard Error | t-ratio | p-value |
|---|---|---|---|---|
| Stressors (t) Predicted by Physical Anxiety (t-1) | ||||
| Nonacceptance | ||||
| Nonacceptance Intercept (18mo) | 0.36 | 0.14 | 2.62 | .009** |
| Physical Anxiety Slope (t-1) | −0.04 | 0.06 | −0.66 | .51 |
| Nonacceptance × Physical Anxiety Slope | −0.01 | 0.01 | −0.89 | .37 |
| Goal-directed Behavior | ||||
| Goals Intercept (18mo) | 0.37 | 0.13 | 2.81 | .005** |
| Physical Anxiety Slope (t-1) | −0.04 | 0.06 | −0.72 | 0.47 |
| Goals × Physical Anxiety Slope | 0.001 | 0.01 | 0.13 | .90 |
| Stressors (t) Predicted by Social Anxiety (t-1) | ||||
| Nonacceptance | ||||
| Nonacceptance Intercept (18mo) | 0.38 | 0.14 | 2.77 | .006** |
| Social Anxiety Slope (t-1) | −0.07 | 0.06 | −1.04 | 0.30 |
| Nonacceptance × Social Anxiety Slope | −0.02 | 0.01 | −1.34 | .18 |
| Goal-directed Behavior | ||||
| Goals Intercept (18mo) | 0.38 | 0.13 | 2.92 | .004** |
| Social Anxiety Slope (t-1) | −0.07 | 0.06 | −1.05 | .30 |
| Goals × Social Anxiety Slope | −0.0001 | 0.01 | −0.01 | .99 |
| Stressors (t) Predicted by Separation Anxiety (t-1) | ||||
| Nonacceptance | ||||
| Nonacceptance Intercept (18mo) | 0.38 | 0.14 | 2.77 | .006** |
| Separation Anxiety Slope (t-1) | 0.09 | 0.08 | 1.09 | .28 |
| Nonacceptance × Separation Anxiety Slope | −0.01 | 0.01 | −0.99 | .32 |
| Goal-directed Behavior | ||||
| Goals Intercept (18mo) | 0.37 | 0.13 | 2.89 | .004** |
| Separation Anxiety Slope (t-1) | 0.08 | 0.08 | 0.98 | .33 |
| Goals × Separation Anxiety Slope | 0.003 | 0.01 | 0.24 | .81 |
Note. Repeated stress and anxiety scores were entered on Level 1 of the model and BL emotion dysregulation variables were entered on Level 2. Age, gender, and initial depression were included as covariates on Level 2. All interactions represent cross-level interactions; t-1 = time minus 1, i.e., BL thru 15 months; t = time, i.e., 3 thru 18 months; BL = baseline
p < .05
p < .01
Table 5.
Earlier Anxiety Types (t-1) Predicting Later Stressor Types (t) (n = 526)
| Fixed Effect | Unstandardized Effect | Standard Error | t-ratio | p-value |
|---|---|---|---|---|
| Independent Stressors (t) | ||||
| Physical Anxiety | ||||
| Intercept (18mo) | 20.10 | 0.27 | 73.93 | <.001** |
| Physical Anxiety Slope (t-1) | −0.02 | 0.02 | −0.73 | .47 |
| Independent Stress Slope (t-1) | −0.11 | 0.03 | −3.80 | <.001** |
| Time Slope (t-1) | 0.15 | 0.05 | 2.99 | .003** |
| Social Anxiety | ||||
| Intercept (18mo) | 20.10 | 0.27 | 74.32 | <.001** |
| Social Anxiety Slope (t-1) | −0.02 | 0.02 | −0.63 | .53 |
| Independent Stress Slope (t-1) | −0.11 | 0.03 | −3.82 | <.001** |
| Time Slope (t-1) | 0.15 | 0.05 | 2.99 | .003** |
| Separation Anxiety | ||||
| Intercept (18mo) | 20.08 | 0.27 | 74.73 | <.001** |
| Separation Anxiety Slope (t-1) | 0.01 | 0.03 | 0.28 | .78 |
| Independent Stress Slope (t-1) | −0.12 | 0.03 | −4.01 | <.001** |
| Time Slope (t-1) | 0.14 | 0.05 | 2.81 | .005** |
| Dependent Stressors (t) | ||||
| Physical Anxiety | ||||
| Intercept (18mo) | 31.02 | 0.48 | 64.65 | <.001** |
| Physical Anxiety Slope (t-1) | 0.01 | 0.04 | 0.33 | .74 |
| Dependent Stress Slope (t-1) | −.07 | 0.03 | −2.17 | .03* |
| Time Slope (t-1) | 0.21 | 0.09 | 2.26 | .02* |
| Social Anxiety | ||||
| Intercept (18mo) | 31.07 | 0.48 | 64.87 | <.001** |
| Social Anxiety Slope (t-1) | −0.02 | 0.04 | −0.42 | .67 |
| Dependent Stress Slope (t-1) | −0.07 | 0.03 | −2.29 | .02* |
| Time Slope (t-1) | 0.22 | 0.09 | 2.46 | .01* |
| Separation Anxiety | ||||
| Intercept (18mo) | 31.00 | 0.48 | 65 | <.001** |
| Separation Anxiety Slope (t-1) | 0.09 | 0.05 | 1.81 | .07+ |
| Dependent Stress Slope (t-1) | −0.08 | 0.03 | −2.82 | .005** |
| Time Slope (t-1) | 0.20 | 0.09 | 2.19 | .03* |
