Tracking Rumination as a Stable Habit (TRASH);Scale Modification and Convergent Validity in a Clinical Sample of Youth With a History of Depression

Abstract

Rumination is a cognitive habit involving repetitive negative thinking (RNT). It is a risk factor for depression and develops during late childhood. Targeted interventions to reduce rumination are critical for reducing risk for and impact of depression. Therefore, quick and sensitive scales are needed to capture concrete, quantifiable changes in rumination habit. We developed and evaluated the Tracking Rumination As a Stable Habit (TRASH) measure, to align with Rumination-Focused Cognitive Behavioral Therapy (RF-CBT). The scale was derived, modified and expanded on from the Habit Index of Negative Thinking (HINT). TRASH assesses automaticity and consistency of the habit, quantifiable elements of the daily habit. As part of a larger trial of RF-CBT, 120 adolescents (age mean [sd]: 15.73 [1.11], 67.5% female) completed all scales. Internal reliability of the scale was high (overall α = 0.86, Automaticity α = 0.85, Consistency α = 0.72). The scale demonstrated moderate to strong convergent validity with existing measures of RNT (rs [118] = 0.43 to 0.46). Associations with anxiety, worry and depression measures were modest and variable (rs [118] = 0.13 to 0.63). Divergent and/or nonrelated validity with scales, included null and negative relationships (rs [118] = −0.23 to 0.39). Exploratory analyses identified a general RNT factor (including worry, rumination), a TRASH subfactor, and distinct affective RNT subfactors (e.g., angry, apathetic). Treatment by time interactions suggest similar changes over time in RNT scales, as well as the RNT main factor and the TRASH subfactor. Additional research is needed for the TRASH scale, if it may be sensitive to concrete changes in rumination. ClinicalTrials.gov Identifier: NCT03859297.

Keywords

rumination repetitive negative thinking scale validation factor analysis

Background

Rumination is a habitual cognitive process characterized by repetitive, passive self-focus on negative emotions and experiences. Rumination has been most extensively researched in the context of depression, guided by the seminal Response Styles Theory, which proposed that rumination is a stable and habitual tendency to dwell on negative information during sad or depressed mood (Nolen-Hoeksema, 1991; Nolen-Hoeksema et al., 2008; Nolen-Hoeksema & Morrow, 1993; Smith & Alloy, 2009; Watkins & Nolen-Hoeksema, 2014). There is now substantial evidence that rumination is a transdiagnostic process that confers increased risk for, and maintenance of, multiple mental health diagnoses (Ehring & Watkins, 2008). In a theoretical integration of existing theory and evidence, Watkins and Roberts’ (2020) H-EX-A-GO-N model proposes that rumination develops as a habitual response (H) to low mood and outlines the role of impairments in the application of executive control (EX), abstract processing (A), and negative biases (N) in susceptibility to rumination, often in the face of goal discrepancies (GO). H-EX-A-GO-N extends Response Styles Theory and Control Theory accounts of rumination to outline distal and proximal mechanisms underlying rumination. Central to both Response Styles Theory and H-EX-A-GO-N is the conceptualization of maladaptive rumination as learned habitual response to stressful experiences that is characterized by more abstract, negative, and passive thinking and behavior. Whilst rumination initially occurs as a normal, goal-directed response to problems and difficulties (goal discrepancies; adjunctive to what may be naturally observed as problem solving; Watkins & Roberts, 2020), repeated pairing of negative emotional states and rumination can lead to this becoming a learned response style (Watkins & Nolen-Hoeksema, 2014). Reinforcement of rumination occurs through avoidance of unpleasant experiences, provision of motivation to complete tasks, and/or perceived contribution to understanding and planning. Frequent and prolonged periods of exposure to stress and low mood can result in repeated rehearsal and consolidation of rumination, producing a learned habitual stimulus-response pairing early in development (Grierson et al., 2016; Shaw et al., 2019). This learned response may contribute to the onset, duration, and recurrence of depression, sustained passivity in action, and an abstract processing style (Smith & Alloy, 2009). Habitual rumination can be difficult to control because it is cued automatically, may be linked to certain locations, internal states and contexts, and it is difficult to target clinically because of its negatively reinforcing consequences.

Rumination is an important treatment target for individuals at risk for depression and its relapse (Just & Alloy, 1997). In individuals with remitted major depressive disorder (MDD), elevated rumination levels predict MDD recurrence (Figueroa et al., 2019). Targeting rumination directly as a maladaptive coping behavior in teens, such as via rumination-focused cognitive behavioral therapy (RF-CBT), may reduce depression recurrence (Bessette et al., 2020; Cook et al., 2019; Jacobs et al., 2016; Topper et al., 2017). Rumination and worry are conceptually and empirically similar RNT cognitive processes (McEvoy et al., 2010); the most frequently used and best validated scales (e.g., Ruminative Response Scale of the Response Styles Questionnaire; Penn State Worry Questionnaire) involve abstract and negative repetitive thought and exacerbate anxiety and sad mood, with potential areas of distinction noted in relation to function (regret vs threat anticipation) and temporal relation (past vs future) (Gustavson et al., 2018; Muris et al., 2004; Stade & Ruscio, 2023). Clinical research demonstrates that rumination and worry are highly correlated, suggesting that rumination and worry have many shared clinical elements for therapeutic targeting (de Jong-Meyer et al., 2009; Fresco et al., 2002; Kalmbach et al., 2016; Kircanski et al., 2018; McEvoy et al., 2013; Puccetti et al., 2025; Stade & Ruscio, 2023).

Repetitive negative thinking (RNT) describes the shared transdiagnostic features of rumination and worry, recognizing these as key risk factors for psychopathology (Ehring & Watkins, 2008). Our H-EX-A-GO-N model of habit-based RNT will benefit from development of specific tools for measuring RNT in daily life. For example, rumination is set in motion not only by stressful experiences, but also by habit-based features like time of day, for example during a morning coffee or in bed at night before sleep. There may be sensory (e.g., darkness and bed covers) or contextual features (e.g., being alone, parents arguing down the hall, social comparison when viewing media). For our purposes, within a habit-framed treatment model of RNT, we focus on the duration, intensity, controllability, and the function(s) of RNT as key variables to be measured given their implications for assessment and treatment (Hoyer et al., 2009; Raes, 2010). For example, these are concrete elements of experience that could be perceivable and are integrated into the diary and therapy parts of RF-CBT. Collection of these measures enables to move beyond “tendency” and “self-described” experiences of rumination to concrete elements that can be changed. Moreover, these in turn can allow for elaboration and nuance within the H-EX-A-GO-N model of RNT.

Despite a strong and consistent link to transdiagnostic clinical disposition and outcomes, rumination is measured infrequently in clinical settings (Baik & Newman, 2023; Ehring, 2021; Kennedy et al., 2024). The most widely used measures of rumination in research focus predominantly on content and frequency of the habitual response style (e.g., Ruminative Responses Scale [RRS]; Nolen-Hoeksema & Morrow, 1991; Treynor et al., 2003), identifying reflective and brooding subtypes. The measure also has a depression subscale, and all three subscales have a modest relationship with depressive symptoms (Nolen-Hoeksema et al., 2008). Brooding rumination is more strongly implicated in transdiagnostic psychopathology relative to reflective rumination (McEvoy et al., 2013; Shaw et al., 2021; Treynor et al., 2003). More recently, scales have been developed to capture specific emotional triggers for rumination like sadness and anger (e.g., Sadness and Angry Rumination Inventory [SARI]; Peled & Moretti, 2010), how individuals think about negative experiences and problems (Bijttebier et al., 2015; Ehring et al., 2011), or frequency and awareness of negative self-thinking (Habit Index of Negative Thinking (HINT); Verplanken et al., 2007).

There are important conceptual distinctions between how a therapist could use rumination scales to try and measure changes in thought content, processing styles, habitual characteristics, in comparison to more tractable, concrete metrics like frequency, intensity, controllability, and situational characteristics (Colvin et al., 2021). For example, whilst the RRS scale used to measure rumination has been highly valuable for developing our understanding of rumination, it is difficult to evaluate week-to-week changes with a scale that is designed to evaluate rumination as a habit, in general, and does not have as many quantifiable elements. Notably, within rumination-focused CBT, there is a therapeutic focus on increasing awareness of rumination, including context and controllability, and practicing alternative, more effective strategies to weaken or replace the ruminative habit (Watkins, 2016). At a theoretical level, key habitual aspects of rumination include the automaticity of rumination (degree of control, how easily is the habit triggered, the initiation of thoughts without awareness or intention) and its consistency (does it occur in similar situations and times, in response to the same stimuli or cues, even in the absence of immediate stressors) (Wood & Neal, 2007). Reliable indices to quantify these characteristics would be valuable to support assessment and ongoing monitoring of the habitual aspects of rumination and to improve their targeting in treatment. In early work, automaticity of negative thought from the HINT scale shows improved predictions of mental health outcomes compared to frequency of RNT (Verplanken et al., 2007). As such, the use of scales, new and existing that tie more concretely to the goals, intentions and strategies of the therapy would aid both patients and therapists in effective delivery of the intervention.

