Infidelity’s aftermath: Appraisals,mental health,and health-compromising behaviors following a partner’s infidelity

Abstract

Guided by transactional stress theory, this research investigated the role of appraisals in noninvolved partners’ mental health and health-compromising behaviors after infidelity. Responses from 232 college students who were recently cheated on revealed that negative appraisals (partner blame, self-blame, and causal attribution) had indirect effects on health-compromising behaviors through mental health (depression, anxiety, and distress). Moderated mediation analyses revealed that gender altered the indirect effect of partner and self-blame on health-compromising behaviors through mental health. Men’s health-compromising behaviors did not differ based on their appraisals or mental health. However, women who reported negative appraisals and high levels of mental health consequences engaged in more health-compromising behaviors. These findings suggest that perceptions of a partner’s infidelity are important, and that those perceptions affect noninvolved partners’ mental health and physical health behaviors.

Keywords

Attribution blame health infidelity mental health romantic relationships stress

Infidelity occurs in many romantic relationships (Mark, Janssen, & Milhausen, 2011). In fact, estimates suggest that infidelity happens in 20–25% of marriages (Laumann, Gagnon, Michael, & Michaels, 1994; Wiederman, 1997) with 22–34% of men and 11–24% of women committing a sexual infidelity in a marital relationship (Allen et al., 2005; Tafoya & Spitzberg, 2007). Additionally, up to 75% of college students have reported engaging in some form of infidelity while in a dating relationship (Shackelford, LeBlanc, & Drass, 2000). Infidelity, which is an emotional, sexual, or romantic involvement that violates an existing relationship’s commitment norms (Glass, 2002), can have damaging effects on the relationship and both partners. In particular, noninvolved partners (i.e., the partners who were cheated on in the relationship) often experience a range of emotional and psychological distress following infidelity, including depression, anxiety, decreased personal and sexual confidence, and decreased self-esteem (Bird, Butler, & Fife, 2007; Gordon & Baucom, 1999; Gordon, Baucom, & Snyder, 2004). The mental health consequences following a partner’s infidelity also may result in an increase in health-compromising behaviors, such as unprotected sex or greater alcohol and drug use, which may have lasting effects on the noninvolved partner’s health. Therefore, experiencing an infidelity can be stressful for the noninvolved partner, with potential harmful consequences (Fife, Weeks, & Stellberg-Filbert, 2013).

The purpose of this study is two fold. First, guided by transactional stress theory (Lazarus & Folkman, 1984, 1987), the present study investigates the mental health consequences and health-compromising behaviors of noninvolved partners following infidelity by examining their negative appraisals of the infidelity (i.e., causal attributions, partner blame, and self-blame). Second, we examine whether noninvolved partners’ mental health links their appraisals to health-compromising behaviors after their partners’ infidelity. Given the prevalence of, and stress caused by, infidelity in romantic relationships, understanding the emotional and physical impacts of infidelity on the noninvolved partner is imperative. This study extends the research in several ways. First, much research on infidelity tends to be atheoretical. This study applies a stress framework to explain the links between appraisal of the infidelity, mental health, and engagement in health-compromising behaviors. Second, infidelity is among the most distressing and damaging events couples face, and it is one of the most difficult problems to treat in therapy (Gordon et al., 2004). This study extends the research on the emotional and health toll of infidelity on the noninvolved partner. Third, this study extends previous research on infidelity by identifying the role of specific personal and relationship factors that might impact the connections between individuals’ appraisals of the infidelity, mental health, and health-compromising behaviors.

Stress, infidelity, and attributions

Lazarus and Folkman’s (1984, 1987) transactional stress theory centers on the importance of cognitive appraisals about an event or situation based on personal and environmental factors, which can result in negative or positive consequences. When an event or situation occurs, the initial appraisal centers on the question, “Is this event or situation stressful?” If a person believes that the situation is not at all stressful, he or she makes a nonthreatening appraisal of the situation, and no reaction is needed. If the person perceives the situation to be stressful, however, the situation can result in negative or positive outcomes for the person. In other words, the individual might see the situation as harmful, threatening, or challenging, and the appraisal helps determine the emotional and physical consequences of the stressful event. However, the impact of the stressor not only hinges on one’s cognitive appraisal, but characteristics of the person and the environment, as well (Lazarus & Folkman, 1984, 1987). Stated another way, features of the person or environment can moderate the severity of the stressful event.

Applying the transactional stress theory to infidelity suggests that the noninvolved partners’ appraisals of the infidelity, as well as features about the individual and the relationship, may impact their responses, health, and well-being. Specific appraisals that have been shown to play a major role in how partners respond and react to stressful events such as infidelity are attributions (Gordon & Baucom, 2003). The focus of attribution theory is on how people evaluate causes of and accountability for behavior, including their own and others’ behaviors (Weiner, 1979). In relationships, causality of a behavior is determined by the locus (whether the cause rests within the partner), stability (whether the behavior will continue over time), and globality (whether the behavior affects other areas of the relationship; Fincham & Bradbury, 1992). In terms of infidelity, if noninvolved partners perceive that the locus is within the cheating partner, that the behavior is not likely to change, and that the behavior impacts other areas of their relationship, they would ascribe causality to the partner, referred to as a causal attribution. Partners who make causal attributions may experience strong emotional reactions (Buunk, 1984), respond negatively, and end the relationship (Hall & Fincham, 2006).

Another appraisal that might influence a noninvolved partner’s health outcomes after infidelity is the attribution of blame. To make sense of an event, people evaluate whether others’ behaviors were intentional and whether they should be blamed (Weiner, 1986). Partner blame attributions concern accountability for a behavior, including the intent (whether the partner’s behavior was on purpose), motivation (whether the behavior was motivated by selfish reasons), and degree to which the partner is blameworthy for the behavior (Fincham & Bradbury, 1992). After an infidelity, noninvolved partners who perceive that the infidelity was intentional, motivated by selfish reasons, and the fault of the partner would ascribe responsibility and blame to the cheating partner. Similar to other relationship stressors (e.g., conflict and dissolution; Gordon, Friedman, Miller, & Gaertner, 2005), noninvolved partners who attribute responsibility and blame to the cheating partner may experience distress and depression.

Likewise, attributions of self-blame may impact individuals’ reaction to a partner’s infidelity. After a relationship breakup, self-blame is associated with greater depression and anxiety (Boelen & Reijntjes, 2009). For situations involving infidelity, the degree to which noninvolved partners blame their unfaithful partners or blame themselves for the infidelity affects whether the relationship can be reconciled (Bird et al., 2007). After a partner’s infidelity, noninvolved partners who experience high self-blame also report greater psychological distress (Shackelford et al., 2000). Therefore, identifying the noninvolved partner’s appraisals of the infidelity, including causal attributions, partner blame, and self-blame, is an important step in understanding the consequences to their mental and physical well-being.

Mental health and health-comprising behaviors following infidelity

Transactional stress theory also emphasizes how appraisals of stressful events impact personal outcomes, such as mental and physical health consequences (Lazarus & Folkman, 1984, 1987). In particular, infidelity, a major relationship stressor, can have negative mental health consequences for the noninvolved partner. After infidelity, noninvolved partners often experience negative emotional reactions including depression, anxiety, and symptoms consistent with those of posttraumatic stress disorder (Cano & O’Leary, 2000; Gordon & Baucom, 1999). Noninvolved partners also report feelings of shame, victimization, powerlessness, and rage following a partner’s infidelity (Bird et al., 2007; Gordon et al., 2004). In addition, wives are 6 times more likely to be diagnosed with a major depressive episode after discovering their husbands’ infidelity (Cano & O’Leary, 2000).

Even though the empirical literature on the association between infidelity and physical health is limited, research linking other stressors to mental and physical health outcomes, including health-compromising behaviors, is well developed (e.g., Hatch & Dohrenwend, 2007). In general, stressful life events often increase the chance of becoming ill (Turner, 2010) and are associated with poor mental and physical health (Hatch & Dohrenwend, 2007). To cope with stressors, individuals often engage in health-compromising behaviors, which can be seen as negative behaviors that increase the risk of poor health outcomes. Health-compromising behaviors include but are not limited to alcohol, nicotine, or other drug use, overeating, limited physical activity, and unprotected sex (Walsh, Senn, & Carey, 2013). In a longitudinal study, Walsh, Senn, and Carey (2013) found that stress, depression, and anxiety predicted increases in health-compromising behaviors; however, partaking in health-compromising behaviors did not impact mental health over time.

In terms of romantic relationships, stressors such as relationship discord and conflict have negative effects on objective and self-rated health outcomes, including cardiovascular disease (Baker et al., 2000) and functional impairments (Choi & Marks, 2008). Relationship breakups, another relationship stressor, can result in poorer physical health (Lepore & Greenberg, 2002) and, among college students, health-compromising behaviors such as increased alcohol use, poorer academic performance, and more disorganized behavior (Field, Diego, Pelaez, Deeds, & Delgado, 2013). In addition, people who experience divorce-related emotional intrusions (e.g., thinking about the separation) have increased resting systolic and diastolic blood pressure (Sbarra, Law, Lee, & Mason, 2009). Based on this body of research linking relationship stress and health, we also then might expect that infidelity would be associated with health, including engagement in health-compromising behaviors.