Note. Repeated stress and anxiety scores were entered on Level 1 of the model. Age, gender, and initial depression were included as covariates on Level 2. The stress slopes (i.e., test-retest paths) reflect the association between baseline and change over time. Negative stress slopes indicate that those who reported higher levels at baseline have lower change across time.; t-1 = time minus 1, i.e., BL thru 15 months; t = time, i.e., 3 thru 18 months; BL = baseline
p < .05
p < .01
Nonacceptance.
Level of emotion nonacceptance significantly predicted greater number and frequency of stressful events at later time points, controlling for stressful events and anxiety at the previous time points and initial depressive symptoms (see Table 4). However, level of nonacceptance did not moderate the relationship between anxiety and stressors.
Goal-directed behavior.
Difficulty with goal-directed behavior when upset significantly predicted a greater number and frequency of subsequent stressors (see Table 4), after controlling for stressors anxiety at the previous time points and initial depressive symptoms. Goal-directed behavior did not moderate the relationship between any anxiety symptoms and stressful events.
Discussion
This is the first study, to our knowledge, to prospectively examine and clarify the bi-directional relationship between specific types of stressors and specific clusters of anxiety symptoms in a longitudinal follow-up of a large sample of community youth. The results of the current study support the stress causation model; that is, stressful events predicted greater subsequent levels of social and physical anxiety symptoms, even after controlling for prior levels of anxiety symptoms, over a 1.5-year period. This relationship was driven primarily by dependent interpersonal stressors, rather than dependent achievement stressors or independent stressors. The current findings did not support the stress generation model, as anxiety symptoms did not predict greater number and frequency of subsequent stressors over this period. Nonacceptance and difficulty with goal-directed behavior each predicted higher levels of anxiety symptoms and greater number and frequency of stressful events, though they did not moderate the relationship between anxiety and stress.
Our findings in support of the stress causation model are consistent with previous research supporting stress causation in anxiety in youth (e.g., Barrocas & Hankin, 2011; Hamilton et al., 2016; Hankin, 2008; Kopala-Sibley et al., 2015). Specifically, interpersonal dependent stressors predicted subsequent increases in social anxiety and physical anxiety symptoms. Findings add specificity to prior literature that has demonstrated that stressful events, broadly defined, predict subsequent anxiety in children (e.g., Allen et al., 2008). This extends prior work that has found that interpersonal stressors are concurrently associated with social anxiety during adolescence (e.g., Hamilton et al., 2016) by prospectively examining this link in a large youth sample. It is also consistent with the previous finding that interpersonal hassles were associated with changes in anxiety symptoms (Hamilton et al., 2016; Kopala-Sibley et al., 2015). This result further parallels the findings in the depression literature that interpersonal stressors are most predictive of later depression among youth (e.g., Hammen, 2009; Hankin et al., 2007; Rudolph et al., 2000; Sheets & Craighead, 2014). Interpersonal relationships gain importance in adolescence (e.g., Laursen, 1996); therefore, interpersonal stressors may be particularly relevant to the development of anxiety and depression.
