Illness, communication and distress: The role of ambivalent communication when faced with potentially terminal illness
Table of Contents List of Tables p. v List of Figures p. viii Introduction p. 1 Emotional Expression p. 1 Written Expression p. 2 Written Disclosure and Illness p. 3 Summary p. 4 Verbal Communication and Life Threatening Illness p. 4 Communication between Family and Dying Persons p. 6 Qualitative Research p. 6 Quantitative Research p. 8 Communication and Psychological Distress p. 9 Summary p. 12 Ambivalence Over Emotional Expression p. 12 Ambivalence and End of Life Communication p. 15 Summary p. 16 The Present Study p. 17 Actor-Partner Interdependence Model p. 18 Hypotheses p. 19 Individual Hypotheses p. 19 Dyadic Hypotheses p. 20
Method p. 21 Participants p. 21 Measures p. 23 Procedure p. 26 Data Analyses p. 27 Individual Analyses p. 27 Dyadic Analyses p28 Results p. 30 Individual Analyses p. 30 Patients p. 30 Spouses p. 36 Dyadic Analyses p. 41 Discussion p. 46 Patients p. 47 Spouses p. 49 Patient-Spouse Dyads p. 50 Limitations p. 53 Clinical Implications p. 54 Future Research p. 56 Conclusion p. 58 References p. 60 Tables p. 68
Figure Captions p. 98 Figures p. 100 Appendix A p. 113 Appendix B p. 114
List of Tables Table 1. Intercorrelations between outcome measures p. 59 Table 2. Descriptive statistics for outcome measures p. 60 Table 3. Descriptive statistics for measures of expressivity p. 61 Table 4. Intercorrelations between communication measures, patients p. 62 Table 5. Correlations between expressivity and outcome scores, patients only p. 63 Table 6. Summary of Hierarchical Regression Analysis for Prediction of p. 64 Psychological Distress (CESD) by Ambivalence over Emotional Expression (AEEQ), Emotional Expressivity (EE), and their Interaction (AEExEE) among patients
Table 7. Summary of Hierarchical Regression Analysis for Prediction of Anxiety p. 65 (STAI) by Ambivalence over Emotional Expression (AEEQ), Emotional Expressivity (EE), and their Interaction (AEExEE) among patients.
Table 8. Summary of Hierarchical Regression Analysis for Prediction of Dyadic p. 66 Adjustment (DAS) by Ambivalence over Emotional Expression (AEEQ), Emotional Expressivity (EE), and their Interaction (AEExEE) among patients.
Table 9. Summary of Hierarchical Regression Analysis for Prediction of p. 67 Psychological Well-Being (SPW) by Ambivalence over Emotional Expression (AEEQ), Emotional Expressivity (EE), and their Interaction (AEExEE) among patients.
Table 10. Summary of Hierarchical Regression Analysis for Prediction of p. 68 Psychological Distress (CESD) by Ambivalence over Emotional Expression (AEEQ), Cancer-Related Communication (IRC), and their Interaction (AEExIRC) among patients.
Table 11. Summary of Hierarchical Regression Analysis for Prediction of p. 69 Anxiety (STAI) by Ambivalence over Emotional Expression (AEEQ), Cancer-Related Communication (IRC), and their Interaction (AEExIRC) among patients.
Table 12. Summary of Hierarchical Regression Analysis for Prediction of p. 70 Dyadic Adjustment by Ambivalence over Emotional Expression (AEEQ), Cancer-Related Communication (IRC), and their Interaction (AEExIRC)
Table 13. Summary of Hierarchical Regression Analysis for Prediction of p. 71 Psychological Well-Being (SPW) by Ambivalence over Emotional Expression (AEEQ), Cancer-Related Communication (IRC), and their Interaction (AEExIRC) among patients.
Table 14. Intercorrelations between communication measures, spouses only p. 72
Table 15. Correlations between expressivity and outcome scores, spouses only p. 73
Table 16. Summary of Hierarchical Regression Analysis for Prediction of p. 74 Psychological Distress (CESD) by Ambivalence over Emotional Expression (AEEQ), Emotional Expressivity (EE), and their Interaction (AEExEE) among spouses.
Table 17. Summary of Hierarchical Regression Analysis for Prediction of p. 75 Anxiety (STAI) by Ambivalence over Emotional Expression (AEEQ), Emotional Expressivity (EE), and their Interaction (AEExEE) among spouses.
