Perceptions of parents and partners and men s depressive symptoms

Transcription

1 22S 03oliver (ds) 17/7/03 2:36 pm Page 621 Perceptions of parents and partners and men s depressive symptoms Luis E. Oliver & Valerie E. Whiffen University of Ottawa ABSTRACT Recent years have shown a marked growth in the literature examining the interpersonal context of depression. However, relatively little is known about the relational correlates of depression in men. The present study examined the relationships among men s recollections of childhood rejection by parents, their adult attachment security in current romantic relationships, and self-reported depressive symptoms. A mediational model of the associations between the childhood and adulthood variables was evaluated in a community sample of men. This model hypothesized that childhood rejection is linked to depressive symptoms through an impact on adult attachment security. Perceived parental rejection, childhood physical abuse, and insecure adult attachment were all correlated with depressive symptoms at the zero-order level, and path analysis supported the mediational model. Adult attachment insecurity fully mediated the link between perceived maternal rejection and higher levels of depression. In addition, paternal rejection was directly associated with adult depressive symptoms. These findings underscore the importance of the interpersonal context in understanding men s emotional functioning. KEY WORDS: childhood experiences depression intimate relationships men The research described in this article formed part of the first author s doctoral dissertation, and was conducted while he was supported by a doctoral fellowship from the Social Sciences and Humanities Research Council of Canada. The authors thank Andreas Reichert for his assistance with data collection and entry. Luis Oliver is now at the Faculty of Human Sciences, Saint Paul University, 223 Main St., Ottawa, Ontario K1S 1C4, Canada. All correspondence concerning this article should be addressed to Valerie Whiffen, School of Psychology, University of Ottawa, Ottawa, Ontario, K1N 6N5, Canada. [ whiff@uottawa.ca]. Barbara Sarason was the Action Editor on this article. Journal of Social and Personal Relationships Copyright 2003 SAGE Publications ( Vol. 20(5): [ (200310) 20:5; ]

2 22S 03oliver (ds) 17/7/03 2:36 pm Page Journal of Social and Personal Relationships 20(5) Although both sexes experience depression, women are twice as likely as men to be diagnosed with the clinical disorder or to report elevated levels of depressive symptoms (Nolen-Hoeksema, 1990). Perhaps as a result of this sex difference, researchers have tended to neglect the topic of men s depression (Cochran & Rabinowitz, 2000). In fact, most models of depression are based on research conducted predominantly with samples of women, with the result that we do not know how well these models generalize to men. In a community sample of men, we explored the associations between two sets of interpersonal variables that are predictive of depressive symptoms in women, perceived rejection by parents in childhood and adult attachment in the current romantic relationship. The first goal of the study was to determine whether these interpersonal variables also are relevant to men s depressive symptoms. The second goal was to evaluate a mediational model linking the childhood and adult variables. This model proposed that rejection by parents in childhood would influence adult attachment security, rendering individuals vulnerable to depressive symptoms. We identified two childhood variables that are relevant to men s depressive symptoms; parental rejection and physical abuse. A recent review concluded that parental rejection and excessive control predict subsequent depression and anxiety among men (Rapee, 1997). Childhood physical abuse also has been implicated as a risk factor for numerous psychological disorders in men, including depression (Duncan, Saunders, Kilpatrick, & Hanson, 1996; Gross & Keller, 1992). There is evidence that a significant number of men were physically abused as children (Malinosky-Rummell & Hansen, 1993), in part because of cultural practices that condone physical discipline and punishment for boys (Lytton & Romney, 1991). Childhood physical abuse may be experienced by the child as a severe form of parental rejection or it may occur as part of a larger context that includes parental rejection. Parental rejection may facilitate the development of a vulnerability factor that puts the individual at risk for later depression (Rohner, 1976). Attachment theory posits that the quality of childhood relationships with caregivers shapes the individual s working models of self and others (Bowlby, 1982). Children who experience rejection by their caregivers are hypothesized to develop a model of self as unlovable and a model of others as rejecting, and these negative working models are believed to be depressogenic (Bowlby, 1980; Cummings & Cicchetti, 1990). The exact mechanism by which insecure attachment is depressogenic has not been specified. Individuals classified as preoccupied and fearful-avoidant appear to rely heavily on interpersonal relationships to give them a sense of self-worth (Griffin & Bartholomew, 1994). However, their negative working models may affect how they experience and interpret stressful or ambiguous interactions and relationship events (Whiffen, Kallos-Lilly, & MacDonald, 2001). Thus, they may expect rejection from attachment figures, and such expectations could operate as self-fulfilling prophecies that contribute in various ways to actual rejection. Thus, one possible explanation for the link