In contrast, interpersonal stressors did not predict separation anxiety symptoms. As separation anxiety involves anxiety about separation from an attachment figure, interpersonal stressors with non-attachment figures (i.e., the majority of the reported interpersonal stressors) may not play a role in the development and maintenance of separation anxiety.
Independent stressors did not predict anxiety symptoms. This finding is consistent with related work that suggests independent stressors have little impact on anxiety sensitivity (Zavos et al., 2012). While some theory suggests uncontrollable and unpredictable (i.e., independent) stressors lead to greater anxiety symptoms in youth relative to controllable (e.g., dependent interpersonal) stressors (Chorpita & Barlow, 1998), it is possible that dependent interpersonal stressors may feel uncontrollable to adolescents, as they are often complex and involve other individuals. Some research indicates that perceived stressor controllability may be more important than actual controllability (Compas et al., 1988). Future work should therefore examine youths’ perceptions of stressor controllability, as well as the match between coping response and event controllability (Clarke, 2006).
Interestingly, the current study did not find evidence for the stress generation model in any domain of anxiety. Existing evidence for stress generation in anxiety is mixed; some studies find that anxiety symptoms lead to more subsequent stressors among youth (Shapero et al., 2013; Uliaszek et al., 2012), while others do not find evidence of this (Barrocas & Hankin, 2011),. The lack of evidence for stress generation in the current study is particularly notable given its contrast to the depression literature, in which studies have consistently shown that symptoms of depression contribute to an exacerbation of life stressors in youth (Hammen, 2006; Liu & Alloy, 2010; March-Llanes et al., 2017). It is possible that there are specific factors contributing to the generation of stressful events that are less commonly found among youth with anxiety than among youth with depression (e.g., hopelessness; Joiner et al., 2005). Thus, it appears that only depression, but not anxiety symptoms, consistently predict subsequent stressors among youth.
Finally, nonacceptance and difficulty with goal-directed behavior while upset predicted greater subsequent anxiety. This extends previous findings that nonacceptance and goal-directed difficulties predict the concurrent relationship with and prospective development of anxiety symptoms over time among youth (Schneider et al., 2016). Neither form of emotion dysregulation moderated the relationship between stressors and anxiety, suggesting that stressful events lead to an increase in anxiety over time regardless of the inability to accept emotions in the face of these stressors or to engage in goal-directed behavior when upset.
Importantly, greater nonacceptance and difficulty with goal-directed behavior also prospectively predicted greater number and frequency of stressors among youth, a novel finding suggesting that difficulties with emotion regulation can contribute to stressful events, above and beyond the contributions of previous levels of stressors, anxiety, and depression symptoms. As emotion dysregulation factors independently contributed to both anxiety symptoms and stressors over time, these factors represent promising targets for prevention and intervention strategies for youth (e.g., acceptance and commitment therapy, Hayes et al., 1999; dialectical behavior therapy, Linehan, 1993; unified protocol for youth, Ehrenreich et al., 2009). Targeting these emotion dysregulation factors may directly reduce anxiety, and could also decrease the number or frequency of stressors generated as a result of emotion dysregulation, which in turn may reduce anxiety symptoms. Results of the current study suggest that interventions targeting emotion dysregulation succeed both by creating symptom change and by reducing stressors.
Limitations and Future Directions
To our knowledge, this is the first study to prospectively examine the bi-directional relationship between specific types of stressors and specific clusters of anxiety symptoms in a longitudinal follow-up of a large sample of community youth. Ours is also the first study to elucidate what role emotion dysregulation might play in this relationship. However, these findings should be considered in light of several limitations. First, the ALEQ contains relatively fewer achievement stress items in comparison to interpersonal and independent stressors, and thus may not have fully captured achievement stress. Future studies should examine the relationship between stressful events and anxiety using additional methods for capturing a broader array of stressors and their impact on one’s life, such as with life stress interviews (e.g., Duggal et al., 2000). In addition, the ALEQ uses weighted items based on their frequency, rather than the subjective impact of each individual stressor. It is possible that stressor frequency varies less within subjects over time, and therefore measures that account for stressor impact may be important to evaluate in future studies. It is also possible that stressor impact, as opposed to stressor frequency, is more predictive of subsequent anxiety symptoms. Since anxiety symptoms were evaluated in a community sample using a self-report measure, it also remains important to examine whether the current findings hold in a clinical sample of youth with anxiety disorders. Finally, the current study focused on the impact of two very specific forms of emotion dysregulation. Future studies could examine the impact of a broader array of emotion regulation strategies (e.g., suppression, reappraisal, rumination, problem-solving), as well as the role of temperament and personality traits (e.g., conscientiousness, openness, neuroticism), which have been linked to the use of specific emotion regulation strategies (Gresham & Gullone, 2012).
