Repetitive negative thinking in adolescence: a mixed methods study

Adolescence is a period of heightened vulnerability for common mental health challenges [1, 2]. The onset of puberty, transitions between primary to secondary schooling, and shifting relationships and self-concepts are just some of the many stressors that young adolescents face [3, 4]. Against this backdrop, the frequent emergence of anxiety and depressive symptoms during early adolescence (ages 10–14; [3]) is unsurprising [2, 5, 6]. Left unaddressed, these symptoms can develop into common mental disorders by mid-to-late adolescence. Recent epidemiological studies suggest that the peak age of onset is 14.5 for social anxiety disorder and obsessive-compulsive disorder, 15.5 for generalised anxiety disorder and panic disorder, and 19.5 for depressive disorders [2]. Interventions aimed at preventing anxiety and depression should be delivered prior to these peak vulnerability periods for disorder onset.

Given the frequent co-occurrence of depression and anxiety and their numerous overlapping risk factors [7, 8], transdiagnostic prevention approaches make clinical and economic sense [9]. A promising target for transdiagnostic preventive interventions is repetitive negative thinking (RNT), the common process of getting stuck in one’s own negative thoughts without arriving at useful insights or solutions [10]. Broadly defined as a thinking style that is recurrent, negatively valenced, and difficult to control, RNT encompasses rumination (repetitive thoughts about past negative events and emotions) and worry (repetitive thoughts about potential future negative outcomes) [11]. RNT predicts not only depression and anxiety symptom severity and comorbidity, but also the development, maintenance, and relapse of depressive and anxiety disorders in both adolescents and adults [12‐16].

Growing recognition of RNT as a promising target in the prevention and treatment of depression and anxiety has led to the development of interventions that specifically aim to reduce RNT. Rumination-focused cognitive behavioural therapy (RFCBT; [17]), its subsequent adaptations [18, 19], and interventions that have incorporated RFCBT principles (e.g., [20, 21]) train adolescents and adults to identify when they are engaging in RNT and to shift towards more helpful thinking processes or other emotionally-adaptive behaviours. RNT-focused interventions have demonstrated effectiveness in both preventing and reducing depression and anxiety [22‐26], further supporting the choice of RNT as a transdiagnostic preventive intervention target. However, existing RNT-focused preventive interventions typically target mid-to-late adolescence (e.g., [19]: ages 15–22; [18]: ages 18–24), developmental periods that coincide with or are subsequent to peak ages of onset for depression and anxiety disorders [2]. There is an absence of RNT-focused preventive interventions for adolescents under the age of 15, the developmental period when depressive and anxiety symptoms are likely to first emerge [23].

This prevention gap warrants attention, since RNT has been repeatedly identified as contributing to depression, anxiety, and various other mental health challenges in young adolescents [27‐29]. RNT may be more malleable during the heightened neural and social plasticity of early adolescence [30, 31], and developing RNT-reduction skills may not only prevent depression and anxiety, but also improve functioning and wellbeing. These views have been echoed by young people with lived experience of depression and anxiety, who emphasised the importance of early RNT-focused interventions (“the earlier the better”; [23]).

To create effective and engaging RNT-focused preventive interventions, we must first develop a comprehensive understanding of how adolescents experience RNT. A mixed methods approach, integrating both quantitative and qualitative methods, provides complementary perspectives to elucidate psychological lived experience [32‐34]. Existing research typically investigates RNT using standardised self-report questionnaires developed by researchers and clinicians. Although such quantitative methods allow for efficient measurement and facilitate comparisons, they should be complemented with qualitative methods for a more nuanced and in-depth understanding of psychological phenomena [33, 34]. The small number of qualitative studies investigating individual experiences of RNT in older populations have provided valuable insights [35‐38]. Specifically, participants across these studies described RNT as overwhelming and difficult to control, often triggered by negative emotions or social interactions, and frequently managed using distraction or interpersonal support, albeit with limited success. These insights have been gathered from clinical or treatment-seeking older adolescents or adults, but the phenomenological lived experience of RNT in early adolescence remains unexamined. Given the distinct context of early adolescence, including developing cognitive capacities, self-awareness, and unique sociobiological stressors [3, 4], a mixed methods study incorporating early adolescent perspectives is important to elucidate not only the prevalence and severity of RNT, but also the subjective meanings and contexts characterising RNT in this developmental period. Furthermore, the qualitative experience of RNT across adolescence has been underexplored. Capturing adolescent lived experience of RNT using both quantitative and qualitative approaches can inform the development of age-appropriate interventions, ensuring the language, examples, and strategies resonate with populations they are being delivered in, to facilitate program engagement.