3 22S 03oliver (ds) 17/7/03 2:36 pm Page 623 Oliver & Whiffen: Depressive symptoms in men 623 between childhood rejection and adult depression is that childhood rejection increases the likelihood of attachment insecurity later in life, which can be depressogenic. This hypothesis suggests that the association between childhood rejection and depressive symptoms is mediated by preoccupation/anxiety and fearful avoidance in adult attachment relationships. There is empirical support for various aspects of this hypothesis. In the developmental literature, adverse relations with parents have a negative impact on attachment security. Infants who are rejected by their parents tend to be insecurely attached to them (see review by Bretherton & Waters, 1985). In particular, they tend to avoid seeking comfort from the attachment figure, despite experiencing high levels of arousal indicative of fear or anxiety (Spangler & Grossman, 1993). Similarly, children who are physically abused by their parents are disproportionately likely to show disorganized patterns of attachment, with, again, high levels of avoidance beneath the disorganization (Lyons-Ruth, 1996). Thus, parental rejection during infancy and childhood clearly is associated concurrently and longitudinally with insecure, particularly avoidant, attachment. There also is evidence that adult attachment insecurity is associated with clinical depression and depressive symptoms. For instance, several studies have found cross-sectional correlations between depressive symptom levels and problematic attachment histories in both clinical (e.g., Pettem, West, Mahoney, & Keller, 1993) and nonclinical (e.g., Pearson, Cohn, Cowan, & Cowan, 1994 ) samples. It is notable that, among married women, clinical depression is associated specifically with fearful-avoidant attachment (Carnelley, Pietromonaco, & Jaffe, 1994, Study 2; Whiffen, Kallos-Lilly, & MacDonald, 2001). Fearful-avoidant individuals yearn for closeness and intimacy, but they also anticipate rejection from their attachment figures, a dilemma that they resolve by remaining emotionally distant even in close relationships (Bartholomew & Horowitz, 1991). Finally, there is evidence that attachment insecurity predicts the exacerbation of depressive symptoms over time among university students (Roberts, Gotlib, & Kassel, 1996), although this effect was not found in a sample of married and clinically depressed women (Whiffen et al., 2001). The extent to which working models formed in childhood are brought forward into adult relationships is presently undetermined. While some investigators believe that adult attachment security has more to do with the quality of adult relationships than with relations with previous attachment figures (Kobak, 1994), others point out that, as schemas, working models should summarize past experience (Baldwin, 1992). Empirical answers are complicated by the fact that infant/child and adult attachment are the foci of separate literatures, which are distinguished both conceptually and empirically by the use of different methods of measuring attachment. Research in the developmental literature suggests that there is stability between working models formed in childhood and those manifested in young adulthood when the same attachment figure is rated at both times (Waters, Merrick, Treboux, Crowell, & Albersheim, 2000). Hazan and Shaver (1987) also found that individuals who reported attachment security

4 22S 03oliver (ds) 17/7/03 2:36 pm Page Journal of Social and Personal Relationships 20(5) as adults described warm relationships with their parents, whereas insecurity was associated with perceptions of both parents as rejecting and uncaring. In two previous studies, we assessed the link between perceptions of parents and of current romantic partners. Although parents are the primary attachment figures of childhood, romantic partners are the primary attachment figures in adulthood. Thus, this research is another way of evaluating the stability of representations about key attachment figures. We found that perceptions of parents as rejecting were associated with perceptions of romantic partners as critical (Thompson, Whiffen, & Aube, 2001), and with anxiety about being rejected or abandoned by attachment figures in adult romantic relationships (Whiffen, Oliver, Emery, Varshney, & MacDonald, 2001). Thus, there seems to be support for the idea that representations of parents are linked to representations of adult romantic partners. Collectively, the research seemed to point in the direction of a mediational model, in which the links between childhood rejection and adult depressive symptoms were mediated by adult attachment security. A path diagram of the hypothesized model appears in Figure 1. Specifically, we hypothesized that attachment insecurity, as manifested by difficulties with intimacy and trust (Closeness) and fears around being abandoned or unloved (Anxiety), would be directly associated with higher levels of Depressive Symptoms. We further hypothesized that perceiving the mother as rejecting (Maternal Rejection) would be associated with avoidant attachment (low Closeness and high Anxiety) because maternal rejection is associated with avoidance in infants. This hypothesis is also in keeping with previous findings linking childhood rejection and avoidance in adulthood (e.g., Carnelley & Janoff-Bulman, 1992; Hazan & Shaver, 1987). However, we reasoned that the childhood relationship with the father would have relatively little impact on men s adult attachment security for two reasons. First, unlike mothers, fathers have not traditionally been the primary caregivers of children. Thus, we suspected that fathers would provide a weaker model of attachment relationships generally. We also believed that the relationship with the mother would have a more powerful impact than the relationship with the father on how a man relates to women in his adult life, in particular to a female attachment figure. Thus, we hypothesized that perceived Paternal Rejection would show weak or nonexistent associations with men s attachment security in romantic relationships. In contrast, we hypothesized that perceiving the father as rejecting would be directly associated with higher levels of depressive symptoms. Specifically, it was hypothesized that perceiving the father as rejecting in childhood (Paternal Rejection) would be associated with higher levels of Depressive Symptoms. Finally, we postulated a direct link between Physical Abuse and adult Depressive Symptoms, as well as an indirect link mediated by adult attachment security because of the negative impact of childhood physical abuse on attachment security during childhood.

