Emotional and Cognitive Coping in Relationship Dissolution

Dissolution of a romantic relationship can adversely affect functioning among college students and represents one primary reason for seeking campus counseling. This study examined the associations among common coping strategies and distress following relationship dissolution. Avoidance and repetitive negative thinking (RNT) were significantly associated with distress, and gender was found to moderate the relationship between RNT and distress. Findings suggest that college counselors should consider emotional coping strategies following relationship dissolutions.

relationship dissolution; repetitive negative thinking; avoidance

Researchers examining relationship dissolution have typically emphasized characteristics of the former relationship (e.g., length of the relationship, time since dissolution, alternative relationship availability, commitment level, initiator status) as important predictors of postdissolution adjustment (Field, Diego, Pelaez, Deeds, & Delgado, [18] ; Fine & Sacher, [19] ; Frazier & Cook, [20] ; Sprecher, [55] ). However, other important variables, such as age and coping strategies, have received less attention.

College students, in particular, are quite vulnerable to adverse outcomes following relationship dissolution. Monroe, Rohde, Seeley, and Lewinsohn ([38] ) reported that relationship dissolution was a significant risk factor for the development of a first episode of depression, a common presentation among college students seeking counseling (Gilbert & Sifers, [21] ; Kelly, [29] ). However, adverse consequences of the distress associated with the end of a romantic relationship in college extend beyond depression and may include academic ruin, social alienation, or even suicide (Davis, Shaver, & Vernon, [14] ; Monroe et al., [38] ). As a result, the treatment of emotional distress following a breakup has been identified as an important area of competency among counselors in university counseling centers (Field et al., [18] ; Moller, Fouladi, McCarthy, & Hatch, [37] ). The current study is designed to fill gaps in the existing literature regarding emotional distress in college students after experiencing relationship dissolution. This more specific focus on emotional coping mechanisms and their association with mental health may assist counselors in providing effective services to students seeking counseling. To contextualize this study in the salient literature, we briefly overview emotional processing and coping studies before providing details regarding the study.

In a meta‐analysis, Aldao, Nolen‐Hoeksema, and Schweizer ([1] ) examined the dispositional correlation between acceptance/reappraisal, rumination, and avoidance with psychopathology and stressful life events. Findings indicated that, after stressful events, rumination and avoidance strategies are associated with a range of disorders, demonstrating large positive effect sizes (fail‐safe N = 351 and 104, respectively). Comparatively, acceptance/reappraisal demonstrated small positive effect sizes to psychopathology (fail‐safe N = 6). The very small number of studies, as indicated by the fail‐safe N, needed to create nonsignificance for acceptance/reappraisal strongly suggests that more research in this area is merited.

Although the broader connection of the aforementioned variables is illuminating, more specific examination of associations between coping strategies, recovery, and relationship dissolution in college students is also important. Use of repetitive negative thinking (RNT), emotion approach coping (EAC; in which acceptance and reappraisal are subsumed), and avoidance after stressful events in other populations, although associated with psychopathology, does not account for possible implications in the already vulnerable population of college students. The intersection of stressors already found in college students could be problematic in treatment, because the use of these strategies could pose further complications (Field et al., [18] ; Pritchard, Wilson, & Yamnitz, [51] ). The following sections address the literature on each of these constructs in turn, with special attention to the contexts of college students and relationship dissolution.

Rumination is characterized by continual repetition of negative thoughts about a situation or the self and is often associated with depression and other maladaptive outcomes (Ehring & Watkins, [16] ; Nolen‐Hoeksema, [42] ). Researchers have established a two‐factor model of rumination: a maladaptive, overgeneral factor (RNT) and a more purposeful, problem‐solving factor (Treynor, Gonzalez, & Nolen‐Hoeksema, [59] ). For the purposes of this study, RNT was assessed to facilitate clarity in interpretation. RNT is both a predictor and a symptom of up to 13 different disorders (e.g., generalized anxiety disorder, depression, posttraumatic stress disorder) and often exacerbates distress because of its unproductive characteristics (Ehring & Watkins, [16] ). RNT has not been examined within the context of relationship dissolution. However, research investigating college students and rumination reveals associations with problematic behaviors, such as binge drinking, self‐injury, social dysfunction, and poor sleep quality (Ciesla, Dickson, Anderson, & Neal, [11] ; Hoff & Muehlenkamp, [25] ; Morrison & O'Connor, [40] ; Zawadzki, Graham, & Gerin, [61] ). If rumination contributes to these negative outcomes for college students and is a frequently used strategy after dissolution, further exploration is needed.

