Introduction
Adolescence is a crucial developmental stage in terms of mental health difficulties (Polanczyk et al.,
2015). Physiological and psychosocial changes that occur during puberty help increase internalizing symptoms—such as depression, anxiety, and stress (Graber,
2013; Romeo,
2013)—with prevalence rates higher among girls than boys (Merikangas et al.,
2010). Considering that mental health problems in youth are associated with mental health problems in adulthood, both predicting and exacerbating them (Copeland et al.,
2020; Johnson et al.,
2018), preventing these first negative mental health symptoms during adolescence is a matter of public concern. Evidence suggests that dispositional mindfulness is associated with adolescents’ mental health and well-being (Pallozzi et al.,
2017) and predicts lower levels of internalizing symptoms over time (Cortazar & Calvete,
2019). However, few studies have examined the temporal relationships between other personal factors related to mindfulness and negative mental health symptoms. This article focuses on impulsivity and rumination, considering that both constructs are associated negatively with mindfulness in adolescents (de Bruin et al.,
2014; García-Rubio et al.,
2019) and are viewed as transdiagnostic vulnerability factors that contribute to the etiology and maintenance of internalizing symptoms throughout adolescence (Cosi et al.,
2011; McLaughlin & Nolen-Hoeksema,
2011). This study used a two-wave longitudinal design to examine the temporal bidirectional associations between mindfulness, rumination, impulsivity, and internalizing symptoms. Furthermore, the model’s sex invariance was tested.
Mindfulness is “the awareness that emerges through paying attention on purpose, in the present moment, and nonjudgmentally to the experience moment by moment” (Kabat-Zinn,
2003, p. 145). Mindfulness can be a state, training, or disposition (Brown et al.,
2011). This article focuses explicitly on mindfulness as a trait or disposition that varies among individuals and can be promoted through training (Brown et al.,
2011). Mindfulness theories state that acting with awareness, a core component of mindfulness, produces its salutary effects by minimizing the automatic, habitual, and impulsive reactive patterns of responding, replacing them with more conscious and adaptive responses to events (Brown et al.,
2007; Williams & Kabat-Zinn,
2013). In this sense, the evidence suggests that acting with awareness is incompatible with getting stuck in repetitive thoughts, which is termed a ruminative thinking pattern (Jury & Jose,
2019). Also, it has been proposed that mindfulness reduces prepotent responses by making the individual aware of their impulsive tendencies (Peters et al.,
2011). Therefore, mindfulness provides space for someone to think before they react by recognizing and having the chance to interrupt the maladaptive cognitive and behavioral patterns that increase suffering and negative mental health symptoms (Williams & Kabat-Zinn,
2013). This is particularly relevant for adolescents; thus, it seems relevant to examine the association between mindfulness and symptoms, and how they are related to other factors, namely rumination and impulsivity.
Rumination
Rumination refers to an emotion-regulation strategy that involves responding to distress repetitively and passively focusing on its symptoms, causes, and consequences (Nolen-Hoeksema et al.,
2008). It is well-documented that rumination is a cognitive vulnerability factor in adolescent psychopathology (Abela et al.,
2012; McLaughlin & Nolen-Hoeksema,
2011; Wilkinson et al.,
2013), particularly among girls (Rood et al.,
2009). Moreover, several longitudinal studies have reported a bidirectional association between psychopathology and rumination (Jose & Weir,
2013; Krause et al.,
2018). Similarly, in line with transactional psychopathology models, it has been suggested that rumination and stressors increase depressive symptoms—which, in turn, increase stress—and both predict long-term rumination (Padilla & Calvete,
2015).
As one might expect, and because a ruminative thinking pattern seems incompatible with acting with awareness, rumination is associated negatively with mindfulness (Chambers et al.,
2015; Pallozzi et al.,
2017; Yu et al.,
2021). Indeed, a study among adolescents found that while rumination aggravated the association between stressors and internalizing symptoms, mindfulness attenuated this association (Marks et al.,
2010). Furthermore, a recent study found that mindfulness reduced rumination’s impact on symptoms (Blanke et al.,
2020). Together, these studies highlighted the need to address rumination as a relevant aspect associated with psychopathology and mindfulness during adolescence.