Table 18. Summary of Hierarchical Regression Analysis for Prediction of p. 76 Dyadic Adjustment (DAS) by Ambivalence over Emotional Expression (AEEQ), Emotional Expressivity (EE), and their Interaction (AEExEE) among spouses.
Table 19. Summary of Hierarchical Regression Analysis for Prediction of p. 77 Psychological Well-Being (SPW) by Ambivalence over Emotional Expression (AEEQ), Emotional Expressivity (EE), and their Interaction (AEExEE) among spouses.
Table 20. Summary of Hierarchical Regression Analysis for Prediction of p. 78 Psychological Distress (CESD) by Ambivalence over Emotional Expression (AEEQ), Cancer-Related Communication (IRC), and their Interaction (AEExIRC) among spouses.
Table 21. Summary of Hierarchical Regression Analysis for Prediction of p. 79 Anxiety (STAI) by Ambivalence over Emotional Expression (AEEQ), Cancer-Related Communication (IRC), and their Interaction (AEExIRC) among spouses.
Table 22. Summary of Hierarchical Regression Analysis for Prediction of p. 80 Dyadic Adjustment by Ambivalence over Emotional Expression (AEEQ), Cancer-Related Communication (IRC), and their Interaction (AEExIRC) among spouses.
Table 23. Summary of Hierarchical Regression Analysis for Prediction of p. 81 Psychological Well-Being (SPW) by Ambivalence over Emotional Expression (AEEQ), Cancer-Related Communication (IRC), and their Interaction (AEExIRC) among spouses.
Table 24. Descriptive statistics for dyadic report of psychological outcomes and p. 82 expressivity measures.
Table 25. Correlations between patient and their spouse‟s report of communication p. 83 and distress measures
Table 26. Correlations between patient/spouse communication and outcome p. 84 variables – direct effects.
Table 27. Structural model fit using APIM to predict patient and spousal report p. 85 of psychological distress (CESD) from patient and spouse communication measures (AEE, EE, IRC).
Table 28. Structural model fit using APIM to predict patient and spousal report p.86 of anxiety (STAI) from patient and spouse communication measures (AEE, EE, IRC).
Table 29. Structural model fit using APIM to predict patient and spousal report p. 87 of relationship adjustment (DAS) from patient and spouse communication measures (AEE, EE, IRC).
Table 30. Structural model fit using APIM to predict patient and spousal report p. 88 of psychological well-being (SPW) from patient and spouse communication measures (AEE, EE, IRC).
List of Figures
Figure 1. Individual hypothesis 4: The relationship between EE/IRC and p. 91 distress will be moderated by AEE.
Figure 2. Illustration of dyad-level hypothesis testing APIM as a model p. 92 for the relationship between communication variables and psychological outcomes.
Figure 3. Patient report of psychological well-being as predicted by emotional p. 93 expressivity and ambivalence over emotional expression
Figure 4. Patient report of relationship adjustment predicted by cancer related p. 94 communication across levels of ambivalence over emotional expression
Figure 5. Patient report of psychological well-being predicted by cancer related p. 95 communication across levels of ambivalence over emotional expression
Figure 6. Full path analysis of dyadic model for prediction of psychological p. 96 distress among patients and partners.
Figure 7. Modified path analysis of dyadic model predicting psychological p. 97 distress among patients and spouses.
Figure 8. Full path analysis of dyadic model for prediction of anxiety among p. 98 patients and spouses.
Figure 9. Modified path analysis of dyadic model predicting anxiety among p. 99 patients and spouses.
Figure 10. Full path analysis of dyadic model for prediction of relationship p. 100 adjustment among patients and spouses.
Figure 11. Modified path analysis of dyadic model predicting relationship p. 101 adjustment among patients and spouses.
Figure 12. Full path analysis of dyadic model for prediction of psychological p. 102 well-being among patients and spouses.
Figure 13. Modified path analysis of dyadic model predicting psychological p. 103 well-being among patients and spouses.