Dissolutions can trigger strong emotional reactions. How these emotions are processed and expressed, referred to as EAC, may affect how distress is experienced postdissolution. EAC has two dimensions: emotional processing and emotional expression. Emotional processing refers to efforts to acknowledge, explore meaning, and understand one's emotions; emotional expression refers to attempted communication of one's emotional experience (Austenfeld & Stanton, [6] ; Stanton, Danoff‐Burg, Cameron, & Ellis, [57] ). Although EAC has been established as a potentially effective coping strategy (e.g., with breast cancer and chronic pain; Austenfeld & Stanton, [6] ), there are contradictory findings. For example, in a study of trauma‐exposed participants, EAC was positively related to symptoms of posttraumatic stress disorder (Amstadter & Vernon, [4] ).

In college student samples, EAC is generally associated with positive outcomes. More specifically, greater expression mediated the relationship between distress and growth after a traumatic event, whereas reduced expression was related to depressive symptoms (Kahn & Garrison, [28] ; Linley, Felus, Gillett, & Joseph, [33] ). Similarly, college students in an emotional‐processing condition exhibited better mood after distressing mood induction (Hunt, [27] ). With respect to relationship dissolution, however, EAC has thus far been neglected. Investigating EAC in a college student sample could address a rich, underexplored area in need of research because of the frequency of relationship dissolutions and greater risk factors.

Avoidance involves the use of strategies to escape painful feelings or situations or to avoid situational or contextual reminders of events (Moulds, Kandris, Starr, & Wong, [41] ). Moos and Schaefer ([39] ) theorized two factors of avoidance: cognitive avoidance, or evasion of mental processing such as thought suppression and distraction, and behavioral avoidance, or preventive actions to avoid reminders or feelings. Avoidance in college student populations has been associated with excessive alcohol use, disordered eating, depression, and anxiety (Levin, Lillis, et al., [31] ; MacNeil, Esposito‐Smythers, Mehlenbeck, & Weismoore, [35] ) and is specifically applicable to relationship dissolution. For example, Lepore and Greenberg ([30] ) found that avoidance strategies in college students after relationship dissolution were correlated with short‐term increases in upper respiratory distress. Additionally, undergraduate students who experienced relationship dissolution showed avoidance at significantly higher levels than did medical students, a colloquially stressed sample (Chung et al., [10] ). Given the concerns surrounding avoidance tactics, such as binge drinking, avoidance is a relevant topic to the college counselor.

Some emotional coping strategies are independently investigated in the extant literature; however, more comprehensive examination is needed in the context of relationship dissolutions and college students. Inspection of these constructs jointly and hierarchically may reveal more specific patterns of association in this population. With this aim in mind, we examined two hypotheses. First, after controlling for the former relationship's characteristics, we predicted that a combination of a less approach‐oriented style and greater use of avoidance and RNT would be associated with significant variance in postdissolution distress. Second, because gender might influence coping strategies differentially, we hypothesized that gender would moderate the association between coping strategies and distress.

Because counselors employed in counseling centers emphasize competencies specific to relationship dissolutions (Field et al., [18] ; Moller et al., [37] ), they are uniquely prepared to offer postdissolution treatment aimed at reducing distress and minimizing adverse consequences in college students. Although relationship instability is common in this population, the extant literature has not comprehensively attended to this topic. This oversight could perhaps be attributable to the perception that relationship dissolution is a fairly normative, frequent occurrence. However, considering the presented risk factors for students, the known adverse effects associated with relationship dissolutions, and the recent emphasis on the need for competent services on campus, the importance of postdissolution research should not continue to be neglected.