In addition to RNT, the adolescent experience of co-rumination, which involves excessive discussion of personal problems, also warrants further investigation. Co-rumination typically occurs in close relationships in which individuals repeatedly rehash issues without reaching new insights or solutions, a relational pattern that can foster closeness but also increase negative social interactions and distress [39]. Co-rumination predicts increased depressive and anxiety symptoms, and more severe depressive episodes over time, including in early adolescence [40‐42]. As an understudied transdiagnostic phenomenon, co-rumination is an important companion construct to RNT that merits further investigation and potential intervention in adolescence.

In light of these considerations, the current study aimed to investigate adolescent experiences of RNT and co-rumination using a combination of qualitative and quantitative methods via an online survey, with inclusion of previously underexamined early adolescent perspectives. As the first mixed methods study in this area, we decided to recruit a 10-18-year-old sample from the community, but to categorise the sample into younger (10–14-year-old) and older (15–18-year-old) adolescents. This allowed us to explore whether there are important developmental differences between how younger and older adolescents experience RNT. To incorporate valuable parental insights, parents of adolescents aged 10–18 from the community were also recruited to participate in our online survey. The survey included open-ended and multiple-choice questions regarding personal definitions, triggers, content, frequency, and duration of adolescent RNT, as well as perceived RNT purpose and coping strategies. To characterise the nature of the sample and measure co-rumination, participants also completed standardised self- or parent-report questionnaires about this construct.

Adolescence is a developmental window of vulnerability for the onset of common mental health challenges [1, 2]. The current mixed methods study explored the lived experience of RNT during adolescence, given evidence that this transdiagnostic process predicts the onset and maintenance of a range of mental disorders [27‐29]. By combining quantitative and qualitative data from both adolescents and parents, the study has provided novel insights that can inform the design of age-appropriate RNT-focused preventive interventions for depression and anxiety. Findings broadly support and extend prior qualitative research on RNT in older adolescents and adults [35‐38], while also revealing developmental differences that highlight the unique vulnerabilities and needs of younger adolescents struggling with RNT.

Mean RNT severity scores among both younger and older adolescents in our study were comparable to clinical samples [53], and substantially higher than typical community samples [45]. This may reflect elevated depression and anxiety in our sample, but also underscores that high levels of RNT can emerge as early as age 10. Furthermore, patterns of RNT frequency and duration among our community-recruited younger and older adolescents were comparable to those observed in clinically depressed older adolescent [36] and community adult samples [35], with most adolescents in our study reporting RNT episodes several times per day lasting between 20 min to an hour. Older adolescents reported slightly longer and more frequent RNT episodes than younger adolescents, indicating a potentially worsening age-related trajectory. Together, these findings suggest that early adolescence is not only a relevant developmental period in which to target RNT via interventions, but may also be an optimal time to introduce RNT management skills before maladaptive thought processes consolidate into habits [55]. Early targeting of RNT is crucial as higher RNT predicts earlier onset, greater severity, and higher likelihood of relapse in depressive and anxiety disorders [10].

Parallelling high levels of RNT, adolescents in our study reported elevated levels of co-rumination with both peers and parents, consistent with prior findings from community samples [46, 54]. Furthermore, parents also reported elevated levels of co-rumination with their adolescent children, again consistent with prior reports from community samples [46]. These findings warrant attention, given that co-rumination predicts heightened depressive and anxiety symptoms over time [40‐42]. The current study confirms prior research suggesting that co-rumination is common even in early adolescence [41], and although speculative, it is possible that parents may inadvertently model or reinforce co-ruminative tendencies when communicating with their children [46]. The co-occurrence of elevated RNT and co-rumination observed in the current study highlight potential bidirectional relationships between these two transdiagnostic processes, and suggest that co-rumination may be an important target to incorporate in RNT-reduction programs.