Introduction The experience of individuals and their family members at the end of life can be quite varied. Research suggests that many factors (i.e., circumstances surrounding illness, demographic, and personality variables) contribute to distress experienced by both a patient and his or her loved ones following diagnosis of serious, life-threatening illness. Unfortunately, this area has been relatively neglected within psychological literature. Although psychological research has generally suggested that emotional expression can be predictive of improved outcomes, this finding has not been clearly established within samples at the end of life. Instead, results suggest that there may be other variables that strongly influence the relationship between emotional expression and psychological distress among those facing possible death. The present study investigated ambivalence over emotional expression as a moderator of the relationship between end of life communication and psychological distress among individuals diagnosed with cancer (patients), partners of cancer patients (spouses), and couples where one partner has been diagnosed with life threatening illness. Emotional Expression Emotional expression has been shown to be linked to the experience of psychological distress within several samples (e.g., Esterling et al., 1999; Rauer & Volling, 2005), including college undergraduates and married couples. Specifically, research has examined the use of both written (e.g., Pennebaker, Kiecolt-Glaser, & Glaser, 1988) and verbal (e.g., Salendar & Spetz, 2002) communication as factors that
may contribute to or protect from the development of psychological distress following a stressful life event. Written Expression The expression of emotion has often been described as a variable facilitating recovery following stressful life events (Esterling et al., 1999; Pennebaker, Kiecolt- Glaser, & Glaser, 1988; Pennebaker & Klihr Beall, 1986; Petrie et al., 2004; Roemer et al., 2001; Rosenberg et al., 2002). In a series of studies examining written disclosure of stressful life events, Pennebaker and colleagues (1988) found that after four consecutive days spent writing about negative life events for just 15 minutes, individuals reported short term physical and psychological distress immediately following disclosure. However, at follow-up, participants who wrote about negative life events were significantly happier and had fewer health center visits than those who wrote about neutral events at three-month follow-up. Results indicate that written disclosure, while stressful in the short-term, can provide long term physical and psychological benefit following negative life events. This model has been applied to various types of stressful events, ranging from sexual assault to motor vehicle accidents to suicide. Similar to the previous study, results indicate that when compared to writing about a control topic, writing about emotional or traumatic experiences predicts a drop in physician visits and self-reported illness (Greenberg & Stone, 1992; Pennebaker & Beall, 1986; Pennebaker & Klihr Beall, 1986; Pennebaker & O‟Heeron, 1984). However, results have been less clear in regards to changes in emotional health following written disclosure of stressful events. Even so, a
meta-analytic review of studies utilizing Pennebaker‟s model (Smyth, 1998) indicated that writing about emotion-laden tasks is significantly associated with reduced emotional distress. The review showed effect sizes (Cohen‟s d) ranging from .22-.82 with an overall average of .47. However, these studies generally failed to consider true traumas and major loss, so the implication of emotional disclosure on more severe events should be considered in greater detail. Written Disclosure and Illness. Rosenberg and colleagues (2002) applied Pennebaker‟s written disclosure paradigm to a prostate cancer population. Although participants in the written disclosure condition showed significant physical health benefits, they did not show improvements in psychological variables. However, given the short-term nature of intervention, it may be premature to conclude that improvement cannot result from written disclosure within this sample. Instead, it is possible that additional, longer writing periods might be required to facilitate relief of psychological distress. Similarly, a study of negative emotion suppression indicated that type of negative emotion (i.e., anger versus anxiety) expressed offered differential benefits for women diagnosed with breast cancer (Lieberman & Goldstein, 2006). Fifty-two women diagnosed with breast cancer were recruited from online bulletin board support groups. Participants completed measures of quality of life and depressive symptoms when they joined the support group and following six months of participation. Postings were analyzed for emotional content during this time period.