Participants included 132 young adults (88 women, 44 men) with a mean age of 19.89 years (range = 18–27). Participants' race/ethnicity was diverse, with 46.2% (n = 61) identifying as Caucasian, 21.2% (n = 28) as African American, 19.7% (n = 26) as Latino, 8.3% (n = 11) as Asian, and 4.5% (n = 6) as other. (Percentages do not total 100 because of rounding.) The average participant relationship lasted 19.7 months (SD = 15.7) before dissolution. Mean dissolution occurred 19.0 weeks (SD = 14.9) before participants completed study measures.

In addition to gathering information on the relationship (i.e., length, time since dissolution, initiator status, availability of perceived alternatives, commitment level of each partner), we administered the following standardized measures.

EAC scales. This measure of EAC contains two scales: Emotional Processing and Emotional Expression, with each scale containing four items (Stanton, Kirk, Cameron, & Danoff‐Burg, [58] ). On a 4‐point Likert‐type scale ranging from 1 (I usually don't do this at all) to 4 (I usually do this a lot), respondents indicated how often each coping strategy is used. A sample item for the Emotional Processing scale is “I acknowledge my emotions,” and an item for the Emotional Expression scale is “I take time to express my emotions.” The maximum score on each scale is 16. The sample mean for emotional processing was 12.0 (SD = 2.8), and the sample mean for emotional expression was 10.6 (SD = 3.3); these scores are consistent with those of the normative sample (Stanton et al., [58] ). A Cronbach's alpha of.91 has been demonstrated within a college study sample, and the scales have demonstrated good test–retest reliability and overall validity in a series of four studies (Stanton et al., [58] ). In the current sample, the Cronbach's alpha was.81 for emotional processing and.88 for emotional expression.

Impact of Event Scale (IES) Avoidance subscale. The IES Avoidance subscale (Horowitz, Wilner, & Alvarez, [26] ) consists of eight items measuring behavioral and cognitive avoidance. The response scale contains 4 points ranging from 1 (not at all) to 4 (often), with a maximum score of 32. Sample items from the Avoidance subscale include “I stayed away from reminders of it” and “I tried to remove it from memory.” The current sample yielded a slightly higher mean (M = 22.2, SD = 4.6) in comparison with the normative study on an outpatient sample (M = 18.2, SD = 10.8); however, this difference is not statistically significant and unlikely to be clinically meaningful. The IES has been used in studies relating to distress and relationship dissolutions (e.g., Boals & Klein, [8] ), with details regarding the validity of the IES reported in research by Zilberg, Weiss, and Horowitz ([62] ). In the current sample, the Cronbach's alpha was found to be.86, which approximates previous reports (e.g., α =.90 in Lepore & Greenberg, [30] ).

Perseverative Thinking Questionnaire (PTQ). The PTQ (Ehring et al., [17] ) measures RNT using 15 items with disorder‐independent content. Factor analysis revealed a general measure of RNT; both convergent validity (with the Response Styles Questionnaire; Nolen‐Hoeksema & Morrow, [45] ) and predictive validity have been previously demonstrated (Ehring et al., [17] ). Similarly, Ehring et al. ([17] ) reported excellent internal consistency (.94). Sample items include “The same thoughts keep going through my mind” and “I feel driven to continue dwelling on the same issue.” Participants respond on a 5‐point Likert‐type scale ranging from 0 (never) to 4 (almost always), with a maximum possible score of 60. The current sample mean was 43.5 (SD = 14.4), with a Cronbach's alpha of.96.

