The term co-rumination refers to talking about one’s problems with another person in a way that is excessive, repetitive, passive, speculative, and focused on negative emotions and consequences1. This behavior emerges as an (often ineffective) attempt to seek social support and manage distress by the disclosing individual2. Unfortunately, co-rumination appears to increase the salience of perceived problems and amplifies negative affect without generating actionable solutions3. Thus, this interpersonal process is theorized to increase depressive thinking patterns (e.g., self-criticism, maladaptive attributions) and exacerbate stress response4. Research has found that adolescent co-rumination with peers and their mothers is indeed linked to internalizing psychopathology like depression567. Although co-rumination is detrimental, persons tend to engage in this type of conversation in the context of close, confiding relationships18. Taken together, findings indicate that co-rumination is an unproductive emotion regulation strategy that adolescents tend to use with close others like parents and peers.

Preliminary data suggest that co-rumination increases risk for depression and other internalizing pathology indirectly via increasing solitary rumination (or the mental tendency to passively and repeatedly focus on one’s distress and its associated causes and consequences) 91011. Importantly, rumination often begins in the context of familial stressors12. Some have argued that caregivers may inadvertently socialize rumination in their children by failing to teach or support more adaptive and action-oriented emotion regulation strategies, especially for events eliciting sadness or anger13. These caregivers may instead suggest or more subtly reinforce disengagement strategies1415. A sizable literature supports the roles of parental modeling and reinforcing avoidant behavior in the etiology of youth worry and anxiety16, which are forms of repetitive negative thinking. Despite evidence pointing to the family environment as a training ground for the ruminative process, literature relevant to the intergenerational transmission of rumination itself is mixed.

Some studies find that maternal rumination reduces interactive parenting behavior and disrupts parent/child relationship quality1718. Research shows that parenting styles characterized by high control and protectiveness are also associated with youth rumination192021. One recent systematic review found a number of associations between parenting behavior and adolescent rumination19. Specifically, adolescent rumination was positively correlated with parental control, emotional abuse, negative affectivity, and authoritarian parenting, and negatively correlated with positive affectivity, low control, parental solicitation, and authoritative parenting. Of particular relevance to co-rumination, one study found maternal encouragement of emotional expressivity predicted increased rumination among daughters22, while another found a family style characterized by passive expression of negative emotions during early childhood predicted greater levels of rumination among youth by adolescence23.

In contrast, one study that more directly investigated intergenerational transmission of rumination failed to find support for this developmental pathway. Dunning and colleagues24 examined whether maternal rumination prospectively predicted their child’s rumination via parenting behaviors (i.e., psychological control) and family functioning (i.e., communication, affective expression, involvement). They also tested whether intergenerational transmission predicted internalizing symptoms among adolescents. Results indicated that maternal rumination did not directly predict adolescent rumination; nor were there indirect effects of maternal rumination on adolescent rumination via parenting behaviors or family functioning. Although higher maternal rumination did predict less effective parenting and more maladaptive family characteristics at follow-up, these factors did not influence adolescent rumination.

In summary, evidence for the intergenerational transmission of rumination is limited to date. However: (1) caregiver socialization around problem-solving and emotion regulation can increase risk for rumination23, (2) co-rumination appears to increase risk for solitary or individual rumination among youth10, and (3) there is an extensive literature supporting the robust link between rumination and internalizing symptoms2526. The pathway from rumination to psychopathology is well established and forms the rationale for rumination-focused cognitive behavioral therapy (RF-CBT)27. While initially developed for adults, our group has adapted RF- CBT for youth and found success in reducing both rumination and the recurrence of depression2830. However, further exploration is needed to examine how factors unique to youth may impact the effectiveness of RF-CBT.

Adolescents are embedded in a family environment that likely influences how psychopathology is perceived, how coping skills are deployed and reinforced, and whether treatment is effective. Youth rarely seek therapy independently, and parents often make vital care decisions based on their perception of youth symptoms. Research shows that parental affect, expectations, and beliefs impact how much attention and emphasis they place on child symptoms31. Yet it is presently unknown whether caregiver engagement with rumination impacts how they view their child’s related symptoms (does caregiver engagement with rumination obscure their understanding of their child’s degree of struggle with depression?). It is also unclear whether familial co-rumination affects the success of youth treatment targeting solitary rumination (does co-rumination dampen RF-CBT efficacy?). Understanding more about the potential influence of the family environment may aid in tailoring interventions to youth and more effectively address factors that can hamper treatment response.

The present study uses data from a randomized clinical trial of adolescent RF-CBT to examine three secondary aims (NCT03859297)3032. First, we sought to examine whether caregiver engagement in ruminative processes with their child (co-rumination) and/or caregivers’ own experiences with solitary rumination are associated with bias in their perception of their adolescent’s current depression symptoms. Parents’ own experience with rumination may either normalize or amplify concerns about their child’s symptoms. Bias was operationalized as a greater discrepancy between caregiver scores and both child and independent evaluator1 scores on the Children's Depression Rating Scale-Revised (CDRS-R; described below)33. Our hypotheses associated with this first aim were somewhat exploratory, as we were unsure a priori whether any potential bias would result in caregivers under or overestimating youth symptoms.