5 22S 03oliver (ds) 17/7/03 2:36 pm Page 625 Oliver & Whiffen: Depressive symptoms in men 625 FIGURE 1 Postulated Causal Path Model Note. Depressive Symptoms = BDI. Maternal Rejection = MPARQ. Paternal Rejection = FPARQ. Physical Abuse = CTS. Closeness = RAAS Closedep. Anxiety = RAAS Anxious. D = error term associated with Depressive Symptoms: A = error term associated with Anxiety; C = error term associated with Closeness. Method Participants and procedure Potential participants were recruited via advertisements in the local media. To be sent a questionnaire package, callers had to meet the eligibility criteria of being male and in a current married or cohabiting intimate relationship of at least two years duration. Over a period of 9 months, a total of 125 inquiries

6 22S 03oliver (ds) 17/7/03 2:36 pm Page Journal of Social and Personal Relationships 20(5) were made regarding participation in the study. Of these potential participants, 25 men did not meet eligibility criteria. Of the 100 packages that were sent out, 80 were returned, yielding an 80% response rate. Of these 80 participants, 3 were excluded because they did not meet eligibility criteria. Another participant was deleted from the analyses due to his being a multivariate outlier, which yielded a final sample of 76 participants. Following data collection, a draw for $200 was held to compensate participants for their time. Measures Background information. Information was obtained on participants age, relationship status and duration, number of biological children, and number of children living with the respondent, as well as level of education, employment, and income. Depressive symptoms. The Beck Depression Inventory (BDI; Beck, Rush, Shaw, & Emery, 1979) is a well-established, 21-item self-report measure of the intensity of depressive symptoms. Meta-analyses by Beck, Steer, and Garbin (1988) yielded a mean alpha coefficient of.86, as well as test retest reliability estimates ranging from.60 to.90 (Beck et al., 1988). In the present study, the BDI yielded an alpha coefficient of.90, indicating excellent internal consistency. The BDI correlates well with other measures of depressive symptoms in clinical and nonclinical samples (Beck et al., 1988). Although the BDI is an established measure of the severity of depressive symptoms, it cannot be used to identify potential cases of clinical depression. Maternal and paternal rejection. The adult version of the Parental Acceptance Rejection Questionnaire (PARQ; Rohner, 1986) is a 60-item, self-report measure of respondents assessments of the level of warmth and acceptance they experienced as children. The psychometric properties of the PARQ have been assessed across numerous settings (Rohner, 1986). The PARQ s test retest reliability is adequate even at intervals of seven years (Cournoyer & Rohner, 1996), and internal consistency estimates range from.86 to.95. (Rohner, 1986). The PARQ has good construct validity as a measure of the acceptance rejection dimension of parenting (Rohner & Cournoyer, 1994). In the present study, participants completed two versions of the PARQ, one with reference to their mother (MPARQ), and the other with reference to their father (FPARQ). Both scales yielded an alpha coefficient of.97. Physical abuse. Four physical aggression items from the Conflict Tactics Scale (CTS; Straus, 1979) served as the measure of physical abuse in childhood. The CTS is a widely used measure of intra-familial conflict that has good test retest reliability (Straus, Hamby, & Sugarman, 1996; Widom & Shepard, 1996) as well as discriminant validity (Widom & Shepard, 1996). In the present study, the CTS yielded an alpha coefficient of.71, indicating an acceptable level of internal consistency. Adult attachment. The Revised Adult Attachment Scale (RAAS; Collins & Read, 1990) is an 18-item, self-report measure of adult attachment re-worded for the present study to apply specifically to the romantic partner or spouse.