Hopkins Symptom Checklist (HSCL). The widely used 58‐item HSCL (Derogatis, Lipman, Rickels, Uhlenhuth, & Covi, [15] ) is an assessment of general distress in outpatient samples. Scores on the HSCL tap into global distress and have demonstrated reliability and validity in many populations, including undergraduates (Derogatis et al., [15] ). Sample items include “Since your breakup how much were you bothered by difficulty speaking with others?” and “Since your breakup how much were you bothered by nausea or stomach upset?” Response options fall on a 4‐point Likert‐type scale ranging from 1 (not at all) to 4 (extremely), with a maximum score of 232. The mean score in this study was 106.5 (SD = 32.4), and the Cronbach's alpha was.97, which is considerably higher than previously reported (previously reported alphas ranged from.82 to.90; Yuan, Guarnaccia, & Hayslip, [60] ).

To ensure that only currently enrolled college students were included, we recruited participants via a psychology department participant web portal (Sona Systems). Screening was conducted to ensure that participants were not married (or formerly married), had experienced relationship dissolution in the past 6 months, and were in that former relationship for at least 2 months. Only individuals meeting the inclusion criteria were enrolled in the study. Following consent, participants completed the study measures while reflecting on their most serious recent relationship dissolution. All participants and their data were treated in accordance with the American Psychological Association ([3] ) ethical code, and this study was conducted with institutional review board approval.

Bivariate correlations among study variables are presented in Table [NaN] . We performed a hierarchical regression to test our hypothesis that coping variables would significantly predict distress, beyond relationship characteristics (see Table [NaN] ). Because age, gender, and race/ethnicity (dummy coded) had significant zero‐order correlations with distress, they were entered in the first block as control variables. This demographic block was associated with 5.0% (adjusted R2) of the variance in distress, F(5, 125) = 2.50, p =.03. Age (β = −.19, p =.03) and gender (β = −.22, p <.01; women reported more distress) were significant independent predictors of distress.

Bivariate Correlations for Relationship Characteristics, Emotional Coping Variables, and Distress

1 Note. N = 132. RNT = repetitive negative thinking. aPoint biserial correlation; coded 1 = women, 2 = men. bCoded 1 = not very committed, 2 = somewhat committed, 3 = strongly committed, 4 = very strongly committed. cCoded 1 = I initiated, 2 = I somewhat initiated, 3 = we both initiated, 4 = my partner somewhat initiated, 5 = my partner initiated. dCoded 1 = strongly disagree about availability of alternatives, 2 = disagree about availability of alternatives, 3 = neutral, 4 = agree about availability of alternatives, 5 = strongly agree about availability of alternatives. *p <.05. **p <.01.

Predictor Correlations With Distress and Summary of Hierarchical Regression Analysis for Variables Predicting Distress

2 Note. N = 132. RNT = repetitive negative thinking. aCoded 1 = women, 2 = men. bCoded 1 = not very committed, 2 = somewhat committed, 3 = strongly committed, 4 = very strongly committed. cCoded 1 = I initiated, 2 = I somewhat initiated, 3 = we both initiated, 4 = my partner somewhat initiated, 5 = my partner initiated. dCoded 1 = strongly disagree about availability of alternatives, 2 = disagree about availability of alternatives, 3 = neutral, 4 = agree about availability of alternatives, 5 = strongly agree about availability of alternatives. *p <.05. **p <.01.

Relationship characteristics (length of relationship, time since dissolution, initiator status, commitment level, and perceived alternatives) were entered into the second block and contributed an additional 6.0% (ΔR2) of variance, ΔF(10, 120) = 2.53, p =.01. Initiator status was the only significant predictor in the second block (β =.25, p <.01), with less perceived initiation of the dissolution associated with greater distress.

The third and final block in the equation included the coping variables (emotional processing, emotional expression, avoidance, and RNT) and explained an additional 33.0% of the variance in distress, ΔF = 23.09, p <.01. As shown in Table [NaN] , only avoidance (β =.20, p <.01) and RNT (β =.53, p <.01) made significant contributions.