Moreover, these studies not only demonstrate the association between relationship stressors and physical health, they also illustrate the links between mental health, physical health, and health-compromising behaviors. Research has shown that negative emotional states are associated with unhealthy physical functioning and illness susceptibility (Cohen et al., 1995) and that behavioral practices greatly influence one’s physical health (Kobau et al., 2013). Likewise, people with mental health issues and those who engage in health-compromising behaviors are at a higher risk of communicable and noncommunicable diseases (Des Jarlais, Semaan, & Arasteh, 2011; Ezzati & Riboli, 2013; Prince et al., 2007). Additionally, negative emotional experiences, such as depression, are related to health-compromising behaviors, including smoking and having unprotected sex, possibly as a way of reducing negative emotions and inducing positive feelings (e.g., Glassman et al., 1990; Kelly & Kalichman, 1998). Overall, these findings suggest that mental health, physical health, and health-compromising behaviors are interrelated. Although people in romantic relationships tend to have better health, a major internal stressor such as infidelity may hinder the typical salubrious effects and instead have mental and physical health consequences, including increased engagement in health-compromising behaviors.

Again, these mental and physical health consequences following infidelity might be explained with transactional stress theory. When examining infidelity as a stressor, the model posits that the appraisal of the infidelity would influence a partner’s emotional response, subjective well-being, and health-compromising behaviors (Lazarus &Folkman, 1987). Appraisals that involve causal attributions and self and partner blame would lead to more mental health issues because noninvolved partners could lose (or have already lost) their relationships and friendships with their unfaithful partners as well as their investment and personal beliefs about ideal relationships. In turn, the poorer mental health would lead to increases in health-compromising behaviors. Therefore, we would expect that noninvolved partners whose appraisals involve causal attributions about the infidelity and self and partner blame would report more depression, anxiety, and distress, which would lead to increases in health-compromising behaviors, such as increased unprotected sex, smoking, and drug and alcohol use.

Person and environment as moderators of infidelity

Another key component of transactional stress theory is that the impact of the stressor also depends on features of the person and the environment (Lazarus & Folkman, 1987). In terms of infidelity, a particular feature of the person that might influence noninvolved partners’ reactions and health following infidelity is their gender. After a partner’s infidelity, women are more likely than men to feel depressed, disappointed, anxious, helpless/abandoned, and undesirable/insecure (Shackelford et al., 2000; Sweeney & Horwitz, 2001). Men are less likely to react emotionally, potentially due to gendered expectations that men should not express their emotions (Baum, 2003). Moreover, women may have more emotional responses to infidelity because their self-construal and identity are often formed by social relationships, including their romantic relationships (Cross & Madson, 1997). Therefore, we might expect that the mental health and health-compromising behaviors of noninvolved partners following infidelity would differ by gender.

In addition, whether or not the couple stayed together after the infidelity may be considered a feature of the relationship that might impact noninvolved partners’ health. Dating couples are more likely to terminate their relationships after infidelity compared to married couples (Sheppard, Nelson, & Mathie, 1995), suggesting that more committed relationships often persist after infidelity. Further, relationship dissolution is already a highly stressful situation (Field, Diego, Pelaez, Deeds, & Delgado, 2009) and dissolution after infidelity creates additional distress, leading to poorer mental health outcomes (Kitson, 1992). We would therefore suspect that whether or not noninvolved partners continued the relationship with their unfaithful partners might moderate the links between noninvolved partners’ appraisals, mental health, and health-compromising behaviors after infidelity.

Overview of the study

The purpose of the present study was to examine the relationships between appraisals, mental health, and health-compromising behaviors of noninvolved partners following infidelity. Guided by transactional stress theory, we contended that infidelity would be considered a relationship stressor in which the noninvolved partners’ appraisals would impact their responses and outcomes following the infidelity. Figure 1 displays the conceptual models tested in this study. Causal attributions, partner blame, and self-blame may pose a threat to the relationship and to one’s well-being because noninvolved partners perceive that the infidelity was due to something about their partners or themselves, rather than something about the situation, leading to increased mental health consequences and health-compromising behaviors. Thus, we expected that high causal attributions, partner blame, and self-blame would be related to increases in mental health consequences as well as increases in health-compromising behaviors. Furthermore, research has shown that mental health and physical health are connected (Chapman, Perry, & Strine, 2005) and that psychological stress associated with relationship stressors can result in poorer physical health (Field et al., 2013; Sbarra et al., 2009). We therefore predicted that increases in mental health consequences would be related to more health-compromising behaviors.

Figure 1.
Models analyzed in this study: (a) basic mediation model and (b) moderated mediation model. Negative appraisals include partner blame, self-blame, and causal attribution; mental health includes depression, anxiety, and distress; and person and relationship characteristics include gender and relationship status with the cheating partner.

In addition to these direct effects, we expected that mental health consequences would mediate the relationships between appraisals and health-compromising behaviors. Specifically, we expected that causal attribution, partner blame, and self-blame would be related to increases in mental health consequences, which would subsequently be related to greater health-compromising behaviors (Figure 1a). We also predicted that gender and relationship status with the cheating partner would moderate the associations between appraisals, mental health, and health-compromising behaviors (Figure 1b).

Method

Participants and procedure

Participants were recruited from a subject pool at a medium-sized western university in the United States to complete an online study for partial fulfillment of research participation credits. The study description and first survey question specified that participants “must have been cheated on (emotionally, sexually/physically, or both) while in a committed relationship in the past 3 months.” Eighty-one individuals had not been cheated on in the past 3 months and were excluded from the study. A total of 232 individuals had been cheated on in the past 3 months and finished the remainder of the study. Of these participants, 15% were still in a relationship with the partner who cheated on them, 11% were not in a relationship with the cheating partner and were dating multiple people, 23% were not in a relationship with the cheating partner and were in a relationship with a different partner, 49% were not in a relationship with the cheating partner and currently were not in a relationship, and 2% did not report their current relationship status. Nearly all of the participants (98%) reported being in a dating relationship with the partner at the time the partner cheated with a mean length of 1.76 years (SD = 1.91 years), ranging from 1 month to 17 years. Participant ages ranged from 18 to 47 years (M = 20.94, SD = 3.97). The majority of participants was female (58%) and Caucasian (63%) followed by Latino (11%), African American (8%), Asian (7%), Pacific Islander (4%), Multi-Ethic (4%), other (2%), and Native American (1%). Participants completed the online questionnaire after verifying that they had been cheated on during the past 3 months. We chose the past 3 months because we expected the infidelity would be fresher in the noninvolved partners’ minds and they may have still been experiencing mental health problems and engaging in health-compromising behaviors.

Measures

The questionnaire included measures about respondents’ perceptions of the infidelity as well as perceptions of change in their feelings and physical health behaviors after they found out that their partners had cheated on them. Descriptive statistics and correlations for the study variables are presented in Table 1.

Table 1.
Intercorrelations, means, standard deviations, and ranges for study variables.

1 2 3 4 5 6 7 8 9 10 11

1. Partner blame —

2. Self-blame −.06 —

3. Causal attribution .40* .18 —

4. Depression .28* .40* .15* —

5. Anxiety .24* .36* .21* .81* —

6. Distress .28* .49* .31* .68* .67* —

7. Mental health composite .32* .50* .25* .93* .87* .89*** —

8. Health-compromising behaviors .14 .17* .14 .32* .29* .27* .31* —

9. Relationship length in years −.04 .09 −.10 .14* .17* .12 .14* −.12 —

10. Sexual/physical infidelity .05 .03 .09 .06 .07 .10 .07 .13 −.00 —

11. Emotional infidelity .25* .02 .18 .16* .08 .08 .13 .04 .05 –.23** —

Mean 4.60 3.00 4.10 4.62 4.92 4.85 4.75 3.17 1.76 5.41 4.42

Standard deviation 1.14 1.60 1.08 1.11 1.32 1.31 1.10 2.77 1.91 1.75 1.93

Range 1–6 1–6 1–6 1-–7 1–7 1–7 1–7 0–13 1–17 1–7 1–7

Note. The range for relationship length is 1 month to 17 years.

*p < .05; **p < .01; p < .001.

Causal attribution of the infidelity

To measure causal attribution of the infidelity, the 3-item subscale of Fincham and Bradbury’s (1992) relationship attribution measure (RAM) was used. Similar to Hall and Fincham’s (2006) adapted version of the RAM, the items were solely regarding infidelity. Participants were instructed to think of how they felt after they found out that their partners had cheated on them (e.g., “I thought the reason my partner cheated on me was not likely to change”). Participants rated each item on a scale from 1 (disagree strongly) to 6 (agree strongly) with higher scores indicating the partner’s infidelity was caused by the partner, a stable behavior, and affected other areas of the relationships. The RAM has shown to be a reliable and valid measure of attributions in relationships (Fincham & Bradbury, 1992), and the causal attribution dimension of the RAM and has shown reasonable internal consistency (α’s = .63–.93; Fincham & Bradbury, 1992). Cronbach’s α was .63 in the current study.