7 22S 03oliver (ds) 17/7/03 2:36 pm Page 627 Oliver & Whiffen: Depressive symptoms in men 627 Test retest and internal consistency indices for the AAS are adequate. Factor analyses originally yielded three underlying dimensions: (1) Close (degree of comfort with closeness and intimacy); (2) Depend (degree of comfort depending on or trusting attachment figures); and (3) Anxiety (fears of being unloved and/or abandoned) (Collins & Read, 1990). However, findings that the Close and Depend scales are highly correlated have led many researchers to use the aggregate of the two scales in their analyses (C. Fraley, personal communication, December, 1999), which also provides a closer fit with Griffin and Bartholomew s (1994) two-dimensional model of adult attachment. Consistent with these findings, the correlation between the two scales in the present sample was found to be.81. Thus, the Close and Depend dimensions were collapsed for the analyses conducted in this study. The aggregate scale, RAAS Closedep, yielded an alpha of.89 for the current study, while the Anxious scale yielded a coefficient of.91. Results Sample characteristics Demographic characteristics of the sample are shown in Table 1. Generally, this was a well-educated, middle socio-economic status sample of men. Means and standard deviations for the study variables also appear in Table 1. Zero-order correlations among the variables are reported in Table 2. All of the study variables correlated significantly, and in the expected directions, with depressive symptoms. The mean BDI score for the sample was (with a standard deviation of 8.28), indicating that, on average, participants reported a mild to moderate level of depressive symptoms. This score was somewhat elevated relative to other community samples of men. Whiffen and Aube (1999), for example, found that the mean BDI score for the 62 men in their community sample was 7.4 (SD = 7.5). Our somewhat elevated score may reflect our recruiting strategy, which called both for men who were feeling relatively satisfied as well as those who were not. Test of the mediational model As mentioned earlier, the zero-order correlations reported in Table 2 show direct effects between all of the independent variables and the dependent variable. However, when the shared variance among the variables was considered in the path model described below, several effects became nonsignificant, lending support to the mediation hypothesis. The mediational model was formally tested with path analysis, using maximum likelihood estimation methods. Testing of the initially postulated model (see Figure 1) revealed a relatively poor degree of fit overall, with 2 (8, N = 76) = 93.04, p =.000; CFI =.93; ECVI = 2.01 (Saturated model =.96; Independence model = 16.64); RMSEA =.38. Although most of the hypothesized paths emerged as statistically significant, no paths from Physical Abuse were significant. No alternative paths from this variable to any other variable in the model were suggested by the modification indices. Post-hoc modifications to the postulated model were considered to bring goodness-of-fit estimates within acceptable parameters. Re-specification of the model took place in two stages. First, a re-specified path model, omitting the Physical Abuse variable, resulted in a significant improvement in fit, with 2

8 22S 03oliver (ds) 17/7/03 2:36 pm Page Journal of Social and Personal Relationships 20(5) TABLE 1 Means and standard deviations on the demographic and study variables M SD Demographic variables Age (years) Duration of relationship (years) Married (%) 66 Childless (%) 34 At least some college or university (%) 85.5 Employed (%) 86.9 Employed in white collar occupations (%) 67.1 Household income > $45,000 Cdn (%) 64.5 Study variables Beck Depression Inventory Mother PARQ Father PARQ Conflict Tactics Scale (Physical Abuse) RAAS Close/Depend RAAS Anxious Notes. Mother PARQ = Maternal Rejection; Father PARQ = Paternal Rejection; RAAS Close/Depend = Comfort with Closeness and Dependability of Partner; RAAS Anxious = Anxiety. TABLE 2 Correlations between the variables BDI MPARQ.47** FPARQ.37**.51** CTS.30**.58**.30* RAAS Closedep.64**.37** RAAS Anxious.55** ** 1.00 Notes. BDI = Depressive Symptoms; MPARQ = Maternal Rejection; FPARQ = Paternal Rejection; CTS=Physical Abuse; RAAS Closedep = Closeness & Dependability; RAAS Anxious = Anxiety. * p <.05; **p<.01. ( df =4, N = 76) = 52.00, p <.005. Second, a path was added from Anxiety to Closeness, as indicated by the modification indices. This path suggested that fears of abandonment or of being unloved by the partner contributed to difficulties with closeness and trust in the romantic relationship, a path that is theoretically tenable. Incorporating this parameter into the model resulted in

