Moderators of the Association Between Co-Rumination and Depressive Symptoms in Emerging Adult Friendships

Moderators of the Association Between Co-Rumination and Depressive Symptoms in Emerging Adult Friendships

Emotion regulation is a set of psychological processes involved in the up- and down-regulation of emotional experiences (Aldao, 2013; Diamond & Aspinwall, 2003; Gross, 2014). For example, rumination is an individual emotion regulation strategy whereby individuals perseverate on negative emotions and perceive problems as worse than they are and unsolvable (Rose et al., 2017). It is consistently and strongly positively associated with depressive symptoms (Piraman et al., 2016). Conversely, cognitive reappraisal involves reinterpreting distressing situations to change their emotional impacts (Sachs-Ericsson et al., 2021) and is negatively correlated with depressive symptoms (Aldao, 2013).

When shared between close individuals, the up- and down-regulation of emotions creates a sense of co-regulation (Diamond & Aspinwall, 2003). Co-rumination is a co-regulation process that occurs when two people ruminate together (Rose, 2002). The association between co-rumination and depressive symptoms is less consistent than it is for individual rumination and depressive symptoms. Because co-rumination is a relational process that involves more than one person regulating their own and their partner’s emotions, differences in friends’ habitual use of a broader range of strategies may account for discrepancies in patterns of association between co-rumination and individual depressive symptoms. However, scant attention has been paid to the influence of friend factors, such as the tendency for friends to ruminate or reappraise their emotional experiences, on the association between co-rumination and depressive symptoms. To address this gap in the literature, the current study assessed whether the association between co-rumination and depressive symptoms is moderated by friend emotion regulation strategy use. We assessed these effects in a sample of emerging adult friendship pairs because change in social relationships, emotion regulation habits, and risk for mental health problems characterize this pivotal period of the life course, linking late adolescence and adulthood.

Developmental systems theory (Lerner, 1998) and its extensions to psychopathology (Sroufe, 1997) hold that individual developmental outcomes, including during the emerging adulthood period of the life course (see Schulenberg et al., 2004), result from transactional processes between individuals and their environments. Starting in infancy, emotions are regulated through transactional processes between children and caregivers (Deater-Deckard et al., 2016). Across childhood and adolescence, transactional processes elaborate into and shape emotion regulation capacity. For example, early experience shapes one’s coping behaviors, cognitions, and dispositions that filter and direct emotional experience (Gross, 2014). In their developmental account, Diamond and Aspinwall (2003) view instances of emotion regulation at particular points in time as discrete expressions of a broader capacity to flexibly activate, coordinate, and direct emotional states within and across developmental periods. From this view, emotion regulation capacity is considered an integral component of the self-regulatory processes that direct development more generally across the entire life course (Carver & Scheier, 1990; Diamond & Aspinwall, 2003; Heckhausen et al., 2010).

Although shaped by earlier experience, emotion regulation capacity continues to develop across the transition to adulthood and the transition from adolescence to adulthood is a period of growth in individual emotion regulation habits (Riediger & Klipker, 2014; Zimmermann & Iwanski, 2014). In turn, this period is characterized by variability in emotional experience, with emotional well-being maturing on average (e.g., Galambos et al., 2006; Ross & Mirowsky, 2008; Schulenberg & Zarrett, 2006). At the same time, prevalence rates of mental health problems that involve the regulation of emotions, such as depression, peak during emerging adulthood relative to earlier in adolescence (Rohde et al., 2013) and to later points in the adult life course (Pearson et al., 2013). Co-rumination with friends during this period may play an important role in predicting heterogeneity in emotional experience.

Compared to later in adulthood, emerging adult social networks tend to be larger than those in later periods of the adult life course (Antonucci et al., 2010). As individuals navigate their way through these networks, they may establish new friendships that are more or less beneficial to their mental health and well-being. Friends are the most preferred confidants during the transition from late adolescence to young adulthood and fulfill all aspects of social support, including emotional support (Carbery & Buhrmester, 1998). However, interpersonal difficulties, such as not being sure how to support friends struggling with their mental health or receiving unsolicited advice perceived as critical or condescending, are frequent during the emerging adulthood period (Kirmayer et al., 2021). In addition, Bagwell et al. (2005) found that higher levels of negative friendship interaction features, such as frequent conflict and antagonism, were associated with clinical symptomatology. Others have similarly found that friendship instability during this period predicts depressive symptoms (Lapierre & Poulin, 2020). These results suggest that friendships are important contributors to emerging adults’ mental health. It is surprising then that little research has examined co-rumination in the context of emerging adult friendships.