Several studies have examined rumination as an underlying mechanism through which mindfulness relates to psychological symptoms. Two cross-sectional studies among adolescents suggested that rumination mediates the association between mindfulness and symptoms. One study found that mindfulness directly impacted adolescents’ internalizing symptoms and indirectly improved their symptoms by reducing rumination (Yu et al.,
2021). Similarly, another study indicated that brooding, a maladaptive form of rumination, mediated the association between mindfulness and depressive symptoms, and also was related with less awareness (Alleva et al.,
2014).
Evidence from longitudinal studies is mixed. One study did not find that mindfulness protected adolescents from rumination long-term (Royuela-Colomer & Calvete,
2016), while other studies found evidence of rumination as a mediator between mindfulness and symptoms in adolescents (Ciesla et al.,
2012) and adults (Jury & Jose,
2019). A similar study among adolescents found that rumination mediated the association between mindfulness and symptoms, and that this association was reciprocal (Tumminia et al.,
2020). These authors emphasized the importance of studying the complex processes that arise in adolescence and comprise a synchronization of regulatory, affective, and motivational systems, considering that it is essential to understand how these cognitive-control regulatory and reactivity processes influence mindfulness. Overall, these results indicate the need to clarify longitudinal associations between negative mental health symptoms and mindfulness and rumination, including other potential factors related with symptoms, mindfulness, and rumination, such as impulsivity.
Impulsivity
Impulsivity refers to “a predisposition toward rapid, unplanned reactions to internal or external stimuli without regard to the negative consequences of these reactions to the impulsive individuals or others” (Moeller et al.,
2001, p. 1784). Impulsivity is related strongly with other concepts, such as cognitive control or executive functions, mainly inhibitory control (i.e., the capacity to inhibit an automatic or prepotent thought, emotion, or behavior in favor of a more desirable response; Diamond,
2013), and sometimes both terms are used interchangeably. Indeed, some authors have suggested that inhibitory control deficits underlie impulsivity (Horn et al.,
2003; Leshem,
2016).
Much extant research supports the hypothesis that impulsivity is associated with externalizing symptoms in adolescence (Fosco et al.,
2019; Stautz & Cooper,
2013). However, little research has been conducted on the association between internalizing symptoms and impulsivity. For example, a recent study reported a higher level of impulsivity in depressed adolescents compared with mentally healthy ones (Onat et al.,
2019). Furthermore, several cross-sectional studies on young adults found a positive correlation between impulsivity and depressive symptoms, anxiety, and stress (Moustafa et al.,
2017; Yu et al.,
2020). Regarding sex differences, boys tend to be more impulsive than girls (Chapple & Johnson,
2007), and research has indicated a stronger association between impulsivity and depressive symptoms in adolescent boys than girls (Regan et al.,
2019). Moreover, some researchers suggested that emotion-regulation strategies, such as rumination, mediate the association between impulsivity and depressive symptoms (d’Acremont & Van der Linden,
2007). Similarly, a longitudinal study reported an association between impulse control difficulties, mindfulness, anxiety, and depressive symptoms among college students, and this association was mediated by limited access to emotion-regulation strategies (Cheung & Ng,
2019). These studies emphasize the need to examine the association between impulsivity and symptoms during adolescence, particularly longitudinal designs.