Results indicated that expression of anger and sadness was negatively associated with depression, whereas expression of anxiety was positively related to depressive symptoms (Lieberman & Goldstein, 2006). Importantly, there was no measure of level of anger, sadness, or anxiety experienced independent of that which was expressed in the online forums. Thus, it is unknown how overall level of negative emotionality may be related to either the expression of emotion, or the benefit received from such expression. Similar results were found within a sample of Chinese cancer survivors, where the inhibition of emotion was positively associated with stress, depression and anxiety, further supporting the idea that emotional disclosure is predictive of psychological health (Ho, Chan, & Ho, 2004). Summary. Overall, it appears likely that the written expression of emotion is a useful tool that may facilitate adjustment to stressful, negative life events. In many cases, expression of negative emotion allows for improved psychological and physical well- being. Such results extend to individuals experiencing both bereavement and diagnoses of serious physical illness. Still, it remains unclear whether written expression is comparable to verbal disclosure of emotional experiences. It remains possible that verbal expression to loved ones offers greater psychological benefits than does written disclosure of these experiences. Verbal Communication and Life Threatening Illness Although the written expression of emotion has been studied with some frequency, the disclosure of emotional experiences, particularly at the end of life, has been less thoroughly addressed. Despite the lack of research examining written
communication at the end of life, several studies have examined the role of verbal expression within close relationships, both when faced with terminal illness and independent of such stressors. Research suggests that in many cases, relationships change at the end of life. Many individuals report that they do not know how to react when faced with a dying loved one. In some cases, dying persons choose to withdraw from social relationships when faced with terminal illness (Nussbaum et al., 2000). Other times, persons might maintain relationships but avoid communication about issues related to death and dying (McGrath, 2004; Wright, 2003), particularly when the relationship is reported as very close (Hinton, 1981; Hinton 1998). Despite a desire to maintain emotional ties at the end of life, research suggests that this does not often occur. Families frequently find it difficult to communicate with doctors, friends, acquaintances, and other family members about the experience of coping with the illness of a loved one, and communication often becomes restricted during the illness (Northouse, 1984). Several qualitative studies (Hinton, 1981; Hinton, 1998; Salander & Spetz, 2002; Zhang & Siminoff, 2003) have shown that communication may become increasingly infrequent as death approaches, and that partners often grow apart during this process. Anxieties and concerns about death are rarely discussed and partners are rarely afforded opportunities to plan for the future (Beach, 1995). At the same time, research suggests that open communication may help the surviving partner to remain physically and emotionally close to their spouse, while accepting the impending death and preparing for life without his or her partner (Kramer, 1997).
Communication between Family and Dying Persons Very few studies (Beach, 1995; Hinton, 1981; Hinton 1998; Kramer, 1997; Northouse & Northouse, 1987; Salander & Spetz, 2002) have examined the role of communication between the dying and their families and friends. Studies have shown that although the patient is typically more willing to discuss their illness with family members than with doctors (McDonald et al., 2003), in most cases, minimal communication between the dying and loved ones occurs. When communication attempts are made, information shared is often limited to the sharing of factual information (i.e. type of illness, treatment information, life expectancy, etc.) between parties; few discuss difficult emotional or relationship issues surrounding the death of their loved one. Conversations about feelings or emotions surrounding death, the changing relationship, or the questions and fears they have about death are rare. Qualitative Research. Most research within death and dying relies on qualitative methods, most often semi-structured interviews that are analyzed for themes that occur during end-of-life communication (Beach, 1995; Salander & Spetz, 2002). These studies indicate that in most cases, caregivers avoid discussing issues related to terminal illness with the patient, often due to a desire to avoid distressing their loved one (Beach, 1995; Zhang & Siminoff, 2003). This has been shown to be particularly true when one or both individuals actively deny the seriousness of illness (Beach, 1995; Salander & Spetz, 2002). When both caregiver and patient shared an open awareness of terminal status, caregivers reported being more easily able to move on following bereavement, perhaps due to a greater ability to accept the impending loss (Beach, 1995). When partners did
not discuss terminal illness, many reported that they “grew apart” over the course of illness (Salander & Spetz, 2002). This suggests that end-of-life communication is likely to not only help ease the grief process following bereavement, but also may help maintain relational ties when faced with serious illness or dying. However, due to the small sample and lack of quantitative measures, this study fails to provide definitive conclusions on this topic. Zhang and Siminoff (2003) relied on qualitative measures to examine how communication occurs between dying persons and their family caregivers. Thirty-seven lung cancer patients and 40 caregivers participated in either focus groups or phone interviews where they discussed their communication about the cancer diagnosis. Data were collected at different times during the course of illness; some families participated at diagnosis and others participated when cancer was more advanced. In general, results showed that both patients and family members were reluctant to discuss the illness. Approximately two thirds of families reported that they struggled to communicate; about half of these families stated that they almost never discussed the illness. The remaining third of participants reported limited end of life communication. Some stated they made deliberate efforts to avoid conversations about cancer, often due to seeing their loved ones as either depressed or anxious. Zhang and Siminoff (2003) further described what they called the “phenomenon of silence.” Ten percent of their sample openly stated that they avoided all discussions of cancer. Many patients and family members reported that they did not want to think about the illness, as their fears and anxieties were reduced through avoidance of the topic.