To more sensitively assess the strength of the habitual rumination in everyday life and its change following targeted treatment, our group created a new scale for use within a randomized clinical trial evaluating RF-CBT by expanding an existing measure of rumination automaticity (Verplanken et al., 2007). Our revised scale (Tracking Rumination as a Stable Habit; TRASH) preserves items relevant to automaticity and adds items intended to assess cue-response consistency. This study examines the psychometrics of TRASH in a clinical sample of adolescents with remitted MDD (rMDD). We hypothesized that the TRASH subscales would have moderate to strong internal reliability and converge with existing measures of rumination and clinical scales of anxiety, depression, and worry. In addition, while we expected strong convergence with existing measures of rumination, we also evaluated whether TRASH assesses unique aspects of rumination not captured in other scales. Moreover, visual analog scales for rumination (rumVAS) were used to determine if we can more directly assess quantifiable changes in frequency, distress, duration, and control of ruminative thought. We examine the relationships between TRASH and impulsivity, distractibility, avoidance, and inhibition, to test conceptual divergence of these constructs that are related to clinical depression but are not part of RNT. These are constructs (e.g., impulsivity, distractibility, avoidance, and inhibition) that are aligned with the experience of an internalizing disorder like depression. These constructs can even be related to rumination (e.g., avoidance) for some, but not all, individuals who engage in habitual rumination. We included these constructs and measures to look at specific aspects of change in rumination with RF-CBT, not more general changes or adaptative behaviors. We hypothesized that these behavioral indices associated with depression may be elevated in individuals who experience mental health challenges, but that they should not relate specifically to RNT.

Method

Participants and Procedures

Baseline data were analyzed for 120 participants recruited for the R61 and R33 phases of a pre-registered RCT examining RF-CBT for adolescents with rMDD and elevated rumination (Langenecker et al., 2024; Roberts et al., 2021). Adolescents were recruited from the Salt Lake City, UT and Columbus, OH areas through radio and community advertisements. Eligible participants were ages 14 to 18 with a history of MDD, with at least partial remission for two weeks. Eligible participants also had RRS scores above a T score of 50 by age and sex. All participants who completed baseline assessments were included in factor analyses, including those who were ineligible for the trial (NCT03859297). Within the trial (described in Roberts et al., 2021), the R61 included randomization to RF-CBT or treatment as usual (Langenecker et al., 2024), and the R33 phase randomization is to RF-CBT or Relaxation therapy (as in Watkins et al., 2009). In a sub analysis to respond to reviewer requests, MBT (only) used unblinded data up through Jan 2026 to evaluate scale by treatment interactions for 55 youth enrolled in the trial with TRASH data before and after treatment. Informed consent/assent was obtained for all participants in line with local institutional review board requirements and the Declaration of Helsinki.

Measures

Rumination Response Scale (RRS) is a 22-item self-report measure that assesses ruminative response tendencies to feelings of sadness and depression, including brooding and reflection (Nolen-Hoeksema & Morrow, 1991; Treynor et al., 2003). The RRS had strong internal consistency (α = 0.93) in a sample of adolescents (Liang & Lee, 2019). Scores were computed by summation of responses for all items, and subsets of items for each subscale.

Sadness and Anger Rumination Inventory (SARI) is a 22-item self-report questionnaire that measures ruminations about one’s anger (SARI anger) or sadness (SARI sadness). The original 22-item version demonstrated excellent internal consistency in adolescents (sadness α = 0.95, anger α = 0.96; (Peled & Moretti, 2007). For our analyses, we used the recommended 19-item version (Peled & Moretti, 2010). Scores were sums of items from each respective subscale.

Behavioral Inhibition Scale (BIS) is a 7-item scale (sum of scores) that measures trait-like responses to avoid negative consequences (Carver & White, 1994; Figueroa et al., 2019), a subscale of the larger Behavioral Inhibition System and Behavioral Activation System (BIS/BAS) questionnaire which measures sensitivity to punishment, non-reward, and novel stimuli. The BIS demonstrated good internal consistency in young adults (α = 0.75; Carver & White, 1994).

Impulsive Behaviors Scale for Children (UPPS-P-C) is a 40-item self-report measure that assesses 5 factors of impulsivity: negative urgency, premeditation, perseverance, sensation seeking, and positive urgency (Zapolski et al., 2010). In an adolescent sample, internal consistency for each factor ranged from α = 0.79 to α = 0.95 (Tomko et al., 2016). Scoring followed the standard procedure of summating the responses and reversed responses for items in each subscale.

Screen for Child Anxiety Related Disorder (SCARED) is a 41-item self-report measure of anxiety, including panic and associated somatic, generalized worries, separation, social, and school refusal anxiety symptoms (Birmaher et al., 1997, 1999). Internal consistency of SCARED subscales ranged from good to excellent in an adolescent sample, with α ranging from 0.74 to 0.94 (Birmaher et al., 1997). Scores were sums of items from each respective subscale.

Penn State Worry Questionnaire for Children (PSWQ-C) is a 14-item self-report measure of worry in children and adolescents (Chorpita et al., 1997) adapted from the adult-based PSWQ. It assesses generality, intensity, and uncontrollability of worry (Păsărelu et al., 2017). The PSWQ-C demonstrated excellent internal consistency in a clinical adolescent sample (α = 0.91; Pestle et al., 2008). Scores were calculated by summing all items, with items 2, 7, and 9 reversed.

Rumination Habit Visual Analog Scale (rumVAS) is a 4-item self-report measure that we developed for quantification of elements of the experience of the rumination habit through a 10-point visual analog scale (scores are on the 10 point scale), modified from a previously published scale that examines rumination frequency and distress caused by ruminative behavior (Wichelns et al., 2016). rumVAS assesses the frequency, sense of control, duration, and distress experienced during rumination.

Mind Wandering Deliberate and Spontaneous Scales (MW-D & -S) are each 4-item self-report measures of deliberate and spontaneous mind wandering behaviors (Carriere et al., 2013), where scores are summations of all respective items. In adolescent samples, the MW-D and MW-S demonstrate good internal consistency (α = 0.82 and α = 0.81, respectively; Carciofo & Jiang, 2021).

Multidimensional Experiential Avoidance Questionnaire for Adolescents (MEAQ-A) is a 39-item self-report measure of one’s tendency to avoid negative internal experiences, adapted for adolescents. Subscales assess distress aversion and the use of distraction/suppression and repression/denial to avoid distress (Gámez et al., 2011). The MEAQ-A demonstrates internal consistencies ranging from α = 0.70 to 0.80. Subscale scores are sums of their respective items.

Children’s Depressive Rating Scale Revised (CDRS-R) is a 17-item, clinician-administered scale that combines child and caregiver report in a clinician-determined scale to assess depressive symptoms (Poznanski & Mokros, 1996) and was used to assess past week symptoms for the present study. The CDRS-R demonstrated good internal consistency in adolescent samples, ranging from α = 0.74 to 0.92 (Mayes et al., 2010). Scoring is adjusted weighting, sum of scores from parent and child interviewed ratings.

Treating Rumination as a Stable Habit (TRASH) asks participants to self-report aspects of ruminative habit over the previous two weeks. We devised this as a shorter, modified form of HINT (Verplanken et al., 2007) for repeated measurement within a clinical trial. TRASH was developed to have two subscales with five items each (see Table 3 for all items): Automaticity (happens without awareness, e.g., “I start before I realize I’m doing it”) and Consistency of the ruminative response to the same cues/contexts (similar locations, situations, times of day, e.g., “I do it in the same place(s)”). Responses are assessed on a 5-point Likert scale, and items are summed for sub and total scores. Understanding and increasing awareness of the ruminative habit are early key goals in the RF-CBT therapeutic model before the therapeutic process can effectively target the rumination habit.

Analyses

All analyses were conducted in R version 4.5.2 (R Core Team, 2024). Factor analyses were conducted with the psych package (Revelle, 2024). Missingness structure was evaluated with the naniar package (Tierney & Cook, 2023). A description of missing data handling is included in the supplement.

TRASH Reliability and Factor Structure

Internal reliability in this sample was assessed using Cronbach’s α across all items, and for Automaticity and Consistency subscales. An exploratory factor analysis (EFA) examined the underlying structure of TRASH. Parallel analysis and Bayesian Information Criterion (BIC) identified the best fitting number of factors. Factors were extracted by principal axis factoring due to non-normality of the data and obliquely rotated via the oblimin algorithm as we anticipated correlation between factors. A confirmatory factor analysis was also run with lavaan package (Rosseel, 2012) to examine fit estimates of the hypothesized two factor structure of TRASH. Model fit was evaluated against a saturated model (chi-square) and a null independence model (CFI, TLI).

TRASH Construct Validity

To assess construct validity, we examined correlation coefficients between TRASH and hypothesized convergent and divergent scales. Correlation probability values were adjusted with the Benjamini-Hochberg procedure. For convergent validity, we anticipated a strong positive correlation between TRASH, existing measures of rumination (RRS, SARI), and to a lesser extent clinical scales of anxiety, depression, and worry (SCARED, CDRS-R, PSWQ respectively). For divergent validity, we anticipated weak positive or negative correlations between TRASH and scales related to impulsivity (UPPS), distractibility (MWDS), avoidance (MEAQ), and inhibition (BIS), although our specific research sample with history of depression who are seeking treatment - presents some potential for restriction of range or bias in this regard.