Concerning the association between impulsivity and mindfulness, the literature suggests a negative association between these two variables (for a review Lu & Huffman,
2017). Most of these studies are cross-sectional and comprise college student samples (Lyvers et al.,
2014; Maltais et al.,
2020; Murphy & MacKillop,
2012; Peters et al.,
2011), and to our knowledge, only one study has examined this correlation among adolescent samples (García-Rubio et al.,
2019). Furthermore, three studies among adolescents reported that mindfulness was related with better inhibitory control, with both gauged using self-report and behavioral measures (Oberle et al.,
2012; Riggs et al.,
2015; Shin et al.,
2016). Finally, some studies have suggested that impulsivity and inhibition deficits are associated with rumination in adolescence. For example, one study found that rumination was associated with difficulty inhibiting negative information (Hilt et al.,
2014). Moreover, a longitudinal study reported a bidirectional association between rumination and impulsivity among college students, and that both helped intensify depressive symptoms (Hasegawa et al.,
2018). Therefore, it is crucial to target impulsivity as a hallmark of psychopathology in adolescence.
So far, the studies presented highlight the need to include rumination and impulsivity when examining longitudinal associations between mindfulness and symptoms throughout adolescence, considering that both have been considered transdiagnostic factors for adolescent psychopathology (Cosi et al.,
2011; McLaughlin & Nolen-Hoeksema,
2011). Furthermore, in line with transactional psychopathology models—which suggest that the associations between stress, cognitive vulnerabilities, and negative mental health symptoms are bidirectional and influence each other (Calvete et al.,
2015)—it is crucial to examine how symptoms and vulnerability factors influence each other, and how they influence and are influenced by mindfulness levels. Thus, understanding longitudinal associations between mindfulness, internalizing symptoms, rumination, and impulsivity is essential to promoting mental health and developing adequate treatment strategies during adolescence.
Discussion
Previous studies suggest that mindfulness predicts fewer internalizing symptoms in adolescents (Cortazar & Calvete,
2019). According to mindfulness theories, acting with awareness reduces automatic, habitual, impulsive cognitive and behavioral responses (such as rumination and impulsivity) by facilitating a more flexible and adaptive response to events (Brown et al.,
2007; Williams & Kabat-Zinn,
2013). Nevertheless, when studying longitudinal associations between mindfulness and negative mental health symptoms, few studies have included its associations with rumination and impulsivity, which are viewed as transdiagnostic factors for psychopathology throughout adolescence (Cosi et al.,
2011; McLaughlin & Nolen-Hoeksema,
2011). Therefore, this study examined, over a one-year period, longitudinal associations between mindfulness, rumination, impulsivity, and internalizing symptoms (depressive symptoms, anxiety, and stress) in adolescents and whether the model was sex invariant.
Several interesting findings emerged from the autoregressive, cross-lagged path analysis. First, mindfulness prospectively predicted less stress, depressive symptoms, and impulsivity, but not anxiety or rumination. Second, rumination did not predict any variable in W2. Third, a bidirectional negative association between mindfulness and impulsivity was found. Fourth, impulsivity positively predicted stress, but not rumination, anxiety, or depressive symptoms. Concerning internalizing symptoms, only anxiety predicted other symptoms (i.e., depressive symptoms and stress) and rumination in W2. These results were observed controlling for all variables in W1, i.e., the associations between variables in W1 and W2 are not spurious due to their associations in W1. Thus, the model was sex invariant.
In line with previous studies, this study found that mindfulness was negatively correlated cross-sectionally with internalizing symptoms, rumination, and impulsivity (García-Rubio et al.,
2019; Pallozzi et al.,
2017; Tan & Martin,
2012). Regarding longitudinal associations, the results partially support the hypothesis. Consistent with previous longitudinal studies among youth, mindfulness predicted fewer depressive and stress symptoms and impulsivity in the one-year follow-up (Cheung & Ng,
2019; Cortazar & Calvete,
2019; Royuela-Colomer & Calvete,
2016). It seems that acting with awareness could make adolescents less impulsive and less susceptible to depressive and stress symptoms. Furthermore, individuals high in mindfulness could be more able to notice how the mind reacts to thoughts, sensations, and information, recognizing and avoiding habitual patterns that unconsciously guide behavior, thereby selecting better and healthier ways of responding to experiences. Indeed, some studies suggested that mindfulness buffers stressors’ impact on depressive symptoms among adolescents (Thomas et al.,
2021). Similarly, considering that previous studies reported that mindfulness predicts a reduction in stressors in the long term among adolescents (Calvete et al.,
2017), it could be that having fewer stressors reduces stress and depressive symptoms.