9 22S 03oliver (ds) 17/7/03 2:36 pm Page 629 Oliver & Whiffen: Depressive symptoms in men 629 improved fit, with 2 ( df =1, N = 76) = 36.55, p <.005. As the modification indices suggested that no further significant improvements in model fit could be made, this was considered the final model. Goodness-of-fit indices for the final model generally indicated a good to excellent degree of fit on all indices, with 2 (3, N = 76) = 4.49, ns; CFI =.998; ECVI = 0.53 (Saturated model = 0.55; Independence model = 10.01); and RMSEA =.08. The final model is presented in Figure 2. As hypothesized, significant paths emerged from the attachment dimensions (Closeness and Anxiety) to Depressive Symptoms. Significant paths also were FIGURE 2 Final Path Model Note. Parameter estimates are Standardized Regression (Beta) Weights indicating level of association between variables as specified by that particular path, and are significant at p <.05. Depressive Symptoms = BDI. Maternal Rejection = MPARQ. Paternal Rejection = FPARQ. Clseness = RAAS Closedep. Anxiety = RAAS Anxious. D = error term associated with Depressive Symptoms; A = error term associated with Anxiety; C = error term associated with Closeness.

10 22S 03oliver (ds) 17/7/03 2:36 pm Page Journal of Social and Personal Relationships 20(5) observed from perceived Maternal Rejection to both attachment dimensions, as predicted. The postulated path from perceived Paternal Rejection to Depressive Symptoms also was supported. Moreover, as anticipated, there was no support for paths from Paternal Rejection to adult attachment. Contrary to expectation, however, none of the hypothesized paths from Physical Abuse to the other variables emerged as significant. Discussion The primary goal of this study was to evaluate mediating hypotheses about the relationships among childhood rejection, adult attachment, and depressive symptoms. Additionally, the study sought to extend the research literature by studying these variables, which have been studied extensively among women, in a community sample of men. The data supported a number of the hypotheses regarding mediating relationships among the variables. Before describing the final path model, it should be noted that, at the zero-order correlation level, all the independent variables were significantly associated with depressive symptoms in the expected directions. However, when the shared variance among the variables was considered in the path model, a new picture regarding the relationships among these variables emerged. The significant childhood variables will be discussed first, followed by the significant adulthood variables and those variables that fell out of the path analysis. The data were consistent with the hypothesis that perceived maternal rejection in childhood is linked to men s depressive symptoms. Experiencing the mother as rejecting, hostile, critical, and indifferent as a child was associated with higher levels of depressive symptoms as an adult. These findings are congruent with those of previous research in suggesting that parental acceptance rejection is an important dimension in the development of adults emotional well-being (Rohner, 1986). The path model also supported the hypothesis that the link between maternal rejection and adult depressive symptoms is entirely mediated by adult attachment security. Specifically, feeling rejected by one s mother in childhood was associated with fears of being unloved and/or abandoned by the spouse, and with difficulties in developing closeness and trust. In turn, attachment insecurity was associated with depressive symptoms. These data are congruent with other research suggesting a degree of continuity in attachment representations across the lifespan (e.g., Waters et al., 2000). Although Bowlby (1980) did not believe that attachment was permanently set in childhood, he did suggest that there is a strong pull for continuity, perhaps because working models operate as cognitive schema that selectively process relational information (Baldwin, 1992). The present findings are consistent with the notion that attachment representations acquired via parent child interactions in childhood may be implicated in the creation of similarly insecure attachment relationships across the lifespan. This does not imply that a positive interpersonal relationship in