When asked to demonstrate conceptual understanding of RNT, adolescents in our study showed comprehensive collective understanding of the multifaceted nature of worry, but limited familiarity with the concept of rumination. In line with findings from an adult sample [35], adolescent definitions of worry from our study referenced cognitive, emotional, and physiological aspects of worry. However, a significant proportion of adolescents were unfamiliar with the term “rumination”, with more than double the percentage of younger compared to older adolescents reporting not having previously encountered the concept. This difference may be due to linguistic or developmental differences between younger and older adolescents, and highlights the importance of tailoring language in RNT programs for early adolescents to avoid disengagement. Furthermore, adolescent definitions of rumination mostly referenced cognitive features (e.g., stuck thinking), in contrast to previous qualitative studies with older adolescents and adults, in which emotional features (e.g., distress) were more frequently referenced [36‐38, 56]. The relative absence of emotional features from adolescent definitions of rumination in our sample may be reflective of their unfamiliarity with the concept. Nonetheless, adolescent definitions of worry and rumination in our sample showed considerable overlap in referencing repetitive, excessive, and negative thinking, supporting findings on the similarities between these two thought processes [57‐59] and further strengthening the validity of targeting both processes simultaneously via the umbrella RNT construct.

Adolescent perspectives of RNT content revealed notable similarities and differences between younger and older adolescents. In both age groups, RNT content tended to be slightly more focused on past than future events, consistent with the findings from studies with adult samples [35, 37], although potential overlap between temporal orientations should also be acknowledged. On a more granular level, the same topics were endorsed as the most common content of worry and/or rumination across both age groups: schoolwork, friendships, things adolescents dislike about themselves, and why things happened to them. Nonetheless, several notable discrepancies in content were noted across adolescent age groups. Younger adolescents were more than twice as likely to worry and/or ruminate about social media interactions than their older peers, potentially reflecting their relative unfamiliarity with this mode of interaction or heightened self-consciousness during vulnerable periods [60‐62]. In contrast, compared to their older peers, younger adolescents were markedly less likely to worry and/or ruminate about their health, friendships, romantic relationships, schoolwork, and how they feel, reflecting age-related differences in key stressors and potentially lower self-reflexivity in early adolescence. Nonetheless, substantial proportions of younger adolescents endorsed abstract, over-generalised styles of thinking as common features of their RNT (e.g., almost a third reported commonly worrying and/or ruminating about “why I feel a certain way”). Abstract thought content is a key feature of RNT [63, 64], and our findings suggest this may be detectable even during early adolescence.

Most adolescents in our study reported experiencing RNT in a predominantly verbal format, consistent with common conceptualisations of RNT and worry [11]. These conceptualisations propose that RNT is a verbal strategy that serves to suppress distressing mental imagery of threats [65]. However, our results suggest that this verbal dominance may be less pronounced in younger adolescents, with over a third of younger adolescents in our sample experiencing RNT as more visual than verbal or a roughly equal mix of verbal and visual content, compared to approximately a fifth of older adolescents reporting the same. Previous studies have also noted the presence of imagery in both adolescent and adult RNT [36, 37]. Together, these findings underscore the need for early adolescent RNT interventions to include strategies to reduce both verbal and imagery-based RNT.

Interpersonal and social events were commonly identified as external triggers of RNT by both adolescents and parents in our study, consistent with the findings of prior studies in adolescents and adults [35, 36, 38]. Although we did not explore the extent to which interpersonal RNT triggers experienced by our sample coincided with co-rumination, moderate levels of co-rumination were reported by younger and older adolescents as well as parents, in the context of both friend and parent-child interactions. Additionally, a prior study involving older adolescents identified co-rumination as a common trigger for RNT [36]. Together, these findings further underscore a need for targeted support in helping adolescents develop skills to manage co-rumination as a pathway to reducing RNT. In addition to external RNT triggers, somatic symptoms and fatigue were identified by adolescents and parents as common internal triggers for RNT. However, many adolescents, especially younger adolescents, were unable to identify specific RNT triggers, indicating the importance of psychoeducation and structured self-monitoring to help adolescents detect and respond to emerging RNT more effectively.