However, contrary to expectations, mindfulness did not predict anxiety or rumination in the long term. Although a previous study among adults reported a longitudinal association between mindfulness and rumination (Jury & Jose,
2019), a study among adolescents did not find a longitudinal association between the acting-with-awareness facet of mindfulness and rumination (Royuela-Colomer & Calvete,
2016). Thus, this discrepancy might suggest a developmental difference in rumination’s role. Furthermore, the results did not indicate that mindfulness predicts anxiety, which contradicts some previous findings (Cortazar & Calvete,
2019). One reason could be that these authors included a combined measure of depressive symptoms and anxiety, and did not examine the effect from mindfulness for each symptom individually. Thus, it is possible that when a longitudinal association between mindfulness and internalizing symptoms is analyzed for each symptom separately, it is significant for some symptoms (stress and depressive symptoms), but not for others (anxiety). Furthermore, this also could suggest an interaction between mindfulness and stress, e.g., individuals with low mindfulness levels and high stress levels could be at higher risk of experiencing more anxiety symptoms. In support of this idea and similar to the results, a recent study did not find a direct longitudinal association between mindfulness and anxiety (Cheung & Ng,
2019). This finding could suggest that other mechanisms or facets of mindfulness, such as non-judging inner experience or non-reactivity, could influence the longitudinal association between mindfulness and anxiety symptoms.
Although rumination was associated cross-sectionally with all the variables, the path analysis results indicate that rumination was not associated longitudinally with any variable. Despite rumination’s importance as a cognitive vulnerability factor for psychopathology in adolescence (McLaughlin & Nolen-Hoeksema,
2011), one study found that rumination longitudinally did not predict internalizing symptoms over and above previous symptoms and mindfulness (Schut & Boelen,
2017). Indeed, this could explain why rumination did not predict the rest of the variables, with the results indicating that rumination in W1 does not explain W2 variables’ variability over and above the rest of the model’s variables. Thus, rumination’s maladaptive role may depend on interaction with other variables. For instance, some studies proposed that rumination could act as a moderator, worsening negative mental health symptoms, e.g., when combined with stressors or previous symptoms (Abela & Hankin,
2011; Ciesla et al.,
2012; Cohen et al.,
2014; Marks et al.,
2010).
For impulsivity, as previous studies suggested (Cheung & Ng,
2019; Maltais et al.,
2020; Peters et al.,
2011), the results confirm a cross-sectional and longitudinal association between impulsivity and mindfulness. Interestingly, this study builds on previous literature using cross-sectional designs by indicating that this association is bidirectional. The obtained results suggest that mindfulness might reduce impulsivity in the long term, and simultaneously, higher impulsivity levels might decrease mindfulness in the long term. Surprisingly, contradicting the hypothesis, impulsivity did not predict depressive symptoms, possibly because other underlying mechanisms influence the association between impulsivity and depressive symptoms. For instance, the results could suggest an indirect effect from impulsivity on depressive symptoms by reducing mindfulness, i.e., impulsivity might predict more depressive symptoms in the long term by lowering mindfulness levels. It should be noted that although the model was invariant across sex, impulsivity predicted depressive symptoms only in boys. Another study also reported a stronger association between impulsivity and depressive symptoms in boys (Regan et al.,
2019). As these authors proposed, boys and girls might experience depressive symptoms differently, and the present study’s results suggest that impulsivity predicts more depressive symptoms in boys. Thus, the sex differences highlight the importance of targeting impulsivity as a risk factor in boys’ depressive symptoms.
Interestingly, it was found that impulsivity predicted more stress symptoms in the long term. Considering that impulsivity is associated commonly with maladaptive behaviors in adolescence—such as alcohol abuse, delinquency, and risky behaviors (Fosco et al.,
2019; Romer,
2010; Stautz & Cooper,
2013)—it could be that more impulsive adolescents engage in maladaptive behaviors, thereby increasing stressors and stress symptoms. To our knowledge, this is the first longitudinal study that has examined the association between impulsivity and stress symptoms among adolescents, but future longitudinal studies are needed to understand these results and examine whether other mechanisms are influencing this association.