Exploratory Bifactor Analyses

While we expected strong convergence with existing measures of rumination, we also evaluated whether TRASH assesses unique aspects of rumination not captured in other scales (i.e., Automaticity and Consistency). We conducted two exploratory bifactor analyses (EBFA). The first EBFA (narrow) included items from a set of scales specifically measuring rumination and worry (TRASH, RRS, SARI, PSWQ, rumVAS). The second EBFA (broad) included items from all scales used for convergent/unrelated/divergent validity evaluation (TRASH, RRS, SARI, PSWQ, rumVAS, SCARED, CDRS-R, UPPS, MWDS, MEAQ, BIS), the results of which are in the supplement (Table S2). For both analyses, we posited an RNT factor on which all items load and allowed items to load onto additional group factors with oblique rotation. The number of factors to extract was determined by parallel analysis and BIC. Individual loadings on factors were determined through extraction by principal axis factoring using each item’s squared multiple correlations (EBFA 1), or by minimum residual (EBFA 2). Bifactor rotation (Jennrich & Bentler, 2011) using the biquartimin function provided final factor structures. A confirmatory factor analysis modeling each of TRASH automaticity, TRASH consistency, RRS, PSWQ-C, SARI anger, SARI sadness as separate factors provided fit estimates.

Treatment Sensitivity

To evaluate the sensitivity of the TRASH to rumination focused RFCBT relative to existing scales, we conducted analyses (unblinding for MBT only) aimed at elucidating pre-post differences in automaticity, consistency, and closely related RNT constructs (RRS, PSWQ-C). Only participants with all relevant data at both time points were included (n = 55). Items at both pre and post treatment were standardized by calculating z-scores using the mean and standard deviation from pre-treatment data. Factor scores for each time point were computed using pre-treatment factor loadings. Linear mixed-effects models (LMMs) were constructed with the lme4 (Bates et al., 2015) and lmerTest (Kuznetsova et al., 2017) packages, including an interaction effect between treatment group and time point and a random intercept for subject. Post hoc estimates were generated via the emmeans package (Lenth, 2025). Results were visualized with ggplot2 (Wickham, 2016).

Results

Demographics and Clinical Variables

A sample of 120 adolescents was used in the primary analyses (see the supplement for a description of missing data). The sample included a larger proportion of females than males, consistent with epidemiological samples of depression in adolescents. Sample characteristics are included in Table 1. Existing rumination and other related scales demonstrated good to excellent internal reliability in our sample comparable to previous reports (see Table 2).

Table 1.

Baseline Sample Characteristics of the 120 Adolescents

Variable/Scale	Mean (SD)/N (%)
Age	15.73 (1.11)
Sex assigned at birth
Female	81 (67.5 %)
Male	39 (32.5%)
Race
Asian	2 (1.67%)
Black/African American	4 (3.33%)
Native American	1 (0.83%)
Other	5 (4.17%)
Pacific Islander	0 (0.0%)
White	108 (90.0%)
Ethnicity
Hispanic or Latin(e)	12 (10.0%)
Not Hispanic or Latine	108 (90.0%)
RRS total	56.1 (11.22)
TRASH automaticity	20.01 (3.56)
TRASH consistency	18.44 (3.31)
SARI sadness	37.29 (6.18)
SARI anger	35.05 (8.55)
Rumination VAS frequency	58.32 (21.88)
Rumination VAS distress	62.33 (25.28)
Rumination VAS control	58.13 (23.31)
Rumination VAS duration	59.95 (23.05)
CDRS-R total	31.84 (9.12)
SCARED-GAD	12.31 (4.43)

Note. RRS, Rumination Response Scale. TRASH, Tracking Rumination As a Stable Habit. SARI, Sadness and Anger Rumination Inventory. VAS, Visual Analogue Scale. CDRS-R, Children’s Depression Rating Scale Revised. SCARED-GAD, Screen for Child Anxiety Related Disorders-Generalized Anxiety Disorder.

Table 2.

Internal Consistency of Rumination and Related Scales

Scale	Cronbach’s α
RRS total	.91
SARI-19 item	.93
Rumination VAS	.68
CDRS-R total	.90
SCARED	.95
PSWQ-C	.92
UPPS-P-C	.89
MW-D-S	.87
MEAQ-A	.85
BIS	.76

Note. N = 120. RRS, Rumination Response Scale. SARI, Sadness and Anger Rumination Inventory. VAS, Visual Analogue Scale. CDRS-R, Children’s Depression Rating Scale Revised. SCARED, Screen for Child Anxiety Related Disorders. PSWQ-C, Penn State Worry Questionnaire for Children. UPPS-P-C, Impulsive Behaviors Scale for Children. MW-D-S, Mind Wandering Deliberate and Spontaneous Scales. MEAQ-A, Multidimensional Experiential Avoidance Questionnaire for Adolescents. BIS, Behavioral Inhibition Scale.

Internal Consistency and Factor Structure of TRASH

Internal reliability (n = 120) was strong for TRASH overall (α = .86) and the Automaticity subscale (α = .85, ω = .85), and moderate for the Consistency subscale (α = .72, ω = .48). Item-level descriptive statistics are included in Table S1. Parallel analysis and BIC suggested two factors (Table 3). Items hypothesized to capture the automatic and consistent aspects of habitual rumination loaded significantly onto separate factors. All items loaded onto only one factor, except for item 6 (“I do when I’m alone”) which loaded onto both. These factors accounted for 47% of item variance. The factors were moderately correlated (r = .57, p < .001), supporting the use of the oblique rotation. Confirmatory factor analysis using the a priori two factor structure (n = 120) indicated moderately good fit (χ² [34] = 50.7, p = .03, RMSEA = 0.064, SRMR = .055; CFI = 0.96, TLI = 0.95).

Table 3.

TRASH Items and EFA Loadings

Item	Subscale	TRASH item	Automaticity	Consistency
1	Automaticity	I do automatically	0.87
9	Automaticity	I start before I realize I’m doing it	0.69
3	Automaticity	I find it hard not to do	0.75
5	Automaticity	I do unintentionally	0.69
7	Automaticity	That requires mental effort to stop	0.65
6	Consistency	I do when I’m alone	0.41	0.33
2	Consistency	I do in the same place(s)		0.72
4	Consistency	I do when I feel sad, anxious or annoyed		0.32
8	Consistency	I do at the same time(s) of the day		0.58
10	Consistency	I do in response to the same situation(s)		0.51
		Initial eigenvalues	4.5	1.21
		SS loadings	3.16	1.51
		Variance explained	.32	.15

Note. Factor analysis of TRASH items identified automaticity and consistency factors, Accounting for a combined 47% of variance. Items from automaticity and consistency subscales loaded uniquely onto their respective factors, except for item 6 which loaded onto both factors. Only item loadings greater than 0.3 are presented. N = 120.

Convergent and Divergent Validity

TRASH Automaticity and Consistency subscales (n = 118) demonstrated moderate to strong convergent validity with established RNT measures, including RRS (Automaticity r [n = 118] = .43, p < .001; Consistency r [118] = .46, p < .001), SARI sadness (Automaticity r [118] = .46, p < .001; Consistency r [118] = .46, p < .001), PSWQ (Automaticity r [118] = .63, p < .001; Consistency r [118] = .44, p < .001). Weak correlations were observed for SARI anger (Automaticity r [118] = .27, p < .001; Consistency r [118] = .29, p < .001). For clinical scales, correlations were mixed, with the SCARED (Automaticity r [118] = .54, p < .001; Consistency r [118] = .48, p < .001), CDRS-R (Automaticity r [118] = .13, p < .001; Consistency r [118] = .08, p < .001). Non-RNT trait measures had generally lower correlations; MEAQ-A (Automaticity r [118] = .07 to .29, p < .001; Consistency r [118] = .09 to .26, p < .001), and MWDS (Automaticity r [118] = .20 to .35, p < .001; Consistency r [118] = .19, p < .001), and UPPS-P-C (Automaticity r [118] = −.02 to .09; Consistency r [118] = −.23 to .08), although see BIS (Automaticity r [118] = .40, p < .001; Consistency r [118] = .29, p < .001). See Figure 1 for an arrayed, color-coded representation of RNT-related relationships, and Figure S1 for all scales.

Figure 1.

Correlation coefficient matrix between RNT-related scales, adjusted with the Benjamini-Hochberg procedure. *p < .05 & p ≥ .01, **p < .01 & p ≥ .001, ***p < .001

Exploratory Bifactor Analysis 1 (Narrow, RNT Focused Analysis of Worry and Rumination Scales Only)

For our hypotheses, we expected a main bifactor for RNT, and separate subfactors for TRASH, worry. Parallel analysis and BIC suggested seven factors (Table 4, n = 120). We interpreted the factors to reflect RNT (primary factor), anger rumination, worry engagement, worry frequency, apathetic/abulic rumination, ruminative habit, and reflective thinking. The bifactor solution demonstrated adequate general factor saturation with substantial unique variance accounted for by the group factors (ω_h = .79, ω_t = .98). Nine of ten items loaded significantly onto the RNT factor, except for item 8: “I do at the same time(s) of the day.” Correlations between the factors were not significant (Table 5). Eight of ten TRASH items loaded onto the ruminative habit subfactor. A structure with each scale modeled on a separate factor yielded poor fit (χ² [2127] = 3770.57, p < .001, RMSEA = 0.08, SRMR = .075; CFI = 0.7, TLI = 0.68). The broad EBFA is reported in the Supplement (Figure S2).

Table 4.