11 22S 03oliver (ds) 17/7/03 2:36 pm Page 631 Oliver & Whiffen: Depressive symptoms in men 631 adulthood cannot reorganize attachment in the direction of enhanced security in relationships, simply that the probability of such a corrective interpersonal relationship occurring may be low. With respect to the father son relationship, the data were congruent with the hypothesis that perceiving the father as rejecting, critical, indifferent, and hostile during childhood is associated directly with higher levels of adult depressive symptoms. Again, these findings are in keeping with those of previous research linking parental rejection to later emotional difficulties (Rohner, 1986). Although comparatively little empirical research has been conducted on the role that fathers play in the development of psychopathology (Phares & Compas, 1992), our findings lend credence to a substantial clinical literature suggesting that boys relationships with their fathers are of great importance for understanding their emotional functioning as adults (e.g., Corneau, 1991; Pittman, 1993; Real, 1997). As hypothesized, there was no evidence that the association between the father son relationship and adult depressive symptoms was mediated by adult attachment. Rather, the direct link between paternal rejection and depressive symptoms suggests that being rejected by one s father is depressogenic per se. It also is possible that another variable, not examined in this study, mediates this relationship. For example, the internalization of a rejecting father s excessively high standards could result in the development of perfectionism and/or self-criticism, personality traits that are linked to depression (Blatt, 1974; Hewitt & Flett, 1991). The above findings also are consistent with the notion that the impact of parental rejection may depend partly upon the sex of the rejecting parent, such that the relationship with the opposite-sex parent may be a particularly important model for later heterosexual relations. As mentioned earlier, we speculate that, because of their traditional role in the home, fathers may provide a weaker model of attachment for men, or at least for adult attachment in romantic relationships with women. Although relatively little research has been published on the differential effects of the same versus opposite sex parent on children (Conte, Plutchik, Picard, Buck, & Karasu, 1996), there is some evidence supporting this hypothesis. For instance, Collins and Read (1990) found similarities in college students perceptions of romantic partners and their perceptions of the caregiving style of their parents, which were most pronounced for participants opposite-sex parent. With regard to the adult variables under study, the data were consistent with the hypothesis that the nature of men s attachment to their spouses is linked to their levels of depressive symptoms. The findings supported the hypothesis that, for men, avoidance of closeness, lack of trust, and fears about rejection and abandonment in the primary attachment relationship are linked to higher levels of depressive symptoms. These findings support the links postulated by attachment theory between attachment insecurity and depression (Bowlby, 1980), and are consistent with a body of empirical evidence that links fearful-avoidant attachment, in particular, to clinical depression among married women (Carnelley

12 22S 03oliver (ds) 17/7/03 2:36 pm Page Journal of Social and Personal Relationships 20(5) et al., 1994; Whiffen et al., 2001). Interestingly, the modification indices suggested that a direct path from the Anxiety dimension of attachment to the Closeness dimension would contribute to a better fit of the model to the data. This could indicate that men who feel emotionally unsafe in relationships avoid closeness and depending upon their partners as a selfprotective strategy. Contrary to expectations, the path analysis did not support any links between childhood physical abuse and any other variables in the model. This variable was significantly correlated at the zero-order level with depressive symptoms. However, when considered in conjunction with the other variables in the model, its contribution was rendered statistically nonsignificant. This suggests that being physically abused in childhood was not predictive of depressive symptoms above and beyond the variance shared with parental rejection. Physical abuse may be experienced by the child as an extension of global parental maltreatment and rejection, which is a primary determinant of depressive symptoms later in life (Rohner, Bourque, & Elordi, 1996). One possible reason that childhood physical abuse has been linked to depressive symptoms may be that it tends to cooccur with parental rejection. Thus, it is possible that the association between depressive symptoms and physical abuse is better explained by the construct of parental rejection. In more general terms, our findings have important implications for the psychology of men. Contrary to cultural stereotypes, our results suggest that men s emotional functioning is very much intertwined with the quality of their interpersonal relationships. Our findings underscore the view that adult attachment needs are universal, and that their emotional impact transcends sex. The findings also highlight the importance of acceptance by parents in the healthy development of boys, and the damaging effects of rejection, despite a cultural emphasis on emotional toughness (Real, 1997). Thus, the findings support a large literature suggesting that much of an adult s emotional state is dependent upon the quality of relationships with others, whether that individual is male or female. There are several important limitations to the present study. The exclusive use of self-report measures limits the strength of the study s conclusions. Because self-reported depressive symptoms are not necessarily equivalent to a diagnosis of depression (Coyne, 1994), for example, the findings may not generalize to clinical depression. Future studies should extend this research to samples of clinically depressed men. Similarly, the use of retrospective self-report measures of childhood rejection raises questions about the accuracy of participants recall. However, a review of the research by Brewin, Andrews, and Gotlib (1993) concluded that adults generally do recall childhood events relatively accurately, and that there is little compelling evidence that depressed individuals selectively recall, exaggerate, or misinterpret the level of childhood adversity they report. Thus, there is little reason to believe that the results we obtained were due to a bias on the part of the dysphoric men in our study to perceive their parents negatively. Our study was also limited in terms of the specificity of