While depressive symptoms and stress did not predict other variables longitudinally, anxiety symptoms predicted depressive and stress symptoms and rumination. Thus, it can be suggested that anxiety symptoms are a risk factor for other internalizing symptoms and rumination in the long term. The explanation of anxiety as a risk factor is consistent with direct causal anxiety models, suggesting that anxiety precedes depression in adolescence (Mathew et al.,
2011), and the results extend this to stress and rumination. Indeed, a limitation from previous studies is that internalizing symptoms are studied as a unitary construct, or only including depressive symptoms, and that depression and anxiety are different and follow different developmental trends (Graber,
2013). Specifically, it has been suggested that anxiety increases throughout childhood, while depressive symptoms increase during adolescence (Graber,
2013). The findings also suggest that anxiety predicts rumination, with previous studies indicating that rumination could be viewed as a transdiagnostic risk factor during adolescence (McLaughlin & Nolen-Hoeksema,
2011). However, rumination did not predict any variable in W2, indicating that rumination’s maladaptive role could be better explained with a diathesis-anxiety model (Cohen et al.,
2014). Such models suggest that anxiety interacts with cognitive vulnerabilities, such as rumination, to predict other symptoms. Therefore, it could be that adolescents with high anxiety levels, when combined with rumination, are at a higher risk of depressive symptoms. Future studies should test this hypothesis. This study’s findings are interesting for extending knowledge on developmental models of psychopathology and support previous evidence that anxiety must be a target risk factor that predicts psychopathology.
Contrary to expectations, this study did not find a significant sex difference in the predictive model, but sex differences were found in the variables. As suggested in the literature (Brown et al.,
2011; Graber,
2013; Rood et al.,
2009), the results indicate that girls have higher anxiety and rumination levels, and lower mindfulness levels at both time points, whereas differences in depressive symptoms and stress are only present in W2. One reason that might explain the absence of sex differences for depressive symptoms and stress in W1 could be because, as opposed to sex differences in rumination and anxiety appearing earlier during development (Graber,
2013), sex differences in depressive symptoms appear later, during adolescence, and it could be the same for stress symptoms. It also should be noted that, as opposed to previous studies (Chapple & Johnson,
2007), no sex differences in impulsivity were found at any time point. However, the results indicate that the predictive association between impulsivity and depressive symptoms was significant only in boys, suggesting that impulsivity could be a mechanism underlying depressive symptoms in boys, but not in girls.
The most important limitation in this study is attrition rate. Regarding the longitudinal nature of the research and the long time interval studied (1 year), some participants failed to complete the second assessment. Moreover, significant differences in depressive symptom levels and age were found between those who completed both time waves and those who only completed W1, compelling us to be cautious about generalizing these results to younger adolescents with higher depressive symptoms levels. A second limitation is that all participants were from only one high school, and most had medium/high SES, limiting the results’ generalizability. Third, this study used only two time points, and future studies should use at least three time points to examine mediation mechanisms. Finally, it should be noted that, similar to other studies’ results,
r-square coefficients were low (Cortazar & Calvete,
2019; Schut & Boelen,
2017). Because adolescence is a complex developmental period, other contextual factors, such as maturation processes or stressors, also could have impacted outcome variables, and future studies should consider other variables. Despite these limitations, the current study has several strengths. First, this study’s longitudinal nature enables examination of bidirectional associations over a one-year period, building on results from previous cross-sectional studies. Second, this research was based on a sample of high school adolescents and builds on previous studies’ knowledge using college student samples. Third, this study builds on previous longitudinal studies that assessed the relationships between mindfulness and negative mental health symptoms by including the associations with two important factors for adolescent development and mindfulness theories: rumination and impulsivity.