Exploratory Bifactor Analysis Item Loadings

	RNT	Anger rumination	Worry engagement	Worry frequency	Apathetic/abulic rumination	Ruminative habit	Reflective thinking
RRS 1 (alone)	0.43	−0.13	0.03	−0.30	0.18	0.12	−0.05
RRS 2 (job)	0.47	0.06	−0.10	0.04	0.13	0.07	0.14
RRS 3 (fatigue)	0.37	0.13	−0.09	0.14	0.30	0.04	−0.05
RRS 4 (hard to concentrate)	0.43	0.00	−0.01	0.12	0.36	−0.05	−0.05
RRS 5 (deserve)^b	0.66	−0.19	0.03	0.00	0.10	−0.40	0.04
RRS 6 (passive)	0.46	0.04	−0.16	−0.04	0.36	0.06	0.05
RRS 7 (analyze events)	0.41	−0.06	−0.05	−0.14	0.01	−0.03	0.55
RRS 8 (don’t feel)	0.43	−0.15	−0.26	−0.25	0.26	−0.05	−0.05
RRS 9 (get going)	0.47	0.03	−0.18	−0.02	0.36	0.07	0.11
RRS 10 (react)^b	0.64	0.03	−0.05	−0.08	0.32	−0.10	0.20
RRS 11 (go away)	0.54	−0.06	−0.04	0.04	0.09	−0.17	0.41
RRS 12 (write down)	0.18	−0.02	0.08	−0.10	0.01	0.17	0.47
RRS 13 (recent situation)	0.55	0.03	0.09	−0.03	0.26	−0.12	−0.04
RRS 14 (concentrate feeling)	0.61	−0.02	0.05	−0.05	0.21	0.00	0.23
RRS 15 (problems)	0.64	0.00	0.01	0.01	0.04	−0.34	0.03
RRS 16 (handle things)^b	0.68	−0.03	−0.06	−0.04	0.25	−0.25	0.16
RRS 17 (sad)	0.58	−0.07	−0.17	−0.05	0.27	−0.07	−0.10
RRS 18 (shortcomings)	0.66	−0.08	−0.06	0.01	0.35	−0.11	−0.11
RRS 19 (up to doing)	0.54	−0.03	−0.18	−0.12	0.26	0.15	−0.05
RRS 20 (analyze personality)	0.45	−0.08	0.03	−0.23	0.09	−0.04	0.47
RRS 21 (someplace alone)	0.43	−0.05	−0.03	−0.05	0.06	−0.05	0.45
RRS 22 (angry)	0.64	0.01	0.03	−0.13	0.33	−0.19	−0.02
TRASH 1 (automatically)^a	0.57	−0.02	0.21	−0.04	−0.01	0.49	0.02
TRASH 2 (same places)	0.37	−0.05	−0.04	−0.06	−0.03	0.20	0.18
TRASH 3 (hard not to)^a	0.63	−0.02	0.23	−0.01	0.10	0.48	−0.23
TRASH 4 (sad anx annoyed)^c	0.47	0.23	−0.08	0.10	0.02	0.36	0.14
TRASH 5 (unintentionally)	0.47	−0.06	−0.12	0.03	0.06	0.49	−0.01
TRASH 6 (when alone)^c	0.58	0.03	−0.15	0.03	0.03	0.32	0.13
TRASH 7 (effort to stop)	0.48	−0.09	0.08	0.00	0.11	0.39	−0.03
TRASH 8 (same times)	0.25	−0.06	−0.06	0.11	−0.11	0.41	0.21
TRASH 9 (start before realize)^a	0.53	−0.10	0.04	0.07	−0.15	0.25	−0.03
TRASH 10 (same situations)^c	0.41	−0.15	−0.06	0.04	−0.06	0.35	0.09
SARI angry 1 (past experiences)	0.38	0.52	0.06	−0.07	0.04	0.02	−0.11
SARI angry 2 (stop thinking)	0.40	0.73	0.10	−0.07	0.09	0.07	−0.09
SARI angry 3 (reasons)	0.34	0.57	−0.08	0.03	0.06	−0.01	0.19
SARI angry 4 (more upset)	0.51	0.60	0.00	−0.07	−0.08	−0.09	−0.08
SARI angry 5 (absorbed)	0.48	0.55	0.06	0.00	−0.02	−0.02	0.01
SARI angry 7 (over and over)	0.37	0.38	−0.01	−0.03	−0.13	0.00	0.43
SARI angry 8 (tire)	0.45	0.61	−0.01	0.16	0.00	0.03	0.11
SARI angry 9 (keep thinking)	0.43	0.66	0.06	−0.17	−0.09	−0.01	0.02
SARI angry 10 (past events)	0.45	0.70	−0.07	0.17	0.09	−0.03	0.04
SARI angry 11 (more angry)	0.30	0.57	−0.15	0.01	−0.09	−0.06	−0.15
SARI sad 1 (past experiences)	0.37	0.66	−0.03	0.00	−0.08	−0.01	−0.05
SARI sad 2 (stop thinking)	0.61	0.00	−0.14	0.13	−0.36	0.07	0.03
SARI sad 3 (reasons)	0.65	0.00	−0.15	0.06	−0.22	−0.22	−0.15
SARI sad 4 (upset)	0.53	−0.18	−0.23	0.07	−0.34	0.11	0.13
SARI sad 5 (absorbed)	0.63	0.11	−0.03	−0.03	−0.34	−0.13	−0.14
SARI sad 7 (over and over)	0.62	0.03	−0.03	0.08	−0.35	−0.17	−0.03
SARI sad 8 (tire)	0.39	−0.09	0.02	−0.03	−0.33	0.09	0.38
SARI sad 9 (keep thinking)	0.67	−0.01	−0.01	0.12	−0.31	−0.09	−0.09
SARI sad 10 (past events)	0.59	0.09	0.12	−0.12	−0.37	−0.11	0.06
Rum VAS 1 (frequency)	0.64	0.04	−0.19	0.22	−0.17	−0.20	−0.13
Rum VAS 2 (control)	0.51	−0.10	−0.15	0.05	−0.24	−0.14	−0.18
Rum VAS 3 (duration)	0.57	0.11	0.06	−0.12	−0.35	−0.01	−0.15
Rum VAS 4 (distress)	0.52	−0.10	0.08	−0.01	−0.03	0.30	0.07
PSWQ-C 1 (worries bother)	0.29	−0.05	0.17	−0.15	0.12	0.01	−0.25
PSWQ-C 2 (don’t worry things)	0.39	0.13	−0.04	−0.09	−0.04	0.28	−0.28
PSWQ-C 3 (things worry me)	0.55	0.00	0.01	−0.13	−0.16	0.35	−0.11
PSWQ-C 4 (can’t help worry)	0.58	−0.15	0.39	−0.01	0.09	0.01	0.07
PSWQ-C 5 (worry under pressure)	0.27	0.01	0.09	0.82	−0.05	−0.06	−0.03
PSWQ-C 6 (always worry)	0.56	0.07	0.57	0.16	0.02	0.03	0.03
PSWQ-C 7 (easy stop worry)	0.53	0.06	0.57	0.08	−0.01	0.09	0.10
PSWQ-C 8 (worry everything else)	0.57	0.01	0.47	0.02	0.05	0.00	−0.01
PSWQ-C 9 (never worry)	0.64	−0.09	0.53	−0.04	−0.09	−0.02	−0.07
PSWQ-C 10 (worry all my life)	0.16	0.00	0.04	0.81	0.09	−0.05	−0.03
PSWQ-C 11 (notice I worry)	0.55	−0.01	0.50	0.14	0.00	0.00	−0.09
PSWQ-C 12 (can’t stop worry)^w	0.21	−0.02	−0.03	0.89	−0.03	0.09	−0.02
PSWQ-C 13 (worry all time)^w	0.51	0.06	0.47	−0.06	−0.13	−0.01	0.02
PSWQ-C 14 (worry until done)^w	0.41	0.00	0.41	0.20	−0.01	0.03	0.10
Initial eigenvalues	19.84	5.18	4.67	3.05	2.77	2.44	2.17
SS loadings	18.67	4.44	3.33	2.96	2.65	2.60	2.31
Variance explained	0.26	0.06	0.05	0.04	0.04	0.04	0.03

Note. Exploratory bifactor analysis of rumination and worry related items produced a primary factor of RNT and subfactors: anger rumination, worry engagement, worry frequency, apathetic/abulic rumination, ruminative habit, and reflective thinking. The three items with highest loadings onto the RNT factor are indicated for each construct: Brooding (b), Automaticity (a), Consistency (c), and Worry (w).

Note. Item loadings ≤ −0.3 or ≥ 0.3 were interpreted and are shown in bold. N = 120.

Table 5.

Factor Score Correlations

	RNT	Anger rumination	Worry engagement	Worry frequency	Apathetic/abulic rumination	Ruminative habit	Reflective thinking
RNT	1
Anger rumination	0.00	1
Worry engagement	0.00	−0.02	1
Worry frequency	0.00	0.08	0.16	1
Apathetic/abulic rumination	0.00	−0.11	−0.05	−0.10	1
Ruminative habit	0.00	−0.07	0.20	0.03	0.00	1
Reflective thinking	0.00	−0.05	−0.12	−0.13	0.01	0.06	1

Note. No correlations were statistically significant. Lack of correlation between the primary factor and subfactors, and low correlations between subfactors, is expected in a bifactor analysis despite using an oblique rotation (Jennrich & Bentler, 2011). Subfactors can be interpreted as distinct variables. N = 120.