We next sought to determine whether gender moderated the association between coping and distress. Four separate tests of moderation were conducted, one for each coping variable (emotional processing, emotional expression, avoidance, and RNT), following the procedures recommended by Cohen, Cohen, West, and Aiken ([10] ). Tests of moderation were not significant for the coping variables measuring emotional processing, emotional expression, and avoidance. However, findings suggested that gender moderated the association of RNT with distress. In Block 1 of the equation, gender and RNT were collectively associated with 42.0% of the variance in distress, F(2, 128) = 48.17, p <.001. The interaction term, in the second block of the equation, was associated with a significant increase in explained variance (β = −.57), ΔF = 4.36, p =.04. Figure [NaN] illustrates the role of gender in moderating the association between RNT and distress.

College students are especially prone to high rates of relationship dissolution, and these breakups can result in the development of significant distressing consequences (Lepore & Greenberg, [30] ; Monroe et al., [38] ). The results of the current study demonstrate that coping strategies contribute to the experience of distress following college students' relationship dissolutions. Gender plays an important role as well. In particular, college women self‐report greater distress following dissolutions, and gender acts as a moderator between RNT and distress.

The results of this study demonstrate that after controlling for demographics and relationship characteristics, avoidance and RNT are significant contributors to global distress after students' relationship dissolutions. These constructs accounted for a 33.0% increase in total variance explained, which strongly underscores the importance of these factors in the consideration of global distress following dissolution of a relationship.

The variables of avoidance and RNT demonstrate clear, lasting effects in the current sample, given that the average time since breakup was 19 weeks prior to participation. Avoidance research and RNT research in college student samples have independently been associated with numerous negative outcomes (Ciesla et al., [11] ; Cribb, Moulds, & Carter, [13] ; Hoff & Muehlenkamp, [25] ; Levin, Hildebrandt, Lillis, & Hayes, [31] ; Morrison & O'Connor, [40] ; Ottenbreit & Dobson, [47] ; Zawadzki et al., [61] ), including disorders recognized by the Diagnostic and Statistical Manual of Mental Disorders (4th ed., text rev.; American Psychiatric Association, [2] ; see also Blalock & Joiner, [7] ; Nolen‐Hoeksema, [43] ). Therefore, the presence of both coping strategies after relationship dissolution could contribute to major difficulties. On the basis of these findings, college counselors should be vigilant for signs of avoidance and RNT, because the use of these coping strategies may complicate the overall clinical picture.

EAC (both emotional processing and emotional expression) showed very little association with distress in the overall model. Literature using these constructs has focused primarily on health research and chronic illness. The nature of the stressor may be very important in the impact of these variables (Anshel, [5] ; Parkes, [48] ). Researchers studying events with a full range of chronicity and severity could provide more information as to the positive and negative aspects of EAC. Use of a college student sample, with relatively few health difficulties and a predetermined stressor, may not capture the construct found in previous studies.

The sole significant predictor in the relationship characteristics block was initiator status. Initiator status has been significantly predictive of distress in several studies, with less participation in the initiation of dissolution correlating with worse mental health outcomes (Davis et al., [14] ; Perilloux & Buss, [49] ; Sprecher, Felmlee, Metts, Fehr, & Vanni, [56] ). However, other researchers have not found this relationship (Locker, McIntosh, Hackney, Wilson, & Wiegand, [34] ; Sbarra, [52] ; Simpson, [54] ). Opposing findings in these studies are not dependent on sample; researchers using college student samples have found both positive and null associations of initiator status and distress. Counselors should be vigilant to greater distress in those students who were not the primary initiators of their breakup.

Findings for gender effects in relationship dissolution have been inconsistent in the dissolution literature. Our study found that, overall, women reported more distress after relationship dissolution, which is consistent with the broader literature on gender differences in self‐reported distress. Some previous findings support this difference (Simon & Barrett, [53] ), but others have found no difference between men and women (Davis et al., [14] ; Field et al., [18] ; Fine & Sacher, [19] ; Locker et al., [34] ; Moller et al., [37] ). Only one study found that men experienced more distress after dissolutions (Hill, Peplau, & Rubin, [24] ). Many of these studies used college samples, indicating that inconsistency in findings is common.