Conclusion
The present study evaluated the stress causation and stress generation hypotheses for anxiety symptoms in a large longitudinal sample of youth. Findings supported the stress causation hypothesis; stressful events predicted increases in physical and social anxiety symptoms among youth over time. Interpersonal stressors, rather than independent stressors or achievement-related stressors, drove this association. Although emotion dysregulation did not moderate this link, it predicted subsequent anxiety and stressors, suggesting that emotion dysregulation uniquely contributes to the development of anxiety and stress. Together, these findings suggest that increased interpersonal stressors create vulnerability for increased anxiety symptoms among youth, elucidating a window for prevention and intervention efforts.
References
- Allen JL, Rapee RM, & Sandberg S (2008). Severe life events and chronic adversities as antecedents to anxiety in children: A matched control study. Journal of Abnormal Child Psychology, 36(7), 1047 10.1007/s10802-008-9240-x [DOI] [PubMed] [Google Scholar]
- Asselmann E, & Beesdo-Baum K (2015). Predictors of the course of anxiety disorders in adolescents and young adults. Current Psychiatry Reports, 17(2), 1–8. 10.1007/s11920-014-0543-z [DOI] [PubMed] [Google Scholar]
- Baldwin JS, & Dadds MR (2007). Reliability and validity of parent and child versions of the multidimensional anxiety scale for children in community samples. Journal of the American Academy of Child & Adolescent Psychiatry, 46(2), 252–260. 10.1097/01.chi.0000246065.93200.a1 [DOI] [PubMed] [Google Scholar]
- Barrocas AL, & Hankin BL (2011). Developmental pathways to depressive symptoms in adolescence: A multi-wave prospective study of negative emotionality, stressors, and anxiety. Journal of Abnormal Child Psychology, 39(4), 489–500. 10.1007/s10802-010-9482-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brown S, Fite PJ, & Poquiz J (2016). Moderating effects of gender on outcomes associated with stressful life events among elementary school-age youth. Child Psychiatry & Human Development, 47(4), 593–602. 10.1007/s10578-015-0592-5 [DOI] [PubMed] [Google Scholar]
- Chorpita BF, & Barlow DH (1998). The development of anxiety: the role of control in the early environment. Psychological Bulletin, 124(1), 3 10.1037/0033-2909.124.1.3 [DOI] [PubMed] [Google Scholar]
- Clarke AT (2006). Coping with interpersonal stress and psychosocial health among children and adolescents: A meta-analysis. Journal of Youth and Adolescence, 35(1), 10–23. 10.1007/s10964-005-9001-x [DOI] [Google Scholar]
- Cole DA, Nolen-Hoeksema S, Girgus J, & Paul G (2006). Stress exposure and stress generation in child and adolescent depression: A latent trait-state-error approach to longitudinal analyses. Journal of Abnormal Psychology, 115(1), 40 10.1037/0021-843X.115.1.40 [DOI] [PubMed] [Google Scholar]
- Compas BE, Malcarne VL, & Fondacaro KM (1988). Coping with stressful events in older children and young adolescents. Journal of Consulting and Clinical Psychology, 56(3), 405 10.1037/0022-006X.56.3.405 [DOI] [PubMed] [Google Scholar]
- Connolly NP, Eberhart NK, Hammen CL, & Brennan PA (2010). Specificity of stress generation: A comparison of adolescents with depressive, anxiety, and comorbid diagnoses. International Journal of Cognitive Therapy, 3(4), 368–379. 10.1521/ijct.2010.3.4.368 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Duggal S, Malkoff-Schwartz S, Birmaher B, Anderson BP, Matty MK, Houck PR, … & Frank E (2000). Assessment of life stress in adolescents: Self-report versus interview methods. Journal of the American Academy of Child & Adolescent Psychiatry, 39(4), 445–452. 10.1097/00004583-200004000-00013 [DOI] [PubMed] [Google Scholar]