Treatment Effects

Significant negative interactions (9 models; n = 29 RF-CBT, 11 TAU, 15 Relax Tx) for the RF-CBT treatment group by time were observed for 5 models: RRS (t [55] = −2.37, p = .02), TRASH consistency scale (t [55] = −2.23, p = .03), PSWQ-C (t [55] = −2.11, p = .04), RNT factor (t [55] = −2.61, p = .01), and TRASH automaticity factor (t [55] = −2.29, p = .03). The treatment by time interaction for TRASH automaticity scale failed to reach significance (t [55] = −1.87, p = .07), and these were not evident for any other measure (consistent with a preliminary analysis with insufficient statistical power). Post hoc, within-group pre-post comparisons for RF-CBT group only indicated significant changes in RRS (t-ratio [58.2] = 5.82, p < .001), TRASH automaticity scale (t-ratio [58.2] = 6.17, p < .001), TRASH consistency scale (t-ratio [58.2] = 4.58, p < .001), PSWQ-C (t-ratio [58.2] = 4.62, p < .001), RNT factor (t-ratio [58.2] = 6.34, p < .001), worry engagement factor (t-ratio [58.2] = 3.24, p = .02), and ruminative habit factor (t-ratio [58.2] = 3.98, p = .002). Pre-post differences in the Relaxation group were only observed for the RNT factor (t-ratio [58.2] = 3.09, p = .03). Model-based estimates are included in the supplement (Tables S3 & S4) and plotted in Figure 2.

Figure 2.

Post hoc estimated marginal group means with 95% confidence intervals for constructs hypothesized to respond to RF-CBT. All data were standardized to pre-treatment data prior to entry into the model and factor scores were calculated using the pre-treatment factor structure loadings. Panels A through D are standard scores for the scales. TRASH automaticity (factor scores) and TRASH consistency (factor scores) reflect factor scores from the TRASH-only factor analysis (Panels E and F). RNT, worry engagement, and ruminative habit reflect factor scores from the narrow (rumination-related scales) factor analysis (Panels G, H, and I). Δ = Glass’s Δ from raw pre/post means, scaled by pre-treatment SD. N = 55

Discussion

Rumination, a form of RNT, is associated with higher risk for the onset, exacerbation, maintenance, and recurrence of depression (Watkins & Roberts, 2020), and poorer outcomes (Clancy et al., 2020; Mazzer et al., 2019; Zawadzki, 2015). Measures of trait rumination such as the RRS and SARI are designed to assess content and general response tendencies such as frequency. They do not directly assess process or temporal components (i.e., automaticity and consistency) of rumination in everyday life. A brief measure that directly captures the habitual strength of rumination is potentially valuable for tracking therapeutic change, including in future treatment studies and clinical practice that targets rumination. Tracking Rumination As a Stable Habit (TRASH) was designed to enable more sensitive and face valid assessment of the automaticity and consistency of the ruminative habit. Generally, TRASH demonstrated strong internal structure and measured two habit factors, as hypothesized. Moreover, an explicit exploratory bifactor analysis of RNT measures (new and existing) identified a distinct ruminative habit factor that was separable from a general RNT factor. TRASH questions also loaded onto the general RNT factor. Subfactors identified emotional (anger and apathy) as well as cognitive elements of rumination or perseverative negative thought (e.g., worry engagement, frequency).

We found significant correlations between our revised TRASH measure and existing measures of rumination, residual depression, worry, and anxiety. Furthermore, items related to worry (i.e., PSWQ-C) loaded onto a general factor of RNT in addition to a separate worry factor (engagement, frequency; Table 4). Convergence of rumination and worry are expected based upon prior literature (de Jong-Meyer et al., 2009; Fresco et al., 2002; Kalmbach et al., 2016; Kircanski et al., 2018; McEvoy et al., 2013; Puccetti et al., 2025; Stade & Ruscio, 2023). While worry and rumination are sometimes conceptualized with respect to different temporal frameworks (future/past respectively), patients may tend to experience rumination and worry flowing together into ongoing repetitive thought (e.g., repetitive thoughts about past difficulties trigger concerns about the future and vice versa). TRASH and RRS scores were moderately related to overall anxiety, panic and GAD, and weakly or unrelated related to subscales of separation, social, and school avoidance anxieties, consistent with generalized anxieties being more closely related with RNT than context-specific fear-based symptoms.

We hypothesized TRASH items would load onto a discrete subfactor within the bifactor analysis. In alignment with our hypothesis, bifactor analyses were able to delineate a separate “habit rumination” subfactor that included most (8 of 10) of the TRASH items - Automaticity and/or Consistency items (more day-to-day experiences, triggers or situations), separate from more general aspects of worry and rumination. Of the two items from the TRASH scale that did not load on the subfactor, they had subthreshold, positive loadings. These questions pertained to a more general pattern of ruminating (e.g., “when sad/upset/distracted”). Items related to worry, rumination, automaticity, and consistency all loaded onto a single RNT bifactor. Because these items assessing habit strongly loaded with more contextual and experiential items on ruminative content, it is possible that habit and emotional content of rumination are highly consistent in adolescents with remitted MDD. Response Styles Theory explicitly links RNT to moments of feeling sad and down. Alternatively, youth may experience difficulty separating triggers (e.g., things that make me ruminate, in general) and contexts (e.g., places and times and people with whom I ruminate) of rumination before psychoeducation about RNT. We were able to show in specific post hoc analyses that there is (preliminarily) reduction of TRASH and RRS in those with history of MDD who received RF-CBT, at a differential rate compared with other therapies (note, insufficient power as the trial is not yet finished). In addition, ruminative habit measurement may be more amenable to repeat assessments of specific habitual aspects with ecological momentary assessment. The ruminative habit subfactor also included the VAS distress item as a positive loading, while two RRS items were inversely loaded on the subfactor (“What am I doing to deserve this?” and “Why do I have problems others don’t have?”). They are inversely loaded, suggesting they are separate from the TRASH subfactor, which is linked more to automaticity (higher items are automaticity 1, 3, and 5). RF-CBT specifically aims to increase awareness of the habit to reduce automaticity through raising awareness, and practice with replacement behaviors (Watkins, 2016).

These results suggest some modest utility for using TRASH Automaticity and Consistency items but call for consideration of omitting item 9 (…start before realize), and item 2 (…in the same places) in repeated assessments (see Table 3). Shortened bifactor subscales for affective elements (worry, sadness, anger) might be useful for specific experimental manipulations or treatment subcomponents, although these may be more niche and unstable elements of the rumination habit. Notably, the visual analog scales did not appear to have a unifying structure that might aid in clinical use and may not be useful for measuring change during rumination interventions. They may reflect more temporally sensitive, or unstable, elements of rumination.

Measuring change with treatment was the original intent of creating the TRASH, yet it was not the specific goal of this manuscript. Nonetheless, showing specific treatment change with TRASH under RF-CBT provides preliminary support for the clinical potential of the scale. With that in mind, the post hoc results with preliminary pre- and post-treatment data do show greater reduction of both RRS and TRASH in the sample with RF-CBT. The direct wording of the TRASH, with both quantity and location for automaticity and consistency elements, may have great face validity and traction for clinical use relative to the RRS. Subsequent work by our group and others can address this question. Further, the brevity and ease of use of the TRASH may relate to repeated use in clinical contexts. We recommend using an average score for the TRASH subscales and total scale, so that the clinician and patients alike can estimate severity of RNT and degree of change with treatment (e.g., 1 = low/absent habit, 3 = modest habit, 5 = strong habit). As the bifactor model would suggest, rumination and worry do share much variance in the main factor. Clinical distinctions about the focus of worry (future) and rumination (past), along with other nuanced distinctions, may have some value in specific treatment techniques, whereas the broad factor is worthy of targeting as well.