With respect to friendship quality in adolescence, Rose et al. (2017) found prospective associations between co-rumination and friendship quality, in which higher friendship quality exacerbated risk for depressive symptoms by further increasing co-rumination, suggesting a reinforcing effect of friendship quality. However, using moderation analyses, Guassi Moreira and colleagues (2016) found that high levels of co-rumination were only associated with depressive symptoms among first year college students who reported being in low-quality relationships, suggesting a protective effect of higher friendship quality.

The disparate findings on the functional role of friendship quality on the association between co-rumination and depressive symptoms may be explained by friends’ individual emotional habits. Individual emotion regulation capacity increases and social networks expand in emerging adulthood. Thus, whether co-rumination results in an increase in depressive symptoms may depend not only on the quality of the friendship in which co-rumination occurs (Rose & Rudolph, 2006), but qualities of the friends themselves. That is, it may not be friendship quality that determines whether co-rumination is adaptive or harmful, but the emotion regulation habits of the friend with whom an individual is co-ruminating with. For example, Afifi et al. (2013) found that when young adults co-ruminated with a friend who responded in an unsupportive manner, anxiety increased. In contrast, supportive remarks prevented the up-regulation of anxiety. Similarly, the types of emotion regulation strategies that a friend tends to employ on their own may augment the association between co-rumination and depressive symptoms for their peer.

Cognitive reappraisal, an emotion regulating strategy involving reframing emotions in a more positive light, is negatively correlated with psychopathology (Aldao, 2013; Aldao et al., 2010; Cludius et al., 2020) and positively correlated with psychological well-being (Gross & John, 2003). Reappraisal often results in the desired changes in self-reported emotion (Gross, 1998; Jackson et al., 2000), peripheral physiology (Denson et al., 2011; Dillon & LaBar, 2005; Ray et al., 2010), and neural measures of emotion (Chang et al., 2015; Dörfel et al., 2014; Hajcak & Nieuwenhuis, 2006; Kalisch et al., 2005; Lohani & Isaacowitz, 2014; Ochsner et al., 2002; Schaefer et al., 2002; Shahane et al., 2019). These findings gave rise to our first hypothesis: co-ruminating with a friend who relies heavily on reappraisal to manage their experiences of negative emotions and encourage problem-solving may protect oneself from depressive symptoms.

Conversely, meta-analyses on rumination, a form of perseverative cognition that focuses on negative content, have established that rumination has a large positive correlation with symptoms of psychopathology, with an especially large effect on depressive symptoms (Aldao et al., 2010; Compas et al., 2017). Rumination is also more strongly related to overall psychopathology than other maladaptive strategies, such as avoidance and suppression (Aldao et al., 2010). Further, rumination is negatively related to problem solving (Hong, 2007). In light of these findings, we present a second hypothesis: co-ruminating with friends who ruminate on their own may increase one’s likelihood of experiencing depressive symptoms. Individuals who rely heavily on rumination to manage their emotions may drive the co-rumination process by encouraging their friend to dwell on problems and focus on the negative, while failing to find solutions. In turn, this might perpetuate the cycle of co-rumination and lead their co-ruminating peers to be especially vulnerable to depressive symptoms.

Same-gender friendships are specifically important for emerging adults (see Barry et al., 2016). They are more common than heterosocial friendships (Mehta & Strough, 2009) and many emerging adults report a preference for their same-gender friendships (Baumgarte & Nelson, 2009). Emerging adults are also more likely to self-disclose to same-sex friends than parents or romantic partners (Derlega et al., 2011). Hence, gender may also act as a moderator in the link between co-rumination and mental health outcomes in emerging adults. Indeed, Rose et al. (2007) found that co-rumination predicted depressive symptoms in girls, over a 6-month follow-up period, but not in boys. Similarly, Bastin et al. (2018) found that co-brooding, a passive component of co-rumination characterized by the tendency to dwell on negative feelings within a dyadic relationship, predicted high levels of concurrent and future depressive symptoms in girls. In contrast, boys who engaged in co-brooding were more likely to report fewer depressive symptoms. However, other studies have not found gender differences. Cross-sectional studies with teenagers (Hankin et al., 2010; Rose, 2002) and a prospective study with a college sample (Calmes & Roberts, 2008) discovered that the relation between co-rumination and depressive symptoms does not vary as a function of gender, suggesting that co-rumination remains maladaptive for boys, the same way that it does for girls (Stone & Gibb, 2015). Therefore, it remains unclear whether the effects of co-rumination operate the same way in same-sex female dyads as it does in same-sex male dyads or mixed dyads. The examination of individual emotion regulation tendencies may help to clarify these disparate findings.