- Ehrenreich JT, Goldstein CR, Wright LR, & Barlow DH (2009). Development of a unified protocol for the treatment of emotional disorders in youth. Child & Family Behavior Therapy, 31(1), 20–37. 10.1080/07317100802701228 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Grant KE, & McMahon SD (2005). Conceptualizing the role of stressors in the development of psychopathology. Development of psychopathology: A vulnerability-stress perspective, 3–31. 10.4135/9781452231655 [DOI] [Google Scholar]
- Grant KE, McMahon SD, Carter JS, Carleton RA, Adam EK, & Chen E (2014). The influence of stressors on the development of psychopathology In Handbook of Developmental Psychopathology (pp. 205–223). Springer, Boston, MA: 10.1007/978-1-4614-9608-3_11 [DOI] [Google Scholar]
- Gratz KL, & Roemer L (2004). Multidimensional assessment of emotion regulation and dysregulation: Development, factor structure, and initial validation of the difficulties in emotion regulation scale. Journal of Psychopathology and Behavioral Assessment, 26(1), 41–54. 10.1023/B:JOBA.0000007455.08539.94 [DOI] [Google Scholar]
- Gresham D, & Gullone E (2012). Emotion regulation strategy use in children and adolescents: The explanatory roles of personality and attachment. Personality and Individual Differences, 52(5), 616–621. 10.1016/j.paid.2011.12.016 [DOI] [Google Scholar]
- Hamaker EL, Kuiper RM, & Grasman RP (2015). A critique of the cross-lagged panel model. Psychological Methods, 20(1), 102 10.1037/a0038889 [DOI] [PubMed] [Google Scholar]
- Hamilton JL, Potter CM, Olino TM, Abramson LY, Heimberg RG, & Alloy LB (2016). The temporal sequence of social anxiety and depressive symptoms following interpersonal stressors during adolescence. Journal of Abnormal Child Psychology, 44(3), 495–509. 10.1007/s10802-015-0049-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hammen C (1991). Generation of stress in the course of unipolar depression. Journal of Abnormal Psychology, 100(4), 555 10.1037/0021-843X.100.4.555 [DOI] [PubMed] [Google Scholar]
- Hammen C (2005). Stress and depression. Annual Review of Clinical Psychology, 1, 293–319. 10.1146/annurev.clinpsy.1.102803.143938 [DOI] [PubMed] [Google Scholar]
- Hammen C (2006). Stress generation in depression: Reflections on origins, research, and future directions. Journal of Clinical Psychology, 62(9), 1065–1082. 10.1002/jclp.20293 [DOI] [PubMed] [Google Scholar]
- Hammen C (2009). Adolescent depression: Stressful interpersonal contexts and risk for recurrence. Current Directions in Psychological Science, 18(4), 200–204. 10.1111/j.1467-8721.2009.01636.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hankin BL, & Abramson LY (2002). Measuring cognitive vulnerability to depression in adolescence: Reliability, validity, and gender differences. Journal of Clinical Child and Adolescent Psychology, 31(4), 491–504. 10.1207/S15374424JCCP3104_8 [DOI] [PubMed] [Google Scholar]
- Hankin BL (2008). Cognitive vulnerability–stress model of depression during adolescence: Investigating depressive symptom specificity in a multi-wave prospective study. Journal of Abnormal Child Psychology, 36(7), 999–1014. 10.1007/s10802-008-9228-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hankin BL (2009). Development of sex differences in depressive and co-occurring anxious symptoms during adolescence: Descriptive trajectories and potential explanations in a multiwave prospective study. Journal of Clinical Child & Adolescent Psychology, 38(4), 460–472. 10.1080/15374410902976288 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hankin BL, Mermelstein R, & Roesch L (2007). Sex differences in adolescent depression: Stress exposure and reactivity models. Child development, 78(1), 279–295. 10.1111/j.1467-8624.2007.00997.x [DOI] [PubMed] [Google Scholar]