Limitations

There are several limitations to consider for the work presented here. The current analyses are secondary to the clinical trial primary outcomes and utilize a volunteer sample. A volunteer sample does not match well all teens who might arrive for clinical care (e.g., availability, monetary incentives, motivation for change). This sample (the RRS range of 35-85 is wide but elevated and does not include low/normative or occasional rumination) may have restricted our ability to understand rumination nuanced differences with other scales. Missingness can also have untoward influences on results, particularly because attrition is often related to clinical characteristics. The impact of attrition, interaction with clinical characteristics, and implications for use of TRASH with RF-CBT are incomplete here as the trial is still ongoing. When the blind is broken for the investigators, we will be able to fully evaluate this important limitation and understand the implications. There were eight individuals who completed the baseline consent but did not complete the questionnaires that went into these factor analyses. In addition, at this stage, only about 40% (N = 55) of the sample had completed all scales pre and post intervention, so there is not a direct linkage between the factor analyses and convergent/divergent statistics with the clinical change data. Further, exploratory bifactor analyses were conducted on a relatively small sample compared to the number of items included in the analyses; these should be considered exploratory initial results to spur follow-on studies with larger samples. The sample also captured a limited age range of youth (14 to 18-year-olds) at a particularly sensitive developmental stage. Broader age ranges will be necessary to generalize results across the lifespan. While this study of youth with rMDD and elevated rumination provides an excellent sample with clinical experience of depression and elevated rumination for validating clinical sensitivity of the TRASH in relation to the RRS and PSWQ, these data may only represent the positive tail of a skewed distribution. Further, the modal range of most ruminative scores may not enable enough range to psychometrically validate clinical variability. In non-clinical controls with low rumination, the concepts and questions from the TRASH may have minimal or mixed relevance to their day to day lives. The scales and differences here are nuanced, as RF-CBT resulted in decreases in RRS, TRASH scales, PSWQ, and related bifactors, if at slightly different rates (and with only 2/3 of the expected sample in this prelim analysis). Presently, there is a great deal of overlap, and individual clinical models and preferences may drive decision making. However, the TRASH scales provide face validity and ask concrete questions. We also acknowledge that people may not perceive/make the subtle distinctions between the concrete questions (e.g., time, place, context) and general tendencies (e.g., when I am sad, in general). The TRASH subcomponents complement the functional analysis that is applied in RF-CBT, which is centered on gathering more details and specifics regarding the experiences leading up to, during, and following habit-driven rumination episodes. The subcomponents are also well-aligned with the evaluation of changes over the course of treatment that are therapeutically meaningful. There are also considerable differences in scales that were weakly or unrelated to TRASH. For example, UPPS-P Lack of premeditation was negatively correlated and unrelated to TRASH consistency and automaticity respectively. For divergent validity, it is not necessary to have inverse correlations, but rather correlations that sort closer to 0. As an example, residual depression symptoms were correlated with RRS (r = .32), but not with either TRASH scale (r = .13 for automaticity and r = .08 for consistency). Convergent validity can range from weak (e.g., r ∼ .20) to modest (r ∼ .40), to strong (r ∼ .60), to colinear (r > .70). Finally, this study’s recruitment and enrollment period largely coincided with the COVID-19 pandemic and subsequent years. Youth generally reported higher rates of anxiety, depression, and post-traumatic symptoms during the pandemic (Marques de Miranda et al., 2020; Racine et al., 2021). Adolescents enrolled during the pandemic may reflect a unique cohort and experience of the ruminative habit relative to those enrolled pre- or post-pandemic.

Conclusion

We revised and expanded an existing scale (HINT) to implement a shorter, simpler, and face valid form (TRASH) that could be used repeatedly by clinicians to track that habitual strength of rumination over the course of treatment. Our intent was to marshal the specificity of the automaticity-related content from HINT and add consistency-related context items into the abbreviated form, and to demonstrate that it measured more discrete facets of the ruminative habit. The bifactor analysis results corroborates existing literature that RNT is a habit with general and specific emotion-based aspects and that a TRASH subfactor is present and changes with RF-CBT. Future work can focus on the application of shorter scale(s) for tracking rumination habit with automaticity and consistency relative to general RNT changes in clinical trials and in clinical care.

Supplemental Material

Supplemental Material - Tracking Rumination as a Stable Habit (TRASH); Scale Modification and Convergent Validity in a Clinical Sample of Youth With a History of Depression

Supplemental Material for Tracking Rumination as a Stable Habit (TRASH); Scale Modification and Convergent Validity in a Clinical Sample of Youth With a History of Depression by Matthew B. Thompson, Leah R. Thomas, Mindy Westlund Schreiner, Noah H. Kingston, Hannah Ollech, Leo Kalotihos, Caty Escobar, Urvi Wagh, Myah Pazdera, Erin Kaufman, Sheila E. Crowell, Henrietta Roberts, Katie L. Bessette, Scott A. Langenecker, Edward R. Watkins in Psychological Reports

Footnotes

Acknowledgements

We are grateful for the statistical consultation provided by Alexander Demos, University of Illinois Chicago.

ORCID iDs

Matthew B. Thompson

Mindy Westlund Schreiner

Noah H. Kingston

Leo Kalotihos

Sheila E. Crowell

Henrietta Roberts

Katie L. Bessette

Edward R. Watkins

Ethical Considerations

This study was approved by the Institutional Review Board (IRB) of University of Utah (IRB_00113733) on 10/22/2018, whose review was accepted by the IRBs of The Ohio State University (2023X0077) and Nationwide Children’s Hospital (STUDY00003890).

Consent to Participate

Written informed assent and consent to participate in this study were provided by participants and participants’ parent(s)/legal guardian(s), respectively.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: SAL and ERW disclosed receipt of the following financial support for the research for this article: This work was supported by the National Institute of Mental Health grant number MH116080 (mPIs SAL and ERW). KLB received financial support through NIMH T32 (MH073517).

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

Data is available on the National Institute of Mental Health Data Archive ().

Supplemental Material

Supplemental material for this article is available online.

Author Biographies

Matthew B. Thompson is a researcher in the Department of Psychiatry and Behavioral Health at The Ohio State University. His research interests include affective neuroscience and computational psychology.

Leah R. Thomas is a researcher in the Department of Psychiatry and Behavioral Health at The Ohio State University. research interests focus on the underlying risk and protective mechanisms in the development of mood disorders.

Mindy Westlund Schreiner is a Psychologist and Clinical Scholar at Nationwide Children's Hospital and an Assistant Professor in the Department of Psychiatry at The Ohio State University. She received her PhD in Clinical Science and Psychopathology Research from the University of Minnesota. Her research focuses on adolescent depression, self-injurious thoughts and behavior, and neurobiological mechanisms of psychopathology.

Noah H. Kingston is a researcher in the Department of Psychiatry and Behavioral Health at The Ohio State University. His research interests include intergenerational transmission of depression/trauma and developing efficient and culturally appropriate psychological assessment strategies for low and middle income countries.

Hannah Ollech is a researcher Nationwide Children’s Hospital Behavioral Health Center; her research interests include non-suicidal self-injury, models of therapy, and psychopathology.

Leo Kalotihos is a researcher at Nationwide Children’s Hospital. He will be pursuing his PhD in clinical psychology, from Louisiana State University. His research examines self-injurious behaviors in adolescents and young adults, with a particular emphasis on therapeutic interventions, theoretical models of suicide, and underrepresented populations.

Caty Escobar is a Clinical Research Specialist in the Department of Psychiatry and Behavioral Health at The Ohio State University. Her research interests include clinical research methods, community mental health, and behavioral science.

Urvi Wagh is a researcher in the Department of Psychiatry and Behavioral Health at The Ohio State University. Her research interests include clinical science, cognition, and interdisciplinary mental health research.

Myah Pazdera is a Senior Clinical Research Coordinator in the Department of Psychiatry at the University of Utah; her research interests include adolescent mental health, developmental psychopathology, and trauma-informed care.

Erin A. Kaufman is an Assistant Professor in the Department of Psychiatry at the University of Utah; she received her PhD in Clinical Psychology from the University of Utah and her research focuses on identifying and treating factors that contribute to self-inflicted injury, borderline personality disorder, and suicide.

Sheila E. Crowell is a Professor in the Department of Psychology at the University of Oregon and a practicing psychologist. She received her PhD in Child Clinical Psychology from the University of Washington. Her research focuses on emotion dysregulation, self-injury, and lifespan developmental psychopathology.

Henrietta Roberts is a Research Fellow in the Mood Disorders Centre at the University of Exeter. She received her PhD from the University of Exeter. Her research interests include rumination, depression, and mechanisms of change in the prevention and treatment of psychopathology.

Katie L. Bessette is a postdoctoral project scientist in the Department of Psychiatry and Biobehavioral Sciences & Semel Institute for Neuroscience and Human Behavior at the University of California Los Angeles. She received her Ph.D. in Clinical Psychology from the University of Illinois Chicago in 2022. She uses a developmental psychopathology approach to understand the neurodevelopmental trajectories of mood disorders in youth, with a particular interest in multimodal interactions of cognitive control and mood regulation.

Scott A. Langenecker is a Professor in the Department of Psychiatry and Behavioral Health at The Ohio State University. He received his PhD in Clinical Psychology from Marquette University. His research focuses on mood disorders, neuropsychology and neuroimaging, and lifespan brain–behavior relationships.

Edward R. Watkins is a Professor of Experimental and Applied Clinical Psychology in the Department of Psychology at the University of Exeter; he received his PhD from the Institute of Psychiatry, King’s College London; his research focuses on rumination, depression, and cognitive-behavioral therapy.

References

Baik

S. Y.

Newman

M. G.

(2023). The transdiagnostic use of worry and rumination to avoid negative emotional contrasts following negative events: A momentary assessment study. Journal of Anxiety Disorders, 95, Article 102679. https://doi.org/10.1016/j.janxdis.2023.102679

Bates

Mächler

Bolker

Walker

(2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67(1), 1–48. https://doi.org/10.18637/jss.v067.i01

Bessette

K. L.

Jacobs

R. H.

Heleniak

Peters

A. T.

Welsh

R. C.

Watkins

E. R.

Langenecker

S. A.

(2020). Malleability of rumination: An exploratory model of CBT-based plasticity and long-term reduced risk for depressive relapse among youth from a pilot randomized clinical trial. PLoS One, 15(6), Article e0233539. https://doi.org/10.1371/journal.pone.0233539

Bijttebier

Raes

Vasey

M. W.

Bastin

Ehring

(2015). Assessment of repetitive negative thinking in children: The Perseverative Thinking Questionnaire–Child Version (PTQ-C). Journal of Psychopathology and Behavioral Assessment, 37(1), 164–170. https://doi.org/10.1007/s10862-014-9446-x

Birmaher

Brent

D. A.

Chiappetta

Bridge

Monga

Baugher

(1999). Psychometric properties of the Screen for Child Anxiety Related Emotional Disorders (SCARED): A replication study. Journal of the American Academy of Child & Adolescent Psychiatry, 38(10), 1230–1236. https://doi.org/10.1097/00004583-199910000-00011

Birmaher

Khetarpal

Brent

Cully

Balach

Kaufman

Neer

S. M.

(1997). The Screen for Child Anxiety Related Emotional Disorders (SCARED): Scale construction and psychometric characteristics. Journal of the American Academy of Child & Adolescent Psychiatry, 36(4), 545–553. https://doi.org/10.1097/00004583-199704000-00018

Carciofo

Jiang

(2021). Deliberate and spontaneous mind wandering in Chinese students: Associations with mindfulness, affect, personality, and life satisfaction. Personality and Individual Differences, 180, Article 110982. https://doi.org/10.1016/j.paid.2021.110982

Carriere

J. S. A.

Seli

Smilek

(2013). Wandering in both mind and body: Individual differences in mind wandering and inattention predict fidgeting. Canadian Journal of Experimental Psychology, 67(1), 19–31. https://doi.org/10.1037/a0031438

Carver

C. S.

White

T. L.

(1994). Behavioral inhibition, behavioral activation, and affective responses to impending reward and punishment: The BIS/BAS scales. Journal of Personality and Social Psychology, 67(2), 319–333. https://doi.org/10.1037/0022-3514.67.2.319

10.