The only significant moderation model was the association between RNT and distress. In adding the Gender × RNT interaction term, the relationship between RNT and distress strengthened, specifically for college women. According to the literature, women, when compared with men, tend to use more emotional regulation strategies in general, specifically rumination, which may lead to more distress (Nolen‐Hoeksema & Aldao, [44] ). Butler and Nolen‐Hoeksema ([9] ) found that rumination mediated the relationship between gender and depression in college students; when rumination was controlled for, the gender disparity in depression disappeared. The moderation found in this study supports other reports in the literature that there are more negative consequences when college women use RNT compared with college men. College counselors treating students after dissolutions should remain aware of RNT‐like behavior in both genders; however, outcomes for women may be more negative if the RNT problem is not addressed.

There were some study limitations. The use of self‐report data might have introduced shared method variance, although the use of a survey approach might have increased authentic responding. Also, the participants included more women than men, but this is a typical occurrence with a target population of young adult college students because more women than men enroll in college (Pollard, [50] ). In addition, it is a consistent finding that college women are more likely than their male peers to seek counseling, creating applicable information for college counselors (Gonzalez, Alegria, & Prihoda, [22] ).

These findings indicate that a preference for certain coping strategies appears to contribute to the experience of distress after relationship dissolution. This information is useful because it may assist in pinpointing specific areas for counselors to address when seeing a student who has recently experienced a breakup. Because other contributing factors such as initiator status are not circumstances that can be changed, these findings provide more information as to symptoms that might be affected. The following are a list of recommendations for specifically addressing emotional coping strategies based on the results of this study.

Women are more prone to RNT than are men. Although men still ruminate about their breakups, RNT appears less impactful on general mental health. On the basis of the results in this study, interventions targeting RNT such as the defusion techniques found in acceptance and commitment therapy (ACT) or challenging irrational thoughts in cognitive behavior therapy might be useful in treatment with women, but men may benefit more from time spent on interventions addressing rumination in conjunction with emotional expression (Hayes, Levin, Plumb‐Vilardaga, Villatte, & Pistorello, [23] ; Masuda, Hayes, Sackett, & Twohig, [36] ).

With regard to age, counselors should first be aware of the heightened risk for distress in college students. Although relationship dissolutions in older adults may result in greater lifestyle transitions (e.g., child custody, financial burden), the social and psychological development in college students come with their own set of complications. Validation of the student's experience with dissolution and clinical work surrounding identity stability may facilitate positive outcomes. Use of self as context and diffusion techniques, as articulated in ACT, can assist counselors in addressing a student's use of maladaptive coping strategies.

Avoidance was also highly associated with distress in our study. Psychoeducation regarding the function of avoidance is recommended to combat problematic avoidance behaviors. Additionally, the significant association between higher avoidance and less initiation of the breakup could indicate that noninitiation is a risk factor for avoidance behaviors. Students using avoidance techniques could benefit from supportive exposure to undesired thoughts or feelings regarding the dissolution. Confrontation with dissolution‐related content using ACT interventions may reduce overall distress and foster acceptance (Levin, Hildebrandt, et al., [31] ; Öst, [46] ).

This study's findings further evidence that, although relationship dissolutions are normative in the lives of college students, significant distress can result from certain coping strategies and can negatively affect mental health. The present study specifically found that RNT and avoidance significantly contributed to global distress among college students, even after controlling for demographics and characteristics of the former relationship. Additionally, gender moderated the relationship between RNT and distress, with women demonstrating the stronger relationship. Clinically, these findings may assist student counseling centers in targeting specific symptoms that prolong distress, which might allow for a more adaptive response to the relationship's end.

Graph: Relationship Between Repetitive Negative Thinking (RNT) and Gender by Distress Note. Gender moderates the relationship between distress, as measured by the Hopkins Symptom Checklist (HSCL), and RNT. Female R2 linear =.45; male R2 linear =.32.