- Hankin BL, Stone L, & Wright PA (2010). Corumination, interpersonal stress generation, and internalizing symptoms: Accumulating effects and transactional influences in a multiwave study of adolescents. Development and Psychopathology, 22(1), 217–235. 10.1017/S0954579409990368 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hankin BL, Young JF, Abela JR, Smolen A, Jenness JL, Gulley LD, … & Oppenheimer CW (2015). Depression from childhood into late adolescence: Influence of gender, development, genetic susceptibility, and peer stress. Journal of Abnormal Psychology, 124(4), 803 10.1037/abn0000089 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hayes SC, Strosahl KD, & Wilson KG (1999). Acceptance and commitment therapy: An experiential approach to behavior change. Guilford Press. [Google Scholar]
- Joiner TE, Wingate LR, Gencoz T, & Gencoz F (2005). Stress generation in depression: Three studies on its resilience, possible mechanism, and symptom specificity. Journal of Social and Clinical Psychology, 24(2), 236–253. 10.1521/jscp.24.2.236.62272 [DOI] [Google Scholar]
- Kopala-Sibley DC, Zuroff DC, Hankin BL, & Abela JR (2015). The development of self-criticism and dependency in early adolescence and their role in the development of depressive and anxiety symptoms. Personality and Social Psychology Bulletin, 41(8), 1094–1109. 10.1177/0146167215590985 [DOI] [PubMed] [Google Scholar]
- Kovacs M (1985): The Children’s Depression Inventory (CDI). Psychopharmacology Bulletin, 21:995–998. [PubMed] [Google Scholar]
- Laursen B (1996). Closeness and conflict in adolescent peer relationships: Interdependence with friends and romantic partners. The company they keep: Friendship in childhood and adolescence, 186–210. [Google Scholar]
- Linehan M (1993). Cognitive-behavioral treatment of borderline personality disorder. Guilford press. [Google Scholar]
- Liu RT, & Alloy LB (2010). Stress generation in depression: A systematic review of the empirical literature and recommendations for future study. Clinical Psychology Review, 30(5), 582–593. 10.1016/j.cpr.2010.04.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- MacCallum RC, Kim C, Malarkey WB, & Kiecolt-Glaser JK (1997). Studying multivariate change using multilevel models and latent curve models. Multivariate Behavioral Research, 32(3), 215–253. 10.1207/s15327906mbr3203_1 [DOI] [PubMed] [Google Scholar]
- March JS, Parker JD, Sullivan K, Stallings P, & Conners CK (1997). The Multidimensional Anxiety Scale for Children (MASC): factor structure, reliability, and validity. Journal of the American Academy of Child & Adolescent Psychiatry, 36(4), 554–565. 10.1097/00004583-199704000-00019 [DOI] [PubMed] [Google Scholar]
- March-Llanes J, Marqués-Feixa L, Mezquita L, Fañanás L, & Moya-Higueras J (2017). Stressful life events during adolescence and risk for externalizing and internalizing psychopathology: a meta-analysis. European Child & Adolescent Psychiatry, 26(12), 1409–1422. 10.1007/s00787-017-0996-9 [DOI] [PubMed] [Google Scholar]
- McMahon SD, Grant KE, Compas BE, Thurm AE, & Ey S (2003). Stress and psychopathology in children and adolescents: is there evidence of specificity?. Journal of Child Psychology and Psychiatry, 44(1), 107–133. 10.1111/1469-7610.00105 [DOI] [PubMed] [Google Scholar]
- Moksnes UK, Moljord IE, Espnes GA, & Byrne DG (2010). The association between stress and emotional states in adolescents: The role of gender and self-esteem. Personality and Individual Differences, 49(5), 430–435. 10.1016/j.paid.2010.04.012 [DOI] [Google Scholar]
- Moya-Higueras J, Cuevas A, Marques-Feixa L, Mezquita L, Mayoral M, Fañanás L, … Ibáñez MI (2018). Recent Stressful Life Events (SLE) and Adolescent Mental Health: Initial Validation of the LEIA, a New Checklist for SLE Assessment According to Their Severity, Interpersonal, and Dependent Nature. Assessment, 107319111881764 10.1177/1073191118817648 [DOI] [PubMed] [Google Scholar]
- Neumann A, van Lier PA, Gratz KL, & Koot HM (2010). Multidimensional assessment of emotion regulation difficulties in adolescents using the difficulties in emotion regulation scale. Assessment, 17(1), 138–149. 10.1177/1073191109349579 [DOI] [PubMed] [Google Scholar]