Chorpita

B. F.

Tracey

S. A.

Brown

T. A.

Collica

T. J.

Barlow

D. H.

(1997). Assessment of worry in children and adolescents: An adaptation of the Penn State worry questionnaire. Behaviour Research and Therapy, 35(6), 569–581. https://doi.org/10.1016/S0005-7967(96)00116-7

11.

Clancy

Prestwich

Caperon

Tsipa

O’Connor

D. B.

(2020). The association between worry and rumination with sleep in non-clinical populations: A systematic review and meta-analysis. Health Psychology Review, 14(4), 427–448. https://doi.org/10.1080/17437199.2019.1700819

12.

Colvin

Gardner

Labelle

P. R.

Santor

(2021). The automaticity of positive and negative thinking: A scoping review of mental habits. Cognitive Therapy and Research, 45(6), 1037–1063. https://doi.org/10.1007/s10608-021-10218-4

13.

Cook

Mostazir

Watkins

E. R.

(2019). Reducing Stress and Preventing Depression (RESPOND): Randomized controlled trial of web-based rumination-focused cognitive behavioral therapy for high-ruminating university students. Journal of Medical Internet Research, 21(5), Article e11349. https://doi.org/10.2196/11349

14.

de Jong-Meyer

Beck

Riede

(2009). Relationships between rumination, worry, intolerance of uncertainty and metacognitive beliefs. Personality and Individual Differences, 46(4), 547–551. https://doi.org/10.1016/j.paid.2008.12.010

15.

Ehring

(2021). Thinking too much: Rumination and psychopathology. World Psychiatry, 20(3), 441–442. https://doi.org/10.1002/wps.20910

16.

Ehring

Watkins

E. R.

(2008). Repetitive Negative Thinking as a Transdiagnostic Process. International Journal of Cognitive Therapy, 1, 192–205. https://doi.org/10.1521/ijct.2008.1.3.192

17.

Ehring

Zetsche

Weidacker

Wahl

Schönfeld

Ehlers

(2011). The Perseverative Thinking Questionnaire (PTQ): Validation of a content-independent measure of repetitive negative thinking. Journal of Behavior Therapy and Experimental Psychiatry, 42(2), 225–232. https://doi.org/10.1016/j.jbtep.2010.12.003

18.

Figueroa

C. A.

DeJong

Mocking

R. J. T.

Fox

Rive

M. M.

Schene

A. H.

Stein

Ruhé

H. G.

(2019). Attentional control, rumination and recurrence of depression. Journal of Affective Disorders, 256, 364–372. https://doi.org/10.1016/j.jad.2019.05.072

19.

Fresco

D. M.

Frankel

A. N.

Mennin

D. S.

Turk

C. L.

Heimberg

R. G.

(2002). Distinct and overlapping features of rumination and worry: The relationship of cognitive production to negative affective states. Cognitive Therapy and Research, 26(2), 179–188. https://doi.org/10.1023/A:1014517718949

20.

Gámez

Chmielewski

Kotov

Ruggero

Watson

(2011). Development of a measure of experiential avoidance: The multidimensional experiential avoidance questionnaire. Psychological Assessment, 23(3), 692–713. https://doi.org/10.1037/a0023242

21.

Grierson

A. B.

Hickie

I. B.

Naismith

S. L.

Scott

(2016). The role of rumination in illness trajectories in youth: Linking transdiagnostic processes with clinical staging models. Psychological Medicine, 46(12), 2467–2484. https://doi.org/10.1017/S0033291716001392

22.

Gustavson

D. E.

du Pont

Whisman

M. A.

Miyake

(2018). Evidence for transdiagnostic repetitive negative thinking and its association with rumination, worry, and depression and anxiety symptoms: A commonality analysis. Collabra: Psychology, 4(1), 13. https://doi.org/10.1525/collabra.128

23.

Hoyer

Gloster

A. T.

Herzberg

P. Y.

(2009). Is worry different from rumination? Yes, it is more predictive of psychopathology. Psychosocial Medicine, 6, Doc06. https://doi.org/10.3205/psm000062

24.

Jacobs

R. H.

Watkins

E. R.

Peters

A. T.

Feldhaus

C. G.

Barba

Carbray

Langenecker

S. A.

(2016). Targeting ruminative thinking in adolescents at risk for depressive relapse: Rumination-focused cognitive behavior therapy in a pilot randomized controlled trial with resting-state fMRI. PLoS One, 11(11), Article e0163952. https://doi.org/10.1371/journal.pone.0163952

25.

Jennrich

R. I.

Bentler

P. M.

(2011). Exploratory bifactor analysis. Psychometrika, 76(4), 537–549. https://doi.org/10.1007/s11336-011-9218-4

26.

Just

Alloy

L. B.

(1997). The response styles theory of depression: Tests and an extension of the theory. Journal of Abnormal Psychology, 106(2), 221–229. https://doi.org/10.1037/0021-843X.106.2.221

27.

Kalmbach

D. A.

Pillai

Ciesla

J. A.

(2016). The correspondence of changes in depressive rumination and worry to weekly variations in affective symptoms. Australian Journal of Psychology, 68(1), 52–60. https://doi.org/10.1111/ajpy.12090

28.

Kennedy

J. C.

Dunlop

B. W.

Baldwin

S. E.

Liu

C. J.

Mayberg

H. S.

Craighead

W. E.

(2024). Assessing in-session rumination during CBT for depression: Replication and further evaluation of an observational measure. Journal of Mood & Anxiety Disorders, 7, 100060. https://doi.org/10.1016/j.xjmad.2024.100060

29.

Kircanski

Thompson

R. J.

Sorenson

Sherdell

Gotlib

I. H.

(2018). The Everyday Dynamics of Rumination and Worry: Precipitant Events and Affective Consequences. Cognition & Emotion, 32(7), 1424–1436. https://doi.org/10.1080/02699931.2017.1278679

30.

Kuznetsova

Brockhoff

P. B.

Christensen

R. H. B.

(2017). lmerTest package: Tests in linear mixed effects models. Journal of Statistical Software, 82(13), 1–26. https://doi.org/10.18637/jss.v082.i13

31.

Langenecker

S. A.

Westlund Schreiner

Bessette

K. L.

Roberts

Thomas

Dillahunt

Pocius

S. L.

Feldman

D. A.

Jago

Farstead

Pazdera

Kaufman

Galloway

J. A.

Kerig

P. K.

Bakian

Welsh

R. C.

Jacobs

R. H.

Crowell

S. E.

Watkins

E. R.

(2024). Rumination-focused cognitive behavioral therapy reduces rumination and targeted cross-network connectivity in youth with a history of depression. Biological Psychiatry: Global Open Science, 4(1), 1–10. https://doi.org/10.1016/j.bpsgos.2023.08.012

32.

Lenth

(2025). emmeans: Estimated marginal means (Version 1.11.2-8) [Computer software]. Comprehensive R Archive Network. https://CRAN.R-project.org/package=emmeans.

33.

Liang

Lee

Y.-H.

(2019). Factor Structure of the Ruminative Response Scale and Measurement Invariance across Gender and Age among Chinese Adolescents. Advances in Applied Sociology, 9(6), 193–207. https://doi.org/10.4236/aasoci.2019.96016

34.

Marques de Miranda

da Silva Athanasio

Sena Oliveira

A. C.

Simoes-e-Silva

A. C.

(2020). How is COVID-19 pandemic impacting mental health of children and adolescents? International Journal of Disaster Risk Reduction, 51, Article 101845. https://doi.org/10.1016/j.ijdrr.2020.101845

35.

Mayes

T. L.

Bernstein

I. H.

Haley

C. L.

Kennard

B. D.

Emslie

G. J.

(2010). Psychometric properties of the Children’s depression rating scale–revised in adolescents. Journal of Child and Adolescent Psychopharmacology, 20(6), 513–516. https://doi.org/10.1089/cap.2010.0063

36.

Mazzer

Boersma

Linton

S. J.

(2019). Rumination, poor sleep and psychological distress in adolescents. Journal of Affective Disorders, 245, 686–696. https://doi.org/10.1016/j.jad.2018.11.053

37.