Partner blame

The 3-item responsibility-blame subscale of Fincham and Bradbury’s (1992) RAM was used and also adapted to regard a partner’s infidelity (e.g., “I thought my partner deserved to be blamed for cheating on me”). Participants rated the three items on a scale from 1 (disagree strongly) to 6 (agree strongly) with higher scores indicating higher partner blame. The responsibility-blame dimension of the RAM has shown adequate internal consistency (α’s = .70–.93; Fincham & Bradbury, 1992) and was acceptable in the current study (α = .71).

Self-blame

The 9-item self-blame subscale of the Grief Cognitions Questionnaire (Boelen, van den Bout, & van den Hout, 2003) was used to assess the degree to which participants blame themselves for their partners’ infidelity. The original measure is used to assess self-blame after a relationship breakup; therefore, we adapted the items to assess self-blame after infidelity (e.g., “I blamed myself for not having cared better for my partner before he/she cheated on me”). The adapted measure was highly reliable (α = .93). Participants rated each item on a scale from 1 (disagree strongly) to 6 (agree strongly) with higher scores reflecting greater self-blame.

Mental health

To examine perceived change in depression and anxiety symptoms, we asked participants to rate how much they experienced several mental health symptoms after their partners’ infidelity compared to their typical experiences. For example, some participants may have been experiencing mental health issues even before the infidelity and asking them to simply rate their mental health status following the infidelity would not tell us if their mental health changed as a consequence of the infidelity. In this way, assessing change may capture subtle changes for people who regularly experience mental health symptoms on a daily basis, even before the infidelity. Likewise, although some people may experience greater mental health symptoms after infidelity, others may experience growth and therefore fewer symptoms than typically experienced before being cheated on. Therefore, we assessed change in depression and anxiety symptoms to assess perceived changes, whether increases or decreases, in their symptoms after the infidelity.

Change in depression symptoms

The 20-item Center for Epidemiologic Studies Depression Scale–Revised (Eaton, Smith, Ybarra, Mutaner, & Tien, 2004) was used to measure the perceived change in participants’ depressive symptoms after their partners’ infidelity compared to their typical depressive symptoms (e.g., “My appetite was poor”). We adapted the response options to measure the degree to which their depression symptoms increased or decreased after the infidelity compared to their typical experiences. The standard responses options (0 = not at all or less than 1 day during the week* to 4 = nearly every day for 2 weeks) were modified to 1 (this occurred much less than usual) to 7 (this occurred much more than usual) with 4 as the midpoint (no difference/about the same) to assess change in their depression symptoms after the infidelity. Cronbach’s α was .93.

Change in anxiety symptoms

The 7-item Generalized Anxiety Disorder Scale (Spitzer, Kroenke, Williams, & Lowe, 2006) was used to assess the perceived change in participants’ anxiety after their partners’ infidelity compared to their typical anxiety (e.g., “I was not able to stop or control worrying”). Again, we adapted the response options to measure the degree to which their anxiety symptoms increased or decreased after the infidelity compared to their typical experiences (α = .92). The response options in the original scale ranged from 0 (not at all) to 4 (nearly every day) but were changed to 1 (this occurred much less than usual) to 7 (this occurred much more than usual) with 4 as the midpoint (no difference/about the same) to assess change in anxiety symptoms after the infidelity.

Distress

The 16-item Break-Up Distress Scale (Field et al., 2009) was used to assess distress after their partners’ infidelity. The original scale was designed to assess distress after a relationship breakup (e.g., “I feel that life is empty without the person”) but was adapted in this study to assess distress after a partner’s infidelity (e.g., “I felt that life was empty after my partner cheated on me”). Participants rated each item on a scale from 1 (disagree strongly) to 7 (agree strongly), and the measure exhibited an α of .93.

To assess overall mental health consequences following a partner’s infidelity, we created a composite variable comprising depression, anxiety, and distress symptoms after the infidelity as suggested by the high correlations among the three variables (rs = .68–.81). We conducted a confirmatory factor analysis (CFA) to verify the viability of a composite mental health variable. The CFA model was fully saturated (i.e., df = 0), therefore, fit indices were uninformative. However, depression (β = .91), anxiety (β = .89), and distress (β = .75) all had strong loadings on the composite variable. The internal consistency of the composite mental health measure was high (α = .96). Higher scores indicate greater perceived mental health consequences following a partner’s infidelity.

Changes in health-compromising behaviors

A 17-item measure was created by the researchers based on items from the Center for Disease Control and Prevention’s Behavioral Risk Factor Surveillance System (2014) to assess the perceived change in health-compromising behaviors after their partners’ infidelity compared to their typical behaviors. Items included drug and alcohol use (e.g., “Alcohol use”), sexual (e.g., “Not using condoms during sex”), exercise (e.g., “Over-exercising”), and eating behaviors (e.g., “Eating less or not at all”). The response options measured the change in the behaviors, ranging from 1 (significantly decreased) to 7 (significantly increased) with 4 as the midpoint (no change). A “not applicable” option was provided for each item as well (N/A—never engaged in this behavior). For the subsequent analyses, we created a count variable that assesses the number of health-compromising behaviors in which participants engaged following their partners’ infidelity. If participants selected an increase in a behavior (i.e., selected 5–7), we coded that item as 1, whereas if not applicable, no change, or a decrease in a behavior was selected (i.e., N/A, 1–4), the item was coded as 0. We then added each of the items together for an overall index of the number of health-compromising behaviors participants reported increases in after their partners’ infidelity. Scores could range from 0 to 17, with higher scores indicating greater engagement in health-compromising behaviors after a partner’s infidelity. We assessed reliability with the Kuder–Richardson 20, which resulted in an internal consistency value of .75.¹

Moderating and control variables

Gender and whether or not participants were in a relationship with the cheating partner were included in the analyses as potential moderators. We also controlled for the degree to which the infidelity was sexual/physical, the degree to which the infidelity was emotional, and the relationship length with the cheating partner because of their potential impact on noninvolved partners’ reactions (Buss et al., 1999; Pittman, 1989). Sexual/physical and emotional infidelities were assessed with two questions: (1) to what degree was your partner’s infidelity sexual/physical? (1 [not at all sexual] to 7 [completely sexual]) and (2) to what degree was your partner’s infidelity emotional? (1 [not at all emotional] to 7 [completely emotional]).

Results

Preliminary analyses

Means in Table 1 show that noninvolved partners were more likely to blame the partner than themselves and tended to hold causal attributions about their partner’s infidelity. They also reported increases in depression, anxiety, and distress following the infidelity. On average, participants reported an increase in more than 3 health-compromising behaviors, with increases ranging from 0 to 13 behaviors. The most frequent increases in health-compromising behaviors were eating less or not at all (45%), alcohol use (44%), over-exercising (29%), having sex under the influence of drugs or alcohol (27%), and marijuana use (19%).

In addition, correlations among the main variables were in line with our predictions (see Table 1). Each of the three appraisal variables (partner blame, self-blame, and causal attribution) were positively correlated with the composite mental health variable and the three mental health variables (depression, anxiety, and distress) such that increases in negative appraisals were associated with more mental health symptoms following the infidelity. Also, the greater the self-blame, the more likely participants were to be involved in health-compromising behaviors. Partner blame and causal attribution were not correlated with health-compromising behaviors. Mental health was also positively associated with health-compromising behaviors, suggesting that when noninvolved partners had greater depression, anxiety, and distress following the infidelity, they also engaged in more health-compromising behaviors. Noninvolved partners who had longer relationships with the cheating partner also reported experiencing more mental health consequences after the infidelity. In addition, more emotional infidelities were associated with greater partner blame, causal attributions, and depressive symptoms. Independent samples t-tests showed that there were no significant differences on study variables by gender or relationship status with the cheating partner.

Mediational analyses

To test the models specified in Figure 1, we first tested whether mental health mediated the effects of negative appraisals of the infidelity (partner blame, self-blame, and causal attributions) on health-compromising behaviors. We then followed up these analyses with moderated mediational analyses to investigate whether any indirect effects of appraisals on health-compromising behaviors through mental health were moderated by gender and relationship status with the cheating partner. We conducted these analyses using Hayes’ (2013) SPSS PROCESS macro and selected model 4 for the mediation analyses and model 59 for the moderated mediation analyses. Conditional and indirect effects were tested with 95% bias-corrected confidence intervals, and the confidence intervals for the indirect effects were also tested with 10,000 bootstrapped samples. Moderated mediation was assessed with Hayes’ (2015) index of moderated mediation, which was also tested with 10,000 bootstrapped samples and 95% confidence intervals. Moderated mediation is confirmed if the index of moderated mediation does not include zero (Hayes, 2015). Continuous variables were mean centered prior to the moderation analyses to improve interpretability (Aiken & West, 1991), and interaction terms were computed as the product of the mean centered variables. Moderated mediation effects were plotted to depict the conditional effects of the moderator on the indirect effects at one standard deviation above and below the mean. We also controlled for any influences of the type of infidelity (sexual/physical and emotional) and relationship length with the cheating partner by entering these three variables as covariates.

Partner blame, mental health, and health-compromising behaviors

We first tested whether mental health consequences mediated the relationship between partner blame and health-compromising behaviors of noninvolved partners following infidelity, as depicted in Figure 1a, while controlling for the type of infidelity and relationship status with the cheating partner. Table 2 shows that increases in partner blame were related to increases in mental health consequences. Although partner blame was not correlated with health-compromising behaviors, mental health was examined as a potential mediator given that the predictor and dependent variables do not need to be directly related to test the presence of an indirect effect (Hayes, 2009). When health-compromising behavior was regressed onto partner blame and mental health, only mental health predicted health-compromising behaviors. Results from the bootstrapped analyses showed that mental health mediated the relationship between partner blame and health-compromising behaviors (indirect effect = .23, SE = .10, 95% CI [.07, .47]). Hence, partner blame was related to greater mental health consequences, which was related to more health-compromising behaviors of noninvolved partners after infidelity.

Table 2.
Model coefficients for the mediational analyses predicting health-compromising behaviors.

Mental health (M) Health-compromising behaviors (Y)

Predictors b SE 95% CI b SE 95% CI

Partner blame

Partner blame (X₁) 0.31* 0.07 0.17, 0.45 0.03 0.19 −0.34, 0.41

Mental health (M) — — — 0.77* 0.20 0.39, 1.16

Emotional infidelity (U₁) 0.01 0.04 −0.07, 0.10 −0.01 0.11 −0.20, 0.23

Physical infidelity (U₂) 0.06 0.05 −0.03, 0.15 0.12 0.12 −0.11, 0.35

Relationship length (U₃) 0.01* 0.00 0.00, 0.01 −0.01 0.01 −0.02, 0.10

R² = 0.13 R² = 0.11

F(4, 173) = 6.77, p < .001 F(5, 172) = 4.15, p = .001

Self-blame

Self-blame (X₂) 0.31* 0.05 0.22, 0.40 0.08 0.14 −0.20, 0.36

Mental health (M) — — — 0.73* 0.21 0.32, 1.14

Emotional infidelity (U₁) 0.04 0.04 −0.03, 0.12 0.02 0.11 −0.20, 0.36

Physical infidelity (U₂) 0.05 0.04 −0.03, 0.14 0.12 0.12 −0.11, 0.35

Relationship length (U₃) 0.00 0.00 −0.00, 0.01 −0.01 0.01 −0.02, 0.01

R² = 0.24 R² = 0.11

F(4, 173) = 13.90, p < .001 F(5, 1712) = 4.21, p = .001

Causal attribution

Causal attribution (X₃) 0.27* 0.08 0.12, 0.42 0.15 0.20 −0.24, 0.55

Mental health (M) — — — 0.82 0.19 0.41, 1.18

Emotional infidelity (U₁) −0.02 0.04 −0.07, 0.11 0.02 0.11 −0.20, 0.23

Physical infidelity (U₂) 0.04 0.05 −0.05, 0.13 0.13 0.12 −0.10, 0.37

Relationship length (U₃) 0.01* 0.00 0.00, 0.14 −0.01 0.01 −0.02, 0.01

R² = 0.10 R² = 0.12

F(4, 1170) = 4.89, p = .001 F(5, 169) = 4.78, p < .001

Note. All coefficients are unstandardized and based on uncentered data. Relationship length is in months. M = mediator; X = independent variable; Y = dependent variable; U = covariate.

p < .05; p < .01; p < .001.

We then tested the potential moderating effects of gender and relationship status with the cheating partner on the links between partner blame, mental health, and health-compromising behaviors, as depicted in Figure 1b. Gender did not alter the effects of partner blame on mental health nor health-compromising behaviors (lines 1 and 2, respectively, Figure 1b), however, gender did moderate the effects of mental health on health-compromising behaviors (line 3, Figure 1b; see Table 3). Additionally, the overall indirect effect of partner blame on health-compromising behaviors through mental health differed by gender (index of moderated mediation = −.37, SE = .18, 95% CI [−.77, −.05]). As shown in Figure 2, the indirect effect of partner blame on health-compromising behaviors through mental health is stronger for women compared to men. For women, high levels of partner blame led to steeper increases in mental health consequences, resulting in more health-compromising behaviors (conditional indirect effect = .40, SE = .13, 95% CI [.18, .72]), whereas men’s partner blame did not have an indirect effect on engagement in health-compromising behaviors through mental health (conditional indirect effect = .04, SE = .12, 95% CI [−.16, .32]). In contrast, the relationships between partner blame, mental health, and health-compromising behaviors were not dependent upon the relationship status with the cheating partner (index of moderation mediation = −.01, SE = .22, 95% CI [−.56, .36]).

Table 3.
Model coefficients for partner blame, mental health, and health-compromising behaviors moderated by gender.

Mental health (M) Health-compromising behaviors (Y)

b SE 95% CI b SE 95% CI

Partner blame (X₁) 0.30* 0.07 0.15, 0.44 0.05 0.19 −0.33, 0.43

Mental health (M) — — — 0.73* 0.19 0.35, 1.11

Gender (W) −0.12 0.16 −0.44, 0.20 0.59 0.40 −0.21, 1.39

Partner blame × Gender (X₁ × W) −0.09 0.15 −0.38, 0.20 0.58 0.39 −0.19, 1.34

Mental health × Gender (M × W) — — — −1.05** 0.40 −1.82, −0.27

Emotional infidelity (U₁) 0.02 0.04 −0.07, 0.11 −0.02 0.11 −0.23, 0.20

Physical infidelity (U₂) 0.06 0.05 −0.03, 0.15 0.10 0.11 −0.13, 0.33

Relationship length (U₃) 0.01 0.00 −0.00, 0.01 −0.01 0.01 −0.03, 0.01

R² = 0.14 F(6, 171) = 4.65, p < .001 R² = 0.16 F(8, 169) = 3.96, p < .001

Note. All coefficients are unstandardized and based on uncentered data. Relationship length is in months. M = mediator; X = independent variable; Y = dependent variable; U = covariate.

p < .05; p < .01; p < .001.

Figure 2.
A visual representation of the indirect effects of partner blame on health-compromising behaviors through mental health by gender.

Self-blame, mental health, and health-compromising behaviors

Next, we tested whether mental health mediated the relationship between self-blame and health-compromising behaviors of noninvolved partners following infidelity while controlling for the type of infidelity and relationship status with the cheating partner. Results showed that increases in self-blame were related to increases in mental health consequences (see Table 2), and although self-blame was correlated with health-compromising behaviors, when health-compromising behavior was regressed onto self-blame and mental health, only mental health predicted health-compromising behaviors. Results from the bootstrapped analyses revealed that mental health mediated the link between self-blame and health-compromising behaviors (indirect effect = .22, SE = .09, 95% CI [.08, .41]). Thus, self-blame was related to greater mental health consequences, which subsequently was related to more health-compromising behaviors of noninvolved partners.

We also tested the moderating effects of gender and relationship status with the cheating partner on the associations between self-blame, mental health, and health-compromising behaviors of noninvolved partners following infidelity (Figure 1b). As with the model for partner blame, gender did not moderate the effect of self-blame on mental health nor health-compromising behaviors (see Table 4). However, gender altered the effects of mental health on health-compromising behaviors (line 3, Figure 1b) and altered the indirect effect of self-blame on health-compromising behaviors through mental health (index of moderated mediation = −.38, SE = .19, 95% CI [−.77, −.08]). As depicted in Figure 3, these associations were particularly strong for women, such that increases in self-blame led to steeper increases in mental health symptoms, which subsequently led to more health-compromising behaviors (conditional indirect effect = .41, SE = .12, 95% CI [.22, .68]). For men, the indirect effect of self-blame on health-compromising behaviors through mental health was not significant (conditional indirect effect = .02, SE = .12, 95% CI [−.26, .24]). Additionally, relationship status with the cheating partner did not moderate the associations between self-blame, mental health, and health-compromising behaviors (index of moderated mediation = −.05, SE = .24, 95% CI [−.65, .30]).

Table 4.
Model coefficients for self-blame, mental health, and health-compromising behaviors moderated by gender.

Mental health (M) Health-compromising behaviors (Y)

b SE 95% CI b SE 95% CI

Self-blame (X₂) 0.31* 0.05 0.22, 0.40 0.07 0.14 −0.21, 0.34

Mental health (M) — — — 0.68 0.21 0.27, 1.09

Gender (W) −0.06 0.15 −0.35, 0.24 0.60 0.41 −0.20, 1.40

Self-blame × Gender (X₂ × W) −0.10 0.09 −0.28, 0.09 0.30 0.29 −0.27, 0.87

Mental health × Gender (M × W) — — — −1.06* 0.42 −1.90, −0.22

Emotional infidelity (U₁) 0.04 0.04 −0.04, 0.12 0.02 0.11 −0.19, 0.24

Physical infidelity (U₂) 0.05 0.04 −0.04, 0.13 0.12 0.12 −0.11, 0.34

Relationship length (U₃) 0.00 0.00 −0.00, 0.01 −0.01 0.01 −0.02, 0.01

R² = 0.25 F(6, 171) = 9.44, p < .001 R² = 0.15 F(8, 169) = 3.82, p < .001

Note. All coefficients are unstandardized and based on centered data. Gender: 0 = females; 1 = males. Relationship length is in months. M = mediator; X = independent variable; W = moderator; Y = dependent variable; U = covariate.

p < .05; p < .01; p < .001.

Figure 3.
A visual representation of the indirect effects of self-blame on health-compromising behaviors through mental health by gender.

Causal attribution, mental health, and health-compromising behaviors

Finally, we tested whether mental health consequences mediated the relationship between causal attributions about the infidelity and health-compromising behaviors of noninvolved partners following infidelity, as depicted in Figure 1a. Results indicated that increases in causal attributions were related to increases in mental health consequences (see Table 2). As with partner blame, causal attribution was not correlated with health-compromising behaviors; however, mental health was examined as a potential mediator. When health-compromising behaviors were regressed onto mental health and causal attribution, only mental health predicted health-compromising behaviors. Results from the bootstrapped analyses showed mental health mediated the link between causal attributions and health-compromising behaviors (indirect effect = .22, SE = .10, CI [.05, .45]). Thus, increases in causal attributions were related to more mental health consequences, which in turn were related to increases in health-compromising behaviors of noninvolved partners following infidelity.

Once again, we examined the potential moderating effects of gender and relationship status with the cheating partner on the associations between causal attribution, mental health, and health-compromising behaviors, as depicted in Figure 1b. Table 5 shows that gender did not moderate the effects of causal attribution on mental health nor health-compromising behaviors (lines 1 and 2, respectively, Figure 1b); however, gender altered the effects of mental health on health-compromising behaviors (line 3, Figure 1b). Although the link between mental health and health-compromising behaviors differed by gender, the indirect effect of causal attribution on health-compromising behaviors through mental health did not depend on gender (index of moderation mediation = −.31, SE = .18, 95% CI [−.68, .01]). That is, the mediating effect of mental health on the link between causal attribution and health-compromising did not differ for men and women. Instead, only the link between mental health and health-compromising behaviors differed by gender such that men’s health-compromising behaviors did not vary based upon their mental health, whereas women’s health-compromising behaviors increased with high levels of mental health consequences. Additionally, relationship status with the unfaithful partner did not moderate the indirect effect of causal attributions on health-compromising behaviors through mental health (index of moderated mediation = −.25, SE = .27, 95% CI [−.93, .18]).

Table 5.
Model coefficients for causal attribution, mental health, and health-compromising behaviors moderated by gender.

Mental health (M) Health-compromising behaviors (Y)

b SE 95% CI b SE 95% CI

Causal attribution (X₃) 0.27* 0.08 0.12, 0.42 0.12 0.20 −0.27, 0.51

Mental health (M) — — — 0.75* 0.19 0.37, 1.13

Gender (W) −0.09 0.16 −0.41, 0.23 0.63 0.41 −0.17, 1.44

Causal attribution ×Gender (X₃ × W) −0.13 0.15 −0.42, 0.17 0.16 0.39 −0.60, 0.93

Mental health × Gender (M × W) — — — −0.85* 0.39 −1.62, −0.07

Emotional infidelity (U₁) 0.02 0.04 −0.07, 0.11 0.00 0.11 −0.22, 0.22

Physical infidelity (U₂) 0.04 0.05 −0.05, 0.14 0.12 0.12 −0.11, 0.35

Relationship length (U₃) 0.01* 0.00 0.00, 0.01 −0.01 0.01 −0.02, 0.01

R² = 0.11 F(6, 168) = 4.86, p < .001 R² = 0.16 F(8, 166) = 3.96, p < .001

Note. All coefficients are unstandardized and based on centered data. Gender: 0 = females; 1 = males. Relationship length is in months. M = mediator; X = independent variable; W = moderator; Y = dependent variable; U = covariate.

p < .05; p < .01; p < .001.

Alternative models

Although the proposed models were based on transactional stress theory, it is possible that alternative models might also capture the relationships between these variables. For example, mental health symptoms may lead to negative appraisals of a partner’s infidelity, which in turn may result in health-compromising behaviors. We tested this alternative model for each of the three appraisal variables, which showed that the indirect effects of mental health on health-compromising behaviors through appraisals were not significant for partner blame (indirect effect = .11, SE = .07, 95% CI [−.11, .16]), self-blame (indirect effect = .05, SE = .10, 95% CI [−.14, .27]), nor causal attribution (indirect effect = .03, SE = .06, 95% CI [−.06, .17]).

It is also plausible that appraisals might be related to increases in health-compromising behaviors, which then give rise to mental health symptoms following a partner’s infidelity. We tested the mediation models for the three appraisal variables and found that partner blame did not have an indirect effect on mental health through health-compromising behavior (indirect effect = .03, SE = .02, 95% CI [−.01, .08]). Although there was a small indirect effect of causal attribution on mental health through health-compromising behaviors (indirect effect = .04, SE = .03, 95% CI [.00, .11]), the path between causal attribution and health-compromising behaviors was not significant (coefficient = .37, SE = .20, 95% CI [−.03, .77]), indicating that the indirect effect was driven by the strong association between health-compromising behaviors and mental health and that mediation did not occur. Lastly, there was also a small indirect effect of self-blame on mental health through health-compromising behaviors (indirect effect = .03, SE = .02, 95% CI [.00, .07]). Although there was a minor mediation effect for the model with self-blame, taken together, the primary model tested in this study (mental health mediating the relationship between negative appraisals and health-compromising behaviors) was significant for all three appraisal variables.

Discussion

This study extended the literature on the health consequences of noninvolved partners after infidelity by applying a stress framework to understand the impact of appraisals on perceived mental health and health behavior outcomes. Guided by transactional stress theory (Lazarus & Folkman, 1984, 1987), we examined whether noninvolved partners’ negative appraisals of the infidelity were associated with poorer mental health and increased health-compromising behaviors. Specifically, we assessed whether noninvolved partners’ appraisals were indirectly related to health-compromising behaviors through increased mental health consequences. Further extending the research, we tested personal and relationship factors that impact the links between appraisals, mental health, and health-compromising behaviors.

As predicted, findings from the mediation models indicated that effects of negative appraisals on health-compromising behaviors were transmitted through mental health. That is, the greater the partner blame, self-blame, and causal attribution of the infidelity, the more perceived depression, anxiety, and distress symptoms, which in turn were associated with more health-compromising behaviors. These results are consistent with previous research showing that noninvolved partners experience negative emotional reactions following their partners’ infidelity (Bird et al., 2007; Cano & O’Leary, 2000; Gordon et al., 2004), and extend this line of research by identifying partner blame, self-blame, and causal attributions as specific appraisals that link a partner’s infidelity to their depression, anxiety, and distress. Further, this study revealed a connection between noninvolved partners’ mental health and health-compromising behaviors following the infidelity, which has been demonstrated in previous work examining physical health outcomes after other relationship and life stressors (Hatch & Dohrenwend, 2007; Lepore & Greenberg, 2002; Sbarra et al., 2009; Turner, 2010) but until this study has not been demonstrated in research on infidelity.

These findings fit within the transactional stress framework (Lazarus & Folkman, 1984, 1987). When noninvolved partners had negative appraisals of the infidelity by blaming their partners, blaming themselves, and attributing the infidelity to internal, global, and stable causes, those negative appraisals were then linked to poorer mental health. The noninvolved partners may have perceived that the infidelity occurred because of something about themselves, their partners, or their relationships; felt a sense of loss with the friendships with the cheating partners, their investment, and trust in their relationships; or experienced threat to their personal beliefs about themselves and relationships. Then, their greater depression, anxiety, and distress were associated with increased health-compromising behaviors, such as increased alcohol use, drug use, and unprotected sex. Engaging in these types of health-compromising behaviors might be explained as an attempt to cope with the infidelity by decreasing their negative feelings and increasing positive emotions (Glassman et al., 1990; Kelly & Kalichman, 1998).

With respect to potential person and environment moderators, we found that gender is a key component of noninvolved partners’ mental health and health behavior outcomes following infidelity. This study revealed that the strength of indirect effects of partner blame and self-blame on engagement in health-compromising behaviors through mental health symptoms differed by gender. The mediated effect of mental health was stronger for women compared to men. That is, women who reported high levels of self-blame or partner blame also experienced high levels of mental health consequences, which in turn resulted in even greater health-compromising behavior engagement following a partner’s infidelity. This gender difference is consistent with previous research showing that women report greater emotional distress following a partner’s infidelity compared to men (Shackelford et al., 2000; Sweeney & Horwitz, 2001). Additionally, women are more likely to base their self-construal and identity on their relationships compared to men (Cross & Madson, 1997). In this way, women who experience more mental health consequences may also engage in more health-compromising behaviors because their self-construal and identity might have been damaged by the infidelity.

Therefore, the results generally support the idea that appraisals, mental health, and health compromising behaviors are interconnected and important within noninvolved partners’ reactions to infidelity. Although participants in the sample were college students and may not be representative of married couples, most participants were invested in their relationships and committed to their partners. The average relationship length with the cheating partner was 1.76 years, almost all participants (94%) reported that they were emotionally attached or in love, and nearly a third had seriously thought about marriage. Despite the sample consisting of individuals in dating relationships who may not have had the structural constraints of married couples, most participants were in serious and committed relationships with strong emotional and time investments. As such, we would expect that they would still be hurt and experience negative consequences after their partner’s infidelity. Nevertheless, further investigation is needed to assess whether these results differ for older, married adults, potentially with different norms and expectations that are inherent in marriage.

Also, it is interesting that within 3 months following the occurrence of infidelity, 34% of the participants were either in a new relationship or dating multiple people. Contrary to our expectations, however, relationship status with the cheating partner did not modify the associations between appraisals, mental health, and health-compromising behaviors. These nonsignificant findings may be explained by the fact that infidelity is one of the most serious and damaging relationship transgressions, causing distress, anger, and decreased self-worth and self-esteem for the noninvolved partner (Bird et al., 2007; Gordon et al., 2004). These intense negative reactions may occur regardless of whether the noninvolved partner continues the relationship. Although we controlled for the effects of relationship length with the cheating partner, the relationships of noninvolved partners after infidelity in our college student sample may differ for those who are older and married. Further research is needed to examine how post-infidelity relationships compare between dating and married individuals.

This study has additional limitations that should be considered when interpreting the results. One limitation is the use of self-report and retrospective measures. Participants’ recollection of changes in their mental health and behavior engagement following the infidelity may be biased. However, we attempted to decrease personal bias by restricting participation to those who had been cheated on within the past 3 months because we expected the infidelity would be fresher in the noninvolved partners’ minds and they may have still been experiencing mental health problems and engaging in health-compromising behaviors. It is also possible that individuals with poor mental health and who engage in health-compromising behaviors would be in relationships in which infidelity is likely to occur. We attempted to address this concern by asking participants about the change in mental health symptoms and health-compromising behaviors after the infidelity compared to their typical symptoms and behaviors. Although capturing reactions to infidelity in real time could prove challenging, future research may want to track ongoing relationships and follow those relationships in which an infidelity subsequently occurs. By doing so, researchers could also explore the nature of noninvolved partners before infidelity, such as whether they have preexisting mental or physical health issues.

Another limitation is that the data were cross-sectional and causal inferences about the directions of these relationships cannot be made. Cross-sectional data, however, can establish relationships between constructs. Although we cannot establish causality, the findings are in line with a transactional stress framework as appraisals of the infidelity are associated both directly and indirectly with mental health consequences and health-compromising behaviors. Despite the alternative model with self-blame having a small indirect effect on mental health through health-compromising behaviors, our hypothesized models performed much better and were in line with predictions based on transactional stress theory as well as previous research on the associations between mental health and health-compromising behaviors (e.g., Walsh et al., 2013). However, future research should more closely examine the relationship between self-blame and health following an infidelity.

Additionally, future researchers may want to consider using objective measures of physical health to examine noninvolved partners’ reactions and behaviors after infidelity. For example, a physical measure that might help explain noninvolved partners’ responses to the stress of an infidelity is cortisol reactivity produced by the hypothalamic–pituitary–adrenal (HPA) axis, a major component of the physiological stress response. Activation of the HPA axis results in a hormonal response to stress (e.g., cortisol), which is related to physical and mental health problems (Sapolsky, 1999). Future research might consider examining whether negative appraisals of a partner’s infidelity are related to higher cortisol levels and therefore more physiological stress compared to those who make positive appraisals.

Despite these limitations, this study opens several avenues for future research. First, recent research has investigated the possibility of posttraumatic growth after experiencing relationship stressors, such as relationship dissolution (e.g., Tashiro & Frazier, 2003). Despite the stress and turmoil of infidelity, some noninvolved partners might experience growth as they cope. We attempted to examine personal growth following infidelity by assessing a decrease in health-compromising behaviors (e.g., drinking less, having more protected sex). However, the number of instances in which participants reported that a health-compromising behavior decreased was too few, restricting the variance and ability to create a meaningful variable. Nevertheless, it would be useful to understand how and why some people experience growth following the trauma of infidelity. Similarly, future research could examine positive appraisals of an infidelity, which may lead to more posttraumatic growth rather than mental health problems and health-compromising behaviors. In addition, the presence of specific protective factors may buffer the negative effects of infidelity as well as the negative appraisals. For example, noninvolved individuals with greater social support, higher self-esteem and self-confidence, greater life optimism, or a stronger sense of coherence might experience less mental health issues and fewer health-compromising behaviors following a partner’s infidelity. Lastly, noninvolved partners’ infidelity history including whether they have previously been cheated on or cheated themselves may impact their emotional and physical reactions to an infidelity. Future studies could examine whether such infidelity history factors influence noninvolved partners’ mental and physical health outcomes.

In sum, the present research makes several contributions. First, this study applied a stress framework to explain the role of cognitions in noninvolved partners’ mental health consequences and health-compromising behaviors after infidelity. Viewing infidelity as a relationship stressor, the results show that appraisals of the partners’ infidelity impact mental and physical health outcomes of noninvolved partners. Second, the findings reveal that noninvolved partners’ poorer mental health extends to their physical health such that depression, anxiety, and distress are associated with increased health-compromising behaviors after infidelity. Finally, this study may help inform clinical work on infidelity. Specifically, this research helps inform practitioners that noninvolved partners’ perceptions and appraisals of the infidelity are important, and that infidelity affects their physical health behaviors as well as mental health.

	1	2	3	4	5	6	7	8	9	10	11
1. Partner blame	—
2. Self-blame	−.06	—
3. Causal attribution	.40***	.18**	—
4. Depression	.28***	.40***	.15*	—
5. Anxiety	.24***	.36***	.21***	.81***	—
6. Distress	.28***	.49***	.31***	.68***	.67***	—
7. Mental health composite	.32***	.50***	.25***	.93***	.87***	.89***	—
8. Health-compromising behaviors	.14	.17*	.14	.32***	.29***	.27***	.31***	—
9. Relationship length in years	−.04	.09	−.10	.14*	.17*	.12	.14*	−.12	—
10. Sexual/physical infidelity	.05	.03	.09	.06	.07	.10	.07	.13	−.00	—
11. Emotional infidelity	.25***	.02	.18**	.16*	.08	.08	.13	.04	.05	–.23**	—
Mean	4.60	3.00	4.10	4.62	4.92	4.85	4.75	3.17	1.76	5.41	4.42
Standard deviation	1.14	1.60	1.08	1.11	1.32	1.31	1.10	2.77	1.91	1.75	1.93
Range	1–6	1–6	1–6	1-–7	1–7	1–7	1–7	0–13	1–17	1–7	1–7

	Mental health (M)	Health-compromising behaviors (Y)
Partner blame
Partner blame (X₁)	0.31***	0.07	0.17, 0.45	0.03	0.19	−0.34, 0.41
Mental health (M)	—	—	—	0.77***	0.20	0.39, 1.16
Emotional infidelity (U₁)	0.01	0.04	−0.07, 0.10	−0.01	0.11	−0.20, 0.23
Physical infidelity (U₂)	0.06	0.05	−0.03, 0.15	0.12	0.12	−0.11, 0.35
Relationship length (U₃)	0.01*	0.00	0.00, 0.01	−0.01	0.01	−0.02, 0.10
	R² = 0.13	R² = 0.11
	F(4, 173) = 6.77, p < .001	F(5, 172) = 4.15, p = .001
Self-blame
Self-blame (X₂)	0.31***	0.05	0.22, 0.40	0.08	0.14	−0.20, 0.36
Mental health (M)	—	—	—	0.73***	0.21	0.32, 1.14
Emotional infidelity (U₁)	0.04	0.04	−0.03, 0.12	0.02	0.11	−0.20, 0.36
Physical infidelity (U₂)	0.05	0.04	−0.03, 0.14	0.12	0.12	−0.11, 0.35
Relationship length (U₃)	0.00	0.00	−0.00, 0.01	−0.01	0.01	−0.02, 0.01
	R² = 0.24	R² = 0.11
	F(4, 173) = 13.90, p < .001	F(5, 1712) = 4.21, p = .001
Causal attribution
Causal attribution (X₃)	0.27***	0.08	0.12, 0.42	0.15	0.20	−0.24, 0.55
Mental health (M)	—	—	—	0.82**	0.19	0.41, 1.18
Emotional infidelity (U₁)	−0.02	0.04	−0.07, 0.11	0.02	0.11	−0.20, 0.23
Physical infidelity (U₂)	0.04	0.05	−0.05, 0.13	0.13	0.12	−0.10, 0.37
Relationship length (U₃)	0.01*	0.00	0.00, 0.14	−0.01	0.01	−0.02, 0.01
	R² = 0.10	R² = 0.12
	F(4, 1170) = 4.89, p = .001	F(5, 169) = 4.78, p < .001

	Mental health (M)	Health-compromising behaviors (Y)
Partner blame (X₁)	0.30***	0.07	0.15, 0.44	0.05	0.19	−0.33, 0.43
Mental health (M)	—	—	—	0.73***	0.19	0.35, 1.11
Gender (W)	−0.12	0.16	−0.44, 0.20	0.59	0.40	−0.21, 1.39
Partner blame × Gender (X₁ × W)	−0.09	0.15	−0.38, 0.20	0.58	0.39	−0.19, 1.34
Mental health × Gender (M × W)	—	—	—	−1.05**	0.40	−1.82, −0.27
Emotional infidelity (U₁)	0.02	0.04	−0.07, 0.11	−0.02	0.11	−0.23, 0.20
Physical infidelity (U₂)	0.06	0.05	−0.03, 0.15	0.10	0.11	−0.13, 0.33
Relationship length (U₃)	0.01	0.00	−0.00, 0.01	−0.01	0.01	−0.03, 0.01
	R² = 0.14 F(6, 171) = 4.65, p < .001	R² = 0.16 F(8, 169) = 3.96, p < .001

	Mental health (M)	Health-compromising behaviors (Y)
Self-blame (X₂)	0.31***	0.05	0.22, 0.40	0.07	0.14	−0.21, 0.34
Mental health (M)	—	—	—	0.68**	0.21	0.27, 1.09
Gender (W)	−0.06	0.15	−0.35, 0.24	0.60	0.41	−0.20, 1.40
Self-blame × Gender (X₂ × W)	−0.10	0.09	−0.28, 0.09	0.30	0.29	−0.27, 0.87
Mental health × Gender (M × W)	—	—	—	−1.06*	0.42	−1.90, −0.22
Emotional infidelity (U₁)	0.04	0.04	−0.04, 0.12	0.02	0.11	−0.19, 0.24
Physical infidelity (U₂)	0.05	0.04	−0.04, 0.13	0.12	0.12	−0.11, 0.34
Relationship length (U₃)	0.00	0.00	−0.00, 0.01	−0.01	0.01	−0.02, 0.01
	R² = 0.25 F(6, 171) = 9.44, p < .001	R² = 0.15 F(8, 169) = 3.82, p < .001

	Mental health (M)	Health-compromising behaviors (Y)
Causal attribution (X₃)	0.27***	0.08	0.12, 0.42	0.12	0.20	−0.27, 0.51
Mental health (M)	—	—	—	0.75***	0.19	0.37, 1.13
Gender (W)	−0.09	0.16	−0.41, 0.23	0.63	0.41	−0.17, 1.44
Causal attribution ×Gender (X₃ × W)	−0.13	0.15	−0.42, 0.17	0.16	0.39	−0.60, 0.93
Mental health × Gender (M × W)	—	—	—	−0.85*	0.39	−1.62, −0.07
Emotional infidelity (U₁)	0.02	0.04	−0.07, 0.11	0.00	0.11	−0.22, 0.22
Physical infidelity (U₂)	0.04	0.05	−0.05, 0.14	0.12	0.12	−0.11, 0.35
Relationship length (U₃)	0.01*	0.00	0.00, 0.01	−0.01	0.01	−0.02, 0.01
	R² = 0.11 F(6, 168) = 4.86, p < .001	R² = 0.16 F(8, 166) = 3.96, p < .001

Footnotes

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Note

References

Aiken

L. S.

West

S. G.

(1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage.

Allen

E. S.

Atkins

D. C.

Baucom

D. H.

Snyder

D. K.

Gordon

K. C.

Glass

S. P.

(2005). Intrapersonal, interpersonal, and contextual factors in engaging in and responding to extramarital involvement. Clinical Psychology: Science and Practice, 12, 101–130. doi:10.1093/clipsy/bpi014

Baker

Paquette

Szalai

Driver

Perger

Helmers

… Tobe

. (2000). The influence of marital adjustment on 3-year left ventricular mass and ambulatory blood pressure in mild hypertension. Archives of Internal Medicine, 160, 3453–3458.

Baum

. (2003). The male way of mourning divorce: When, what, and how. Clinical Social Work Journal, 31, 37–50. doi:10.1023/A:1021462517875

Bird

M. H.

Butler

M. H.

Fife

S. T

. (2007). The process of couple healing following infidelity. Journal of Couple & Relationship Therapy: Innovations in Clinical and Educational Interventions, 64, 1–25. doi:10.1300/J398v06n04_01

Boelen

P. A.

Reijntjes

(2009). Negative cognitions in emotional problems following romantic relationship break-ups. Stress and Health, 25, 11–19. doi:10.1002/smi.1219

Boelen

P. A.

van den Bout

van den Hout

M. A.

(2003). The role of cognitive variables in psychological functioning after the death of a first degree relative. Behavior Research and Therapy, 41, 1123–1136. doi:10.1016/S0005-7967(02)00259-0

Buss

D. M.

Shackelford

T. K.

Kirkpatrick

L. A.

Choe

j. C.

Lim

H. K.

Hasegawa

… Bennett

. (1999). Jealousy and the nature of beliefs about infidelity: Tests of competing hypotheses about sex differences in the United States, Korea, and Japan. Personal Relationships, 6, 125–150. doi:10.1111/j.1475-6811.1999.tb00215.x

Buunk

(1984). Jealousy as related to attributions for the partner’s behavior. Social Psychology Quarterly, 47, 107–112.

10.

Cano

O’Leary

K. D.

(2000). Infidelity and separations precipitate major depressive episodes and symptoms of nonspecific depression and anxiety. Journal of Consulting and Clinical Psychology, 68, 774–781. doi:10.1037/0022-006X.68.5.774

11.

Centers for Disease Control and Prevention. (2014). Behavioral risk factor surveillance system survey questionnaire. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention.

12.

Chapman

D. P.

Perry

G. S.

Strine

T. W

. (2005). The vital link between chronic disease and depressive disorders. Preventing Chronic Disease [serial online]. Retrieved from URL: http://www.cdc.gov/pcd/issues/2005/jan/04_0066.htm

13.

Choi

Marks

N. F.

(2008). Marital conflict, depressive symptoms, and functional impairment. Journal of Marriage and Family, 70, 377–390.

14.

Cohen

Doyle

W. J.

Skoner

D. P.

Fireman

Gwaltney

J. M.

Newsom

J. T.

(1995). State and trait negative affect as predictors of objective and subjective symptoms of respiratory viral infections. Journal of Personality and Social Psychology, 68, 159–169. doi:10.1037/0022-3514.68.1.159

15.

Cross

S. E.

Madson

(1997). Models of the self: Self-construals and gender. Psychological Bulletin, 122, 5–37. doi:10.1037/0033-2909.122.1.5

16.

Des Jarlais

D. C.

Semaan

S. M.

Arasteh

(2011). At 30 years: HIV/AIDS and other STDs among persons who use psychoactive drugs. In Hall

J. C.

Hall

B. C.

Cockerell

C. J.

(Eds.), Cockerell HIV/AIDS in the post-HAART era: Manifestations, treatment, and epidemiology (pp. 753–778). Shelton, CT: People’s Medical Publishing House.

17.

Eaton

W. W.

Smith

Ybarra

Muntaner

Tien

(2004). Center for Epidemiologic Studies Depression Scale: review and revision (CESD and CESD-R). In Maruish

M. E.

(Ed.), The use of psychological testing for treatment planning and outcomes assessment (Vol. 3, 3rd ed., pp. 363–377). Mahwah, NJ: Lawrence Erlbaum.

18.

Ezzati

Riboli

(2013). Behavioral and dietary risk factors for noncommunicable diseases. New England Journal of Medicine, 369, 954–964. doi:10.1056/NEJMra1203528

19.

Field

Diego

Pelaez

Deeds

Delgado

(2009). Breakup distress in university students. Adolescence, 44, 705–727.

20.

Field

Diego

Pelaez

Deeds

Delgado

(2013). Negative emotions and behaviors are markers of breakup distress. College Student Journal, 47, 516–523.

21.

Fife

S. T.

Weeks

G. R.

Stellberg-Filbert

(2013). Facilitating forgiveness in the treatment of infidelity: An interpersonal model. Journal of Family Therapy, 35, 343–367. doi:10.1111/j.1467-6427.2011.00561.x

22.

Fincham

Bradbury

(1992). Assessing attributions in marriage: The relationship attribution measure. Journal of Personality and Social Psychology, 62, 457–468. doi:10.1037/0022-3514.63.4.613

23.

Glass

S. P.

(2002). Couple therapy after the trauma of infidelity. In Gurman

A. S.

Jacobson

N. S.

(Eds.), Clinical handbook of couple therapy (3rd ed., pp. 488–507). New York, NY: Guilford.

24.

Glassman

A. J.

Helzer

J. E.

Covey

L. S.

Cottler

L. B.

Sterner

Tipp

J. E.

Johnson

(1990). Smoking, smoking cessation, and major depression. Journal of the American Medical Association, 264, 1546–1549. doi:10.1001/jama.1990.03450120058029

25.

Gordon

K. C.

Baucom

D. H.

(1999). A forgiveness-based intervention for addressing extramarital affairs. Clinical Psychology: Science and Practice, 6, 382–399. doi:10.1093/clipsy/6.4.382

26.

Gordon

Baucom

(2003). Forgiveness and marriage: Preliminary support for a measure based on a model of recovery from a marital betrayal. American Journal of Family Therapy, 31, 179–199. doi:10.1080/01926180301115

27.

Gordon

K. C.

Baucom

D. H.

Snyder

D. K.

(2004). An integrative intervention for promoting recovery from extramarital affairs. Journal of Marital and Family Therapy, 30, 213–231. doi:10.1111/j.1752-0606.2004.tb01235.x

28.

Gordon

K. C.

Friedman

M. A.

Miller

I. W.

Gaertner

(2005). Marital attributions as moderators of the marital discord-depression link. Journal of Social and Clinical Psychology, 24, 876–893. doi:10.1521/jscp.2005.24.6.876

29.

Hall

J. H.

Fincham

F. D.

(2006). Relationship dissolution following infidelity: The roles of attributions and forgiveness. Journal of Social and Clinical Psychology, 25, 508–522. doi:10.1521/jscp.2006.25.5.508

30.

Hatch

S. L.

Dohrenwend

B. P.

(2007). Distribution of traumatic and other stressful events by race/ethnicity, gender, SES, and age: A review of the research. American Journal of Community Psychology, 40, 313–332.

31.

Hayes

A. F.

(2009). Beyond Baron and Kenny: Statistical mediation analysis in the new millennium. Communication Monographs, 76, 408–420. doi:10.1080/03637750903310360

32.

Hayes

A. F.

(2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: Guilford Press.

33.

Hayes

A. F.

(2015). An index and test of linear moderated mediation. Multivariate Behavioral Research, 50, 1–22. doi:l0.1080/00273171.2014.962683

34.

Kelly

J. A.

Kalicbman

S. C.

(1998). Reinforcement value of unsafe sex as a predictor of condom use and continued HIV/AIDS risk behavior among gay and bisexual men. Health Psychology, 17, 328–335. doi:10.1037/0278-6133.17.4.328

35.

Kitson

G. C.

(1992). Portrait of divorce: Adjustment to marital breakdown. New York, NY: Guilford.

36.

Kobau

Bann

Lewis

Zack

Boardman

Boyd

… Lucas

(2013). Mental, social, and physical well-being in New Hampshire, Oregon, and Washington, 2010 behavioral risk factor surveillance system: Implications for public health research and practice related to healthy people 2020 foundation health measures on well-being. Population Health Metrics, 11, 19–19. doi:10.1186/1478-7954-11-19

37.

Laumann

E. O.

Gagnon

J. H.

Michael

R. T.

Michaels

(1994). The social organization of sexuality: Sexual practices in the United States. Chicago, IL: University of Chicago Press.

38.

Lazarus

R. S.

Folkman

(1984). Stress, appraisal, and coping. New York, NY: Springer.

39.

Lazarus

R. S.

Folkman

(1987). Transactional theory and research on emotions and coping. European Journal of Personality, 1, 141–169. doi:10.1002/per.2410010304

40.

Lepore

S. J.

Greenberg

M. A.

(2002). Mending broken hearts: Effects of expressive writing on mood, cognitive processing, social adjustment and health following a relationship breakup. Psychology & Health, 17(5): 547–560. doi:10.1080/08870440290025768.

41.

Mark

K. P.

Janssen

Milhausen

R. R.

(2011). Infidelity in heterosexual couples: Demographic, interpersonal, and personality-related predictors of extradyadic sex. Archives of Sexual Behavior, 40, 971–982. doi:10.1007/s10508-011-9771-z

42.

Pittman

(1989). Private lies. New York, NY: Norton.

43.

Prince

Patel

Saxena

Maj

Maselko

Phillips

M. R.

Rahman

(2007). No health without mental health. The Lancet, 370, 859–877. doi:10.1016/S0140-6736(07)61238-0

44.

Sapolsky

R. M.

(1999). Hormonal correlates of personality and social contexts: From nonhuman to human primates. In Panter-Brick

Worthman

C. M.

(Eds.), Hormones, health, and behavior: A socioecological and lifespan perspective (pp. 18–46). New York, NY: Cambridge University Press.

45.

Sbarra

D. A.

Law

R. W.

Lee

L. A.

Mason

A. E.

(2009). Marital dissolution and blood pressure reactivity: Evidence for the specificity of emotional intrusion-hyperarousal and task-related emotional difficulty. Psychosomatic Medicine, 71, 532–540. doi:10.1097/PSY.0b013e3181a23eee

46.

Shackelford

T. K.

LeBlanc

G. J.

Drass

(2000). Emotional reactions to infidelity. Cognition and Emotion, 14, 643–659. doi:10.1080/02699930050117657

47.

Sheppard

V. J.

Nelson

E. S.

Andreoli-Mathie

(1995). Dating relationships and infidelity: Attitudes and behavior. Journal of Sex and Marital Therapy, 21, 202–212.

48.

Spitzer

R. L.

Kroenke

Williams

J. B.

Löwe

(2006). A brief measure for assessing generalized anxiety disorder: The GAD-7. Archives of Internal Medicine, 166, 1092–1097. doi:10.1001/archinte.166.10.1092

49.

Sweeney

M. M.

Horwitz

A. V.

(2001). Infidelity, initiation, and the emotional climate of divorce: Are there implications for mental health? Journal of Health and Social Behavior, 42, 295–309.

50.

Tafoya

M. A.

Spitzberg

B. H.

(2007). The dark side of in delity: Its nature, prevalence, and communicative functions. In Spitzberg

B. H.

Cupach

W. R.

(Eds.), The dark side of interpersonal communication (2nd ed., pp. 201–242). Mahwah, NJ: Erlbaum.

51.

Tashiro

Frazier

(2003). “I’ll never be in a relationship like that again”: Personal growth following romantic relationship breakups. Personal Relationships, 10, 113–128. doi:10.1111/1475-6811.00039

52.

Turner

R. J.

(2010). Understanding health disparities: The promise of the stress process model. In Avison

W. R.

Aneshensel

C. S.

Schieman

Wheaton

. (Eds.), Advances in the conceptualization and study of the stress process: Essays in honor of Leonard I. Pearlin, (pp. 3–21). New York, NY: Springer.

53.

Walsh

J. L.

Senn

T. E.

Carey

M. P.

(2013). Longitudinal associations between health behaviors and mental health in low-income adults. Translational Behavioral Medicine, 3, 104–113. doi:10.1007/s13142-012-0189-5

54.

Weiner

(1979). A theory of motivation for some classroom experiences. Journal of Educational Psychology, 71, 3–25. doi:10.1037/0022-0663.71.1.3

55.

Weiner

(1986). Attribution, emotion, and action. In Sorrentino

R. M.

Higgins

E. T.

(Eds.), Handbook of motivation and cognition: Foundations of social behavior (pp. 281–312). New York, NY: Guilford Press.

56.

Wiederman

M. W.

(1997). Extramarital sex: Prevalence and correlates in a national survey. The Journal of Sex Research, 34, 167–174. doi:10.1080/00224499709551881

	Mental health (M)			Health-compromising behaviors (Y)
Predictors	b	SE	95% CI	b	SE	95% CI
Partner blame
Partner blame (X₁)	0.31***	0.07	0.17, 0.45	0.03	0.19	−0.34, 0.41
Mental health (M)	—	—	—	0.77***	0.20	0.39, 1.16
Emotional infidelity (U₁)	0.01	0.04	−0.07, 0.10	−0.01	0.11	−0.20, 0.23
Physical infidelity (U₂)	0.06	0.05	−0.03, 0.15	0.12	0.12	−0.11, 0.35
Relationship length (U₃)	0.01*	0.00	0.00, 0.01	−0.01	0.01	−0.02, 0.10
	R² = 0.13			R² = 0.11
	F(4, 173) = 6.77, p < .001			F(5, 172) = 4.15, p = .001
Self-blame
Self-blame (X₂)	0.31***	0.05	0.22, 0.40	0.08	0.14	−0.20, 0.36
Mental health (M)	—	—	—	0.73***	0.21	0.32, 1.14
Emotional infidelity (U₁)	0.04	0.04	−0.03, 0.12	0.02	0.11	−0.20, 0.36
Physical infidelity (U₂)	0.05	0.04	−0.03, 0.14	0.12	0.12	−0.11, 0.35
Relationship length (U₃)	0.00	0.00	−0.00, 0.01	−0.01	0.01	−0.02, 0.01
	R² = 0.24			R² = 0.11
	F(4, 173) = 13.90, p < .001			F(5, 1712) = 4.21, p = .001
Causal attribution
Causal attribution (X₃)	0.27***	0.08	0.12, 0.42	0.15	0.20	−0.24, 0.55
Mental health (M)	—	—	—	0.82**	0.19	0.41, 1.18
Emotional infidelity (U₁)	−0.02	0.04	−0.07, 0.11	0.02	0.11	−0.20, 0.23
Physical infidelity (U₂)	0.04	0.05	−0.05, 0.13	0.13	0.12	−0.10, 0.37
Relationship length (U₃)	0.01*	0.00	0.00, 0.14	−0.01	0.01	−0.02, 0.01
	R² = 0.10			R² = 0.12
	F(4, 1170) = 4.89, p = .001			F(5, 169) = 4.78, p < .001