- Olson SL, Schilling EM, & Bates JE (1999). Measurement of impulsivity: Construct coherence, longitudinal stability, and relationship with externalizing problems in middle childhood and adolescence. Journal of Abnormal Child Psychology, 27(2), 151–165. 10.1023/A:1021915615677 [DOI] [PubMed] [Google Scholar]
- Raudenbush SW, Bryk AS, & Congdon R (2013). HLM 7.01 for Windows [Computer software]. Skokie, IL: Scientific Software International, Inc. [Google Scholar]
- Rudolph KD (2008). Developmental influences on interpersonal stress generation in depressed youth. Journal of Abnormal Psychology, 117(3), 673 10.1037/0021-843X.117.3.673 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rudolph KD, Hammen C, Burge D, Lindberg N, Herzberg D, & Daley SE (2000). Toward an interpersonal life-stress model of depression: the developmental context of stress generation. Development and Psychopathology, 12(2), 215 10.1017/S0954579400002066 [DOI] [PubMed] [Google Scholar]
- Schneider RL, Arch JJ, Landy LN, & Hankin BL (2016). The Longitudinal Effect of Emotion Regulation Strategies on Anxiety Levels in Children and Adolescents. Journal of Clinical Child & Adolescent Psychology, 1–14. 10.1080/15374416.2016.1157757 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schuurman NK, Ferrer E, de Boer-Sonnenschein M, & Hamaker EL (2016). How to compare cross-lagged associations in a multilevel autoregressive model. Psychological Methods, 21(2), 206 10.1037/met0000062 [DOI] [PubMed] [Google Scholar]
- Shapero BG, Hankin BL, & Barrocas AL (2013). Stress generation and exposure in a multi-wave study of adolescents: Transactional processes and sex differences. Journal of Social and Clinical Psychology, 32(9), 989–1012. 10.1521/jscp.2013.32.9.989 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sheets E, & Craighead W (2014). Comparing chronic interpersonal and noninterpersonal stress domains as predictors of depression recurrence in emerging adults. Behaviour Research and Therapy, 63, 36–42. 10.1016/j.brat.2014.09.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Uliaszek AA, Zinbarg RE, Mineka S, Craske MG, Sutton JM, Griffith JW, … & Hammen C (2010). The role of neuroticism and extraversion in the stress–anxiety and stress–depression relationships. Anxiety, Stress, & Coping, 23(4), 363–381. 10.1080/10615800903377264 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Uliaszek AA, Zinbarg RE, Mineka S, Craske MG, Griffith JW, Sutton JM, … & Hammen C (2012). A longitudinal examination of stress generation in depressive and anxiety disorders. Journal of Abnormal Psychology, 121(1), 4 10.1037/a0025835 [DOI] [PubMed] [Google Scholar]
- van Gastel W, & Ferdinand RF (2008). Screening capacity of the Multidimensional Anxiety Scale for Children (MASC) for DSM‐IV anxiety disorders. Depression and Anxiety, 25(12), 1046–1052. 10.1002/da.20452 [DOI] [PubMed] [Google Scholar]
- Vrshek-Schallhorn S, Stroud CB, Mineka S, Hammen C, Zinbarg RE, Wolitzky-Taylor K, & Craske MG (2015). Chronic and episodic interpersonal stress as statistically unique predictors of depression in two samples of emerging adults. Journal of Abnormal Psychology, 124(4), 918–932. 10.1037/abn0000088 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wei C, Hoff A, Villabø MA, Peterman J, Kendall PC, Piacentini J, … & Sherrill J (2014). Assessing anxiety in youth with the multidimensional anxiety scale for children. Journal of Clinical Child & Adolescent Psychology, 43(4), 566–578. 10.1080/15374416.2013.814541 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zavos H, Wong CC, Barclay NL, Keers R, Mill J, Rijsdijk FV, … & Eley TC (2012). Anxiety sensitivity in adolescence and young adulthood: the role of stressful life events, 5HTTLPR and their interaction. Depression and Anxiety, 29(5), 400–408. 10.1002/da.21921 [DOI] [PubMed] [Google Scholar]