Second, we aimed to examine whether caregiver-child co-rumination scores would moderate rumination-related outcomes for youth at treatment termination. We hypothesized that co- rumination scores may moderate the relation between RF-CBT exposure (RF-CBT vs. treatment as usual; TAU) and treatment outcomes (rumination scores) for youth, such that higher engagement in co-rumination with a caregiver would dampen treatment benefit between pre- and post-intervention measurements, as compared to family environments characterized by lower co- rumination scores. Third, we examined differences in the rates of treatment effect decay between high versus low co-rumination subgroups and between high versus low caregiver rumination subgroups from pre- to post-treatment and across follow-up time points.

With the present study, we sought to explore how familial processes relevant to rumination may impact the effectiveness of RF-CBT among youth. Specifically, we explored (1) potential caregiver bias when identifying and reporting on their child’s depressive symptoms as a result of their own engagement with the ruminative process, (2) whether youth RF-CBT treatment effects would differ based on experiences of co-rumination with a caregiver, and (3) whether treatment effects on youth rumination would decay more quickly in the context of high (versus low) co- rumination at home. While these results are preliminary due to the smaller sample size for three- way interactions, the main effect of treatment on co-rumination was encouraging.

Our exploration of potential caregiver bias indicates that adult experiences with rumination (whether co-rumination engagement with their child, or independent rumination) do not systematically bias their perception of their child’s depression symptoms in comparison to child report of their own symptoms or ratings made by an independent evaluator. This result is promising, as caregivers are often vital informants for identifying and characterizing youth psychopathology, connecting their children to treatment, encouraging therapy engagement, and maintaining treatment gains. Discrepancy between caregiver and child reports of youth psychopathology is a long-recognized and common issue3940 that may be associated with adverse outcomes41. Our results indicate that caregiver engagement with rumination does not appear to be a source of reporter bias. This is important, as caregivers are often a key informant for conceptualizing youth psychopathology in clinical practice.

We also found that co-rumination engagement did not appear to dampen treatment effects at treatment termination or result in a more rapid/severe decay in benefit post-treatment. In fact, families that reported greater engagement in co-rumination at baseline had youth who benefited more from treatment‚ as measured by reductions on the RRS. Taken together, these results speak to the robustness of beneficial effects of RF-CBT on youth rumination. Treatment effects were observed even in the context of co-rumination in the family. It is possible that higher co- rumination at baseline may relate to greater engagement of both the youth and the caregiver in therapy. Specifically, youth within families who tend to co-ruminate may begin to integrate skills and other therapy content within the context of co-rumination, potentially increasing their opportunity to ingrain more adaptive habits. The current preliminary analysis can be repeated in the R33 phase, when we have a larger sample and a more robust comparison group (e.g., RF- CBT vs. relaxation therapy, which should not change rumination or co-rumination).

As with all research, this study has some notable limitations. Caregiver-child co-rumination is largely understudied in comparison to youth-peer co-rumination and thus, any examination is valuable in the context of youth internalizing pathology. However, we only examined caregiver reports of co-rumination without peer contexts, and we were not able to assess co-rumination shifts over the course of treatment. We also examined these patterns across all caregiver/child dyads and were not powered to examine potentially unique effects of mother/father and daughter/son pairings. As data for these analyses were drawn from a larger study evaluating youth treatment response, a more detailed evaluation of caregiver-child co-rumination with more frequent assessment points and more complete follow-up is warranted. The original trial was not designed for an analysis of the type conducted here, but we were encouraged to explore the relationships based on literature that was published after the trial was started, and to inform hypotheses for future work. In this pilot trial, many participants missed assessment windows during our lengthy follow-up period, resulting in significant reductions in power when examining related questions (see Table 2). The R61 phase of the study was powered for a sample size of 30 per cell to obtain an effect size change of .5 SD with .80 power. This was designed as a simple t-test change difference between the change in the TAU group and the change in the RF-CBT group, in which the actual effects were notably stronger 30. Our sample was also largely white, female, cisgender, and non-Hispanic, which may limit the generalizability of our results. Finally, the effects of caregiver rumination and co-rumination on report-bias were only examined in the context of adolescent depression; caregiver rumination has not yet been assessed with other relevant forms of psychopathology, such as adolescent anxiety.

Despite limitations, our preliminary results indicate that caregiver-child co-rumination did not reduce the effectiveness of RF-CBT among youth or result in greater decay of treatment benefits over time. This positively speaks to the strength of the intervention among youth. However, future work would likely still benefit from the explicit integration of familial involvement with the therapeutic process, as this may further increase the effectiveness of the intervention. Overall, this study highlights the importance of examining environmental factors that may influence treatment outcomes for youth with depression. Continued work is needed to evaluate other unique aspects of youth’s social systems in an effort to most effectively leverage interventions. In particular, further research is needed to test whether similar patterns emerge with respect to co-rumination with friends and peers, given the increasing importance of such relationships during the adolescent period.

This study was supported by the the National Institute of Mental Health (R61MH116080; PI Langenecker; PI Watkins). EAK’s time on this work was supported by the National Institute of Mental Health (K23MH135225). Effort of SAL, ERW, MWS, SEC, KLB, EAF, BF were also supported by MH116080.