McEvoy

P. M.

Mahoney

A. E. J.

Moulds

M. L.

(2010). Are worry, rumination, and post-event processing one and the same? Journal of Anxiety Disorders, 24(5), 509–519. https://doi.org/10.1016/j.janxdis.2010.03.008

38.

McEvoy

P. M.

Watson

Watkins

E. R.

Nathan

(2013). The relationship between worry, rumination, and comorbidity: Evidence for repetitive negative thinking as a transdiagnostic construct. Journal of Affective Disorders, 151(1), 313–320. https://doi.org/10.1016/j.jad.2013.06.014

39.

Muris

Roelofs

Meesters

Boomsma

(2004). Rumination and worry in nonclinical adolescents. Cognitive Therapy and Research, 28(4), 539–554. https://doi.org/10.1023/b:cotr.0000045563.66060.3e

40.

Nolen-Hoeksema

(1991). Responses to depression and their effects on the duration of depressive episodes. Journal of Abnormal Psychology, 100(4), 569–582. https://doi.org/10.1037/0021-843X.100.4.569

41.

Nolen-Hoeksema

Morrow

(1991). A prospective study of depression and posttraumatic stress symptoms after a natural disaster. Journal of Personality and Social Psychology, 61(1), 115–121. https://doi.org/10.1037//0022-3514.61.1.115

42.

Nolen-Hoeksema

Morrow

(1993). Effects of rumination and distraction on naturally occurring depressed mood. Cognition & Emotion, 7(6), 561–570. https://doi.org/10.1080/02699939308409206

43.

Nolen-Hoeksema

Wisco

B. E.

Lyubomirsky

(2008). Rethinking rumination. Perspectives on Psychological Science, 3(5), 400–424. https://doi.org/10.1111/j.1745-6924.2008.00088.x

44.

Păsărelu

C. R.

Dobrean

Balazsi

Predescu

Şipos

Lupu

(2017). The Penn State Worry Questionnaire for Children: Age, Gender and Clinical Invariance. Child Psychiatry and Human Development, 48(3), 359–369. https://doi.org/10.1007/s10578-016-0663-2

45.

Peled

Moretti

M. M.

(2007). Rumination on anger and sadness in adolescence: Fueling of fury and deepening of despair. Journal of Clinical Child and Adolescent Psychology, 36(1), 66–75. https://doi.org/10.1080/15374410709336569

46.

Peled

Moretti

M. M.

(2010). Ruminating on rumination: Are rumination on anger and sadness differentially related to aggression and depressed mood? Journal of Psychopathology and Behavioral Assessment, 32(1), 108–117. https://doi.org/10.1007/s10862-009-9136-2

47.

Pestle

S. L.

Chorpita

B. F.

Schiffman

(2008). Psychometric properties of the penn state worry questionnaire for children. Journal of Clinical Child and Adolescent Psychology, 37(2), 465–471. https://doi.org/10.1080/15374410801955896

48.

Poznanski

E. O.

Mokros

H. B.

(1996). Children’s depression rating scale–revised (CDRS-R). Western Psychological Services.

49.

Puccetti

N. A.

Stamatis

C. A.

Timpano

K. R.

Heller

A. S.

(2025). Worry and rumination elicit similar neural representations: neuroimaging evidence for repetitive negative thinking. Cognitive, Affective, & Behavioral Neuroscience, 25(2), 488–500. https://doi.org/10.3758/s13415-024-01239-z

50.

R Core Team . (2024). R: A language and environment for statistical computing (Version 4.4.1) [Computer software]. R Foundation for Statistical Computing. https://www.R-project.org/ [1, 2]

51.

Racine

McArthur

B. A.

Cooke

J. E.

Eirich

Zhu

Madigan

(2021). Global prevalence of depressive and anxiety symptoms in children and adolescents during COVID-19. JAMA Pediatrics, 175(11), 1142–1150. https://doi.org/10.1001/jamapediatrics.2021.2482

52.

Raes

(2010). Rumination and worry as mediators of the relationship between self-compassion and depression and anxiety. Personality and Individual Differences, 48(6), 757–761. https://doi.org/10.1016/j.paid.2010.01.023

53.

Revelle

(2024). psych: Procedures for psychological, psychometric, and personality research (Version 2.4.6.26) [Computer software]. Northwestern University. https://personality-project.org/r/ [1, 2, 3, 4]

54.

Roberts

Jacobs

R. H.

Bessette

K. L.

Crowell

S. E.

Westlund-Schreiner

Thomas

Easter

R. E.

Pocius

S. L.

Dillahunt

Frandsen

Schubert

Farstead

Kerig

Welsh

R. C.

Jago

Langenecker

S. A.

Watkins

E. R.

(2021). Mechanisms of rumination change in adolescent depression (RuMeChange): study protocol for a randomised controlled trial of rumination-focused cognitive behavioural therapy to reduce ruminative habit and risk of depressive relapse in high-ruminating adolescents. BMC Psychiatry, 21, 206. https://doi.org/10.1186/s12888-021-03193-3

55.

Rosseel

(2012). lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(2), 1–36. https://doi.org/10.18637/jss.v048.i02

56.

Shaw

Z. A.

Conway

C. C.

Starr

L. R.

(2021). Distinguishing transdiagnostic versus disorder-specific pathways between ruminative brooding and internalizing psychopathology. Research on Child and Adolescent Psychopathology, 49(10), 1319–1331. https://doi.org/10.1007/s10802-020-00714-8

57.

Shaw

Z. A.

Hilt

L. M.

Starr

L. R.

(2019). The developmental origins of ruminative response style. Clinical Psychology Review, 74, Article 101780. https://doi.org/10.1016/j.cpr.2019.101780

58.

Smith

J. M.

Alloy

L. B.

(2009). A roadmap to rumination: A review of the definition, assessment, and conceptualization of this multifaceted construct. Clinical Psychology Review, 29(2), 116–128. https://doi.org/10.1016/j.cpr.2008.10.003

59.

Stade

E. C.

Ruscio

A. M.

(2023). A meta-analysis of the relationship between worry and rumination. Clinical Psychological Science, 11(3), 552–573. https://doi.org/10.1177/21677026221131309

60.

Tierney

Cook

(2023). Expanding tidy data principles to facilitate missing data exploration, visualization and assessment of imputations. Journal of Statistical Software, 105(7), 1–31. https://doi.org/10.18637/jss.v105.i07

61.

Tomko

R. L.

Prisciandaro

J. J.

Kutty Falls

Magid

(2016). The Structure of the UPPS-R-Child Impulsivity Scale and its Relations with Substance Use Outcomes Among Treatment-Seeking Adolescents. Drug and Alcohol Dependence, 161, 276–283. https://doi.org/10.1016/j.drugalcdep.2016.02.010

62.

Topper

Emmelkamp

P. M.

Watkins

Ehring

(2017). Prevention of anxiety disorders and depression by targeting excessive worry and rumination. Behaviour Research and Therapy, 90, 123–136. https://doi.org/10.1016/j.brat.2016.12.015

63.

Treynor

Gonzalez

Nolen-Hoeksema

(2003). Rumination Reconsidered: A Psychometric Analysis. Cognitive Therapy and Research, 27(3), 247–259. https://doi.org/10.1023/A:1023910315561

64.

Verplanken

Friborg

Wang

C. E.

Trafimow

Woolf

(2007). Mental habits: Metacognitive reflection on negative self-thinking. Journal of Personality and Social Psychology, 92(3), 526–541. https://doi.org/10.1037/0022-3514.92.3.526

65.

Watkins

E. R.

(2016). Rumination-focused cognitive-behavioral therapy for depression. Guilford Press.

66.

Watkins

E. R.

Baeyens

C. B.

Read

(2009). Concreteness training reduces dysphoria: Proof-of-principle for repeated cognitive bias modification in depression. Journal of Abnormal Psychology, 118(1), 55–64. https://doi.org/10.1037/a0013642

67.

Watkins

E. R.

Nolen-Hoeksema

(2014). A habit-goal framework of depressive rumination. Journal of Abnormal Psychology, 123(1), 24–34. https://doi.org/10.1037/a0035540

68.

Watkins

E. R.

Roberts

(2020). Reflecting on rumination: Consequences, causes, mechanisms and treatment of rumination. Behaviour Research and Therapy, 127, 103573. https://doi.org/10.1016/j.brat.2020.103573

69.

Wichelns

G. A.

Renna

M. E.

Mennin

D. S.

(2016). Preliminary validation of subjective anchor scales for worry and rumination. Cognitive Therapy and Research, 40(5), 645–660. https://doi.org/10.1007/s10608-016-9779-8

70.

Wickham

(2016). ggplot2: Elegant graphics for data analysis. Springer.

71.

Wood

Neal

D. T.

(2007). A new look at habits and the habit–goal interface. Psychological Review, 114(4), 843–863. https://doi.org/10.1037/0033-295X.114.4.843

72.

Zapolski

T. C. B.

Stairs

A. M.

Settles

R. F.

Combs

J. L.

Smith

G. T.

(2010). The measurement of dispositions to rash action in children. Assessment, 17(1), 116–125. https://doi.org/10.1177/1073191109351372

73.

Zawadzki

M. J.

(2015). Rumination is independently associated with poor psychological health. Psychology & Health, 30(10), 1146–1163. https://doi.org/10.1080/08870446.2015.1026904

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.38 MB