Method

Results

Prior to data analysis, data were screened to ensure that all dyads consisted of actual friendship pairs. Dyads were removed if either the target participant or the friend selected the lowest possible option on the screening questions about how well they knew each other and how close they were. Based on this screening, eight pairs were removed. Another pair was also removed as one friend did not provide data on these questions. 15 additional pairs were excluded because at least one of the friendship members reported either being currently or previously in a romantic relationship with their friend, resulting in 87 friend pairs (174 participants).

Means and bivariate correlations are presented in Table 1. Target participant co-rumination and depressive symptoms were not correlated. The correlations between target participant reports of friendship quality and their friend’s report of reappraisal (r = −.23, p < .05) and between target participant reports of co-rumination and their friend’s report of reappraisal (r = −.29, p < .01) were significant and negative, suggesting that individuals were more likely to perceive their friendship quality as higher and co-ruminate more when their friend did not rely heavily on reappraisal. Moreover, the correlation between target participant co-rumination and friendship quality was significant and positive (r = .39, p < .01).

OPEN IN VIEWER

Table 1.

Intercorrelations, Means, Standard Deviations, and Friend Pair Gender Differences.

	Target Participant: Depressive Symptoms	Target Participant: Co-Rumination	Target Participant: Friendship Quality	Friend reappraisal	Friend rumination
Target participant: Depressive symptoms	1.00	—
Target participant: Co-rumination	−.02	1.00	—
Target participant: Friendship quality	−.02	.39**	1.00	—
Friend reappraisal	−.02	−.29**	−.23*	1.00	—
Friend rumination	.04	.19	.06	−.20	1.00
Mean	0.78	3.42	7.08	4.83	2.59
Standard deviation	0.56	0.99	1.01	1.15	0.75
Friend pair gender (female-female = 0, female-male = 1)	−0.17	2.03*	0.52	−1.16	0.88

To assess the moderating influence of a friend’s tendencies to use reappraisal and rumination on the association between target participants’ co-rumination and their own depressive symptoms, we ran a series of structural equation models in Mplus (Muthén & Muthén, 2018). In our first model, we assessed whether the association between the target participants’ report of co-rumination with their friend and their own depressive symptoms was moderated by their friends’ tendencies to use reappraisal or rumination. Depressive symptom scores were regressed onto target participant co-rumination, friend reappraisal, friend rumination, and the two interaction terms: target participant co-rumination and friend reappraisal and target participant co-rumination and friend rumination. The model fit the data well: RMSEA= .00, CFI= 1.00, SRMR= .00. In line with our first hypothesis, target participant depressive symptoms were associated with their reports of co-rumination with the friend they chose for the study (B= 1.40, p= .04), friend’s use of reappraisal (B= 0.85, p= .02), and the interaction between these two variables (B= −1.18, p= .01). Figure 1 shows a visualization of this interaction. However, contrary to our second hypothesis, target participant depressive symptoms were not associated with friend rumination (B= 0.47, p= .31) nor the interaction between target participant co-rumination and friend rumination (B= −0.59, p= .37). See Table 2 for all study results.OPEN IN VIEWER

OPEN IN VIEWER

Table 2.

Reappraisal and Rumination Moderation On The Association Between Co-Rumination And Depressive Symptoms.

	Full sample	Female-Female Pairs	Female-male Pairs
	B	B	B
Co-rumination	1.40*	1.91*	0.61
Reappraisal	0.85*	1.05*	0.76
Rumination	0.47	0.87	−0.23
Reappraisal * Co-rumination	−1.18*	−1.46*	−0.86
Rumination * Co-rumination	−0.59	−1.07	0.16

Note. *p < .05 B = standardized coefficient.

This model was then run separately for the female-female and female-male friend pairs to assess whether this effect was maintained across gender make-up of friendships. For the female-female pairs, the effects were slightly amplified: target friend co-rumination (B= 1.91, p= .02), friend reappraisal (B= 1.05, p= .03), and the interaction between these two variables (B= −1.46, p= .02) were associated with depressive symptoms. However, in the female-male pairs, no paths were significant.