13 22S 03oliver (ds) 17/7/03 2:36 pm Page 633 Oliver & Whiffen: Depressive symptoms in men 633 the findings. It would be interesting to determine whether the model predicts only depressive symptoms, or other forms of psychopathology as well. Finally, the present study used a cross-sectional design to test a longitudinal model. Although the ordering of our variables was consistent with the results of a previous longitudinal study (Roberts et al., 1996), it is important to emphasize that we cannot rule out other potential causal orderings. In particular, the temporal ordering of adult attachment insecurity and depressive symptoms needs to be clarified in samples of couples. Future studies should employ longitudinal designs in addition to causal modeling techniques to more rigorously explore the causal links between these variables. REFERENCES Baldwin, M. W. (1992). Relational schemas and the processing of social information. Psychological Bulletin, 112, Bartholomew, K., & Horowitz, L. (1991). Attachment style among young adults: A test of a four-category model. Journal of Personality and Social Psychology, 61, Beck, A., Rush, A., Shaw, B., & Emery, G. (1979). Cognitive therapy for depression. New York: Guilford Press. Beck, A., Steer, R., & Garbin, M. (1988). Psychometric properties of the Beck Depression Inventory: Twenty-five years of evaluation. Clinical Psychology Review, 8, Blatt, S. (1974). Levels of object representation in anaclitic and introjective depression. Psychoanalytic Study of the Child, 29, Bowlby, J. (1980). Attachment and loss, Volume 3: Loss, separation, and depression. New York: Basic Books. Bowlby, J. (1982). Attachment and loss, Volume 1: Attachment. New York: Basic Books. Bretherton, I., & Waters, E. (1985). Growing points in attachment theory and research. Monographs of the Society for Research in Child Development, Vol. 50. Brewin, C., Andrews, B., & Gotlib, I. (1993). Psychopathology and early experience: A reappraisal of retrospective reports. Psychological Bulletin, 113, Carnelley, K., & Janoff-Bulman, R. (1992). Optimism about love relationships: General vs. specific lessons from one s personal experiences. Journal of Social and Personal Relationships, 9, Carnelley, K., Pietromonaco, P., & Jaffe, K. (1994). Depression, working models of others, and relationship functioning. Journal of Personality and Social Psychology, 66, Cochran, S., & Rabinowitz, F. (2000). Men and depression: Clinical and empirical perspectives. San Diego: Academic Press. Collins, N., & Read, S. (1990). Adult attachment, working models, and relationship quality in dating couples. Journal of Personality and Social Psychology, 58, Conte, H., Plutchik, R., Picard, S., Buck, L., & Karasu, B. (1996). Gender differences in recalled parental childrearing behaviors and adult self-esteem. Comprehensive Psychiatry, 37, Corneau, G. (1991). Absent fathers, lost sons. Boston: Shambala. Cournoyer, D., & Rohner, R. (1996). Reliability or retrospective reports of perceived maternal acceptance rejection in childhood. Psychological Reports, 78, Coyne, J. (1994). Self-reported distress: Analog or ersatz depression? Psychological Bulletin, 116, Cummings, E., & Cicchetti, D. (1990). Toward a transactional model of relations between attachment and depression. In M. Greenberg, D. Cicchetti, & E. Cummings (Eds.),

14 22S 03oliver (ds) 17/7/03 2:36 pm Page Journal of Social and Personal Relationships 20(5) Attachment in the preschool years: Theory, research, and intervention (pp ). Chicago: University of Chicago Press. Duncan, R., Saunders, B., & Kilpatrick, D., & Hanson, R. (1996). Childhood physical assault as a risk factor for PTSD, depression, and substance abuse. American Journal of Orthopsychiatry, 66, Griffin, D., & Bartholomew, K. (1994). Models of the self and other: Fundamental dimensions underlying measures of adult attachment. Journal of Personality and Social Psychology, 67, Gross, A., & Keller, H. (1992). Long-term consequences of childhood physical and psychological maltreatment. Aggressive Behavior, 18, Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52, Hewitt, P., & Flett, G. (1991). Perfectionism in the self and social contexts: Conceptualization, assessment and association with