A further area warranting attention is the timing and context in which RNT tends to arise. Across both age groups, adolescents most commonly reported engaging in RNT at night or when alone, with likelihood of RNT engagement being lowest in the morning and increasing as the day progresses. These patterns are consistent with findings in adult populations [35] and suggest that the end of the day may represent a particularly vulnerable period for the onset or escalation of RNT. Interventions for early adolescents may therefore benefit from incorporating evening- and/or bedtime-specific coping strategies to help mitigate RNT during these high-risk periods. Such strategies are particularly important given evidence that rumination mediates the association between adolescent insomnia and depression [66], and adolescent sleep disturbance predicts increases in internalising symptoms via both general and pre-sleep RNT [29].

Distraction and social interaction were reported to frequently interrupt RNT, functioning both as passive circuit breakers (e.g., an external event capturing attention) and as active coping strategies (e.g., deliberately shifting focus to another activity). These findings are consistent with and extend upon prior adolescent and adult studies, in which participants also cited distraction and social interactions as common ways to stop RNT [35, 37].

Concerningly, almost half of the adolescents in our study reported not actively engaging in any strategies to reduce their RNT and that nothing stops their RNT. These rates are higher than those reported by adults [35], but echo other adolescent qualitative studies where participants described a sense of helplessness and inertia in the face of RNT [56].

Over half of the adolescents in our study reported that RNT was unhelpful or had no purpose, with only a minority citing benefits such as preparation, self-protection, or emotional processing. These findings are consistent with the predominantly negative views depressed adolescents expressed regarding their rumination in a previous qualitative study [36], but diverge from the findings of adult studies in which higher percentages of participants endorsed positive metacognitive beliefs regarding their RNT [35]. Since positive and negative metacognitions can both contribute to RNT [67], interventions could aim to help younger adolescents gain insight into the function of their RNT, as a way to help them develop adaptive strategies to manage their RNT.

The current mixed methods study provides a comprehensive and nuanced understanding of adolescent experiences of RNT by integrating quantitative and qualitative insights from both adolescent and parental perspectives. Findings from the current study form a solid foundation for the design of an engaging and developmentally appropriate transdiagnostic program to reduce RNT, co-rumination, and prevent depression and anxiety in early adolescence. Adolescents and parents described a range of anticipated benefits from reducing RNT, including improved mood, better sleep, increased mental clarity, and enhanced ability to engage in meaningful activities. These benefits can be powerfully leveraged to enhance program engagement among both adolescents and parents.

Several limitations of the current study should also be acknowledged. First, most adolescent and parent participants did not form matched dyads, which precluded direct comparison between adolescent self-reports and parental perceptions due to insufficient power, and limited the interpretability of parent-child findings in the current study. Future studies may benefit from examining within-dyad concordance to better understand divergence in experienced versus observed adolescent RNT. Second, although the study recruited from the general community, the sample was skewed towards a culturally homogenous and predominantly female cohort of adolescents with clinically elevated symptoms of depression and anxiety and self-identified neurodivergence, potentially limiting generalisability to other populations. The predominantly clinical and neurodivergent nature of our self-selecting sample may reflect a participation bias among adolescents and families with existing concerns about mental wellbeing. Third, although the study aimed to recruit broadly across both younger and older adolescents, our adolescent-report sample consisted of significantly more older than younger adolescents. Therefore, caution must be exercised when extrapolating insights from our findings to inform the design of early-adolescent-specific interventions. Fourth, given the cross-sectional and retrospective self-report nature of data collected, causal directionality between variables cannot be established and recall bias may limit the generalisability of findings. Fifth, the extent to which participants distinguished between the constructs of rumination and worry was not explored, and both quantitative and qualitative results may contain conceptual overlap between these constructs.

The current study has provided valuable insights into how young people aged 10–18 perceive and manage RNT. RNT and the related phenomenon of co-rumination are commonly experienced in adolescence, from as early as age 10. RNT causes distress to young people often lacking in effective coping strategies and warrants further attention alongside co-rumination as transdiagnostic intervention targets in preventive mental health programs.