Furthermore, the majority of patients reported that even when they wished to share their feelings about the illness, they sensed that their family did not wish to discuss it. Some reported that their caregivers would refuse to talk, in some cases because the family would refuse to believe the patient was dying. However, it is unclear if desire to share was mutual, or if such a need was only expressed by half the relational dyad. Though some participants noted that anxious feelings were reduced through avoidance of cancer- specific conversations, there was no specific measurement or examination of psychological distress. Although the study provided a detailed account of variables that influence participation in or avoidance of end-of-life communication, measurement of these variables would allow researchers to better understand the relationship between communication patterns and psychological distress. Quantitative Research. Although qualitative researchers have provided a first indication of the need for communication when faced with illness and/or dying, there is a general void of quantitative research within this area. To date, researchers have provided valuable and detailed descriptions of communication at the end of life; however, it is often unclear how the described patterns are related to psychological variables. Kramer (1997) attempted to determine if the relationship a woman shares with her terminally ill husband influences her adjustment post-bereavement. Results indicated that pre-loss acceptance was positively correlated to post-loss adjustment, whereas continued affiliation with one‟s spouse was not significantly related to this outcome. Pre- loss communication was significantly associated with a higher degree of happiness with one‟s life, but not to post-loss adjustment. However, aspects of adjustment studied by
Kramer have not been shown to have a significant relationship with long-term outcomes or with psychological distress. In a rare experimental study of communication in the face of life threatening illness, Hodgson, Shields, and Rousseau (2003) examined the role of disengaging communication in couples faced with breast cancer. Participants completed paper and pencil measures of depression, relationship satisfaction, and a general health survey. Additionally, participants completed a verbal task where they were expected to reach a consensus on a neutral topic. Results indicated that as more disengaging comments were made during the consensus task, higher levels of depression and lower levels of marital satisfaction were found within the dyad. Although it is interesting that stunted communication is related to poorer psychological outcomes, this study failed to examine the role of conversation about end of life issues or emotionally laden topics. It is unclear how disengagement around emotionally stressful conversations may influence distress, though it appears likely that this would be particularly distressing for individuals. With this in mind, it is possible that specific types of conversations might be more related to emotional distress than are others. Specifically, communication about end-of-life issues, diagnosis, and emotions related to this significant life event may be particularly associated with emotional distress experienced when faced with potentially life threatening illnesses. Communication and Psychological Distress. Studies that have examined end of life communication and its relationship to emotional well-being have been limited. Hinton (1981) attempted to determine the
relationship between open communication and levels of depression and anxiety of the spouses of terminally ill patients. Participants were engaged in informal interviews an average of ten weeks prior to their spouse‟s death. Interviews assessed the degree of open communication shared between partners, the feelings shared between the pair, and the level of anxiety and depression expressed by the surviving partner. Results indicated that the degree of open communication had no relationship to levels of depression or anxiety in the surviving partner. It is important to note that Hinton relied on self-report of communication and interviewer ratings of depression and anxiety, rather than utilizing quantitative measures of these variables. In a second study of end of life communication, Hinton (1998) found similar relationships between end of life communication and psychological distress. He found that persons who reported higher levels of anxiety were less open with the patient, whereas those who exhibited higher levels of depression were more open during communication with the patient. Unlike his previous findings, results suggest that communication may be related to both anxiety and depression, indicating a need to further examine these relationships. Additionally, this study only examined communication and outcomes of patients' loved ones; the emotional and psychological experience of patients was not considered. Zhang and Siminoff (2003) again replicated these findings. Twenty-four percent of patients reported symptoms of depression in addition to marked communication difficulties. Although statistical analyses comparing these phenomena were not discussed, the authors contended that failure to openly communicate may have resulted
from psychological distress. However, due to the descriptive nature of this study, it is difficult to conclude the precise relationship between psychological distress and communication. It appears likely that these variables are related; however, as neither was adequately measured, conclusions based solely on this study are premature and causal relationships should not be inferred. Finally, Metzger and Gray (2008) used a series of questionnaires to examine the extent to which communication that occurs at the end of life should match that which was desired by the surviving person. Sixty survivors of the death of a significant other completed measures of communication, depression, and grief following the loss of a significant other. Results indicated that, in most cases, the discreet amount of communication that occurs between a dying person and his/her loved ones is not significantly associated with distress following bereavement. However, as survivors reported higher levels of interaction with a loved one before his/her death, they also reported higher levels of grief following the loss. At the same time, results suggested that the degree to which participants‟ level of expression with their loved ones matched that which they desired was not related to bereavement outcomes. Interestingly, most significantly associated with psychological well-being following bereavement was a survivor‟s ability to accept an impending loss. Thus, those who reported the strongest ability to accept a loved one‟s death pre-bereavement also reported the least psychological distress following bereavement.