We assessed whether these results were consistent when including the target participant perceptions of the quality of their friendship with the friend they brought to the study. Therefore, these models were repeated adding two paths: target participant depressive symptoms were regressed onto target participant report of friendship quality and the interaction between their report of friendship quality and their report of co-rumination. The model fit the data well (RMSEA= .00, CFI= 1.00, SRMR= .00). Contrary to our third hypothesis, friendship quality did not significantly moderate the effect of co-rumination on depressive symptoms. Moreover, results showed that although target participants’ reports of co-rumination were no longer significantly associated with their own depressive symptoms, friend reappraisal (B= 0.87, p= .02), and the interaction between reappraisal and co-rumination (B= −1.21, p= .01) remained significant. Finally, this model was then re-run separately for the female-female and female-male pairs. Again, target friend co-rumination was no longer significantly associated with depressive symptoms, but friend reappraisal (B= 1.02, p= .04) and the interaction between reappraisal and co-rumination (B= −1.44, p= .02) remained significant for the female-female dyads. Lastly, no paths were significant in the female-male pairs. See Table 3 for all results.

OPEN IN VIEWER

Table 3.

Reappraisal, rumination, and friendship quality moderation on the association between co-rumination and depressive symptoms.

	Full sample	Female-Female Pairs	Female-male Pairs
	B	B	B
Target participant: Co-rumination	1.61	1.91	1.86
Target participant: Friendship quality	0.09	−0.07	0.80
Friendship quality * Co-rumination	−0.21	−0.01	−1.32
Friend reappraisal	0.87*	1.02*	0.63
Reappraisal * Co-rumination	−1.21*	−1.44*	−0.72
Friend rumination	0.5	0.85	0.34
Rumination * Co-rumination	−0.64	−1.05	−0.66

Discussion

The current study aimed to address the inconsistent empirical findings with respect to the role of friendship quality for co-rumination and depressive symptoms. Guided by a transactional relational model derived from developmental systems (Lerner, 1998), psychopathology theories (Sroufe, 1997), and their extension to the emerging adulthood period of the life course (Schulenberg et al., 2004), we examined the potential moderating effects of friend cognitive reappraisal and rumination on the association between co-rumination and depressive symptoms within emerging adult friendships. We expected that high levels of friend cognitive reappraisal use would buffer the link between co-rumination and depressive symptoms (Hypothesis 1). Our findings supported this hypothesis. We also hypothesized that high levels of friend rumination use would exacerbate the relation between co-rumination and depressive symptoms (Hypothesis 2). We did not find support for this hypothesis. Overall, results demonstrated that the extent to which a co-ruminating individual experienced depressive symptoms depended on the extent to which their friend relied on reappraisal, not rumination, to manage their emotional experiences. Contrary to past research suggesting that the presence of a maladaptive emotion-regulation strategy (e.g., rumination) is more deleterious than the absence of an adaptive emotion-regulation strategy (e.g., reappraisal), results from the current study indicates a stronger association between reappraisal (vs. rumination) and depressive symptoms.

In light of previous findings showing a direct link between co-rumination and depressive symptoms (e.g., Rose, 2002; Piraman et al., 2016), results from the current study show that friend reappraisal and co-rumination interacted to predict depressive symptoms. This result supported our first hypothesis: emerging adults who co-ruminate with friends who regularly use cognitive reappraisal to manage their emotions are protected from depressive symptoms, perhaps because of their friends’ problem-solving habits. This finding indicates a more complex relation between co-rumination and depressive symptom that fits with the developmental psychopathology model (Cicchetti, 2006; Lerner, 1998; Schulenberg et al., 2004; Sroufe, 1997). This theory holds that individual developmental outcomes result from transactional processes between individuals and their environments. In line with this perspective, the results from our study suggest that within a relational context, individual depressive symptoms depend upon emotion regulation habits of the members involved in the co-ruminative discussion. Furthermore, developmental models of emotion regulation consider emotion regulation capacity as an integral component of the self-regulatory processes that direct development more generally across the lifespan (Carver & Scheier, 1990; Diamond & Aspinwall, 2003; Heckhausen et al., 2010). Our findings suggest that friendships are an important context within which this capacity is shaped during the emerging adulthood period

Past research also supports our finding about the nature of the interaction between co-rumination and cognitive reappraisal in predicting depressive symptoms. In particular, we found that reappraisal plays a role in attenuating depressive symptoms. In line with this result, McRae and Gross’ (2020) review presents compelling evidence for the positive impacts of cognitive reappraisal on well-being and mental health. Additionally, Sachs-Ericsson et al. (2021) found that cognitive reappraisal can also act as a protective factor in the relation between lower social support and depressive symptoms.

Given that rumination has long been associated with psychopathology (Aldao et al., 2010; Compas et al., 2017), our second hypothesis stated that friend rumination would amplify the association between co-rumination and depressive symptoms. This hypothesis was not supported. Results showed that the link between dyad-level co-rumination and target friend depressive symptoms did not depend on the extent to which their friend ruminated. Thus, friend individual rumination may not exacerbate the link between emerging adults’ co-rumination and depressive symptom levels. That is, the negative consequences of co-ruminating might not necessarily be made worse when done with friends who regularly ruminate on their own, given the already negative nature of co-rumination. This finding also suggests that it is the relational aspect of co-rumination, or the act of ruminating together, that may be particularly important for depressive symptoms in emerging adult friendships.

Another finding from the current study emphasizes the role of relational aspects of emerging adult friendships for depressive symptoms. Our results showed that the correlations between friendship quality and friend’s reappraisal, as well as the correlation between co-rumination and friend’s reappraisal, were significant and negative. These results suggest that emerging adults are more likely to perceive their friendship quality as lower and co-ruminate less when their friend relies heavily on reappraisal to manage their own emotions. One possibility is that emerging adults who habitually reappraise their negative emotions discourage the use of co-rumination with their friends, a co-regulation experience that tends to make friends feel connected to one another. Indeed, the association between co-rumination and friendship quality was significant and positive, supporting that this coping style increases feelings of social connectedness among college students (Weaver & Bosson, 2011) and adults (Keshishian et al., 2016). Conversely, findings from adolescence samples have shown that higher rates of co-rumination are associated with negative friendship quality (McLaughlin & Nolen-Hoeksema, 2012), a decreased number of friends over time (Starr & Davila, 2009), less social acceptance (Tompkins et al., 2011) and more interpersonal stress (Rose et al., 2017). More longitudinal or cohort studies are needed to investigate whether co-rumination differently affects the friendship quality of adolescents when compared to emerging adults.

Lastly, the present study revealed some interesting gender patterns. When models were run separately for female-female and female-male dyads, the pattern of results was quite different. The interaction between co-rumination and reappraisal remained significant for female-female pairs, but no paths were significant for female-male pairs. This result may suggest that the impacts that co-rumination and friend reappraisal have on emerging adults’ mental health is more important for young women. Another possible explanation is that women experience higher levels of depressive symptoms because they are more likely to co-ruminate than men. In fact, co-rumination is enacted primarily among female dyads (Calmes & Roberts, 2008; Hankin et al., 2010; Rose, 2002; Rose et al., 2007; Tompkins et al., 2011; Waller & Rose, 2013; White & Shih, 2012), perhaps because of the way that young women learn about communicating. Whereas females are often encouraged to talk about their emotions, the opposite is true for males. For example, boys sometimes tease each other if one discloses information about their feelings, which discourages them from talking about emotions (Miller-Slough & Dunsmore, 2016). Therefore, future research is needed to further investigate whether the impact of co-rumination differs across same-gender and mixed-gender friendships.

Conclusion

Taken together, results from the current study point to promising directions for understanding emotion regulation strategies enacted within emerging adult friendships. Specifically, findings suggest that friendships are important contexts for regulating emotions, and that these co-regulation behaviors have important associations with well-being for young adults. Our findings demonstrate that the link between co-rumination and depressive symptoms depended on individual propensities to engage in cognitive reappraisal, but not on rumination. Future studies should aim to assess other potential moderators, such as stress. A significant contribution of our study is that it took into account individual-coping strategies when evaluating the link between co-rumination and depressive symptoms. It also represents an improvement over past studies by using independent reports of symptoms and coping. Furthermore, conducting this research on emerging adults was an important step as most previous studies have only assessed the relation between co-rumination and depressive symptoms in early or mid-adolescence. Overall, the current study filled an important gap in the field about relational factors that can strengthen versus weaken the extent to which co-rumination leads to depressive experiences in emerging adults.