Summary Overall, much information is available regarding communication patterns at the end of life. It is clear that end of life communication tends to be limited, with much information being withheld amongst loved ones. Many individuals report fear and anxiety surrounding end-of-life conversations and emotional concerns, purportedly due to feelings of affection towards a dying loved one. Research in this area is marked by qualitative methodologies that, while providing rich, detailed description of end-of-life experiences, do not allow adequate conclusions to be drawn regarding the relationships between end-of-life communication and distress experienced by the dying and their loved ones. Further, available descriptions of communication at the end of life suggest that many persons diagnosed with life threatening illness, as well as their loved ones, experience ambivalence regarding appropriate expression of emotion at the end of life (Hinton, 1998; McGrath, 2004). Persons often desire communication, but are concerned that initiation of these conversations could contribute to others‟ suffering. It appears likely that ambivalence over emotional expression could contribute to psychological distress and poor relationship outcomes at the end of life. However, no studies to date have fully examined this possibility. Ambivalence Over Emotional Expression Ambivalence over emotional expression (King & Emmons, 1990) is a construct that describes conflict that individuals experience regarding the expression of emotion. Whereas emotional expressivity refers to the degree to which a person actually expresses emotion (persons might be high or low on this trait), ambivalence refers to the conflict
experienced about such expression. For example, a person might express a high level of emotion, but wish he or she had withheld this emotion. Conversely, one might fail to express emotion, but wish he or she was able to express emotions freely. Both of these people would be high on ambivalence, even though their actual expressivity is quite different. Research suggests that ambivalence over emotional expression (AEE), rather than actual emotional expressivity (EE), is associated with negative outcomes within samples of healthy persons (Emmons & King, 1988; Emmons & Colby, 1995; Katz & Campbell, 1994; King, 1993; King & Emmons, 1990; King, Emmons, & Woodley, 1992; Mongrain & Vettese, 2003; Mongrain & Zuroff, 1994). In a series of studies, King and Emmons (1990) examined the construct of AEE and its relationship to EE and other psychological variables within a sample of college undergraduates. First, results indicated that AEE and EE are negatively correlated; individuals who are highly ambivalent regarding emotional expression tend to express fewer emotions than do those who are less ambivalent. Second, and perhaps more interestingly, AEE was negatively associated with physical health and positively associated with psychological distress. In other words, highly ambivalent individuals reported more physical health problems and greater psychological distress than did less ambivalent persons. Additionally, ambivalence was negatively associated with health center visits and illness diagnoses within one year prior to the study. Interestingly, EE was not significantly associated with physical and psychological well-being, suggesting that AEE, rather than actual expression, is predictive of physical and psychological well-being. With this in mind, emotional
expression may not be the key to healthy emotion regulation, as has been noted in previous research. King and Emmons (1991) extended these results to married couples. Participants completed questionnaires measuring emotional expressiveness (EE), ambivalence over emotional expression (AEE), and emotional control (EC). Additionally, respondents completed measures of psychological functioning, life satisfaction, and the experience of various positive and negative emotions. Finally, participants rated their own and their spouse‟s physical health. Results indicated that EE was significantly negatively associated with AEE and with emotional inhibition and aggressive control. Results further suggested that participants who were highly emotionally expressive articulated less ambivalence, emotional inhibition, and greater control of aggressive impulses (King & Emmons, 1991). Additionally, AEE was positively associated with psychological distress and negatively associated with positive affect. EE was not significantly associated with psychological outcomes, suggesting again that ambivalence about the expression of emotions is more important to psychological well being than is actual expression. As was found previously, AEE was significantly negatively associated with physical well-being. Furthermore, AEE was negatively associated with estimates of one‟s partner‟s physical health. Although causality cannot be determined by this study, it appears possible that individuals were more ambivalent about expressing emotion when their partners were in poorer health. Overall, research notes a strong relationship between ambivalence about emotional expression and both physical and psychological well-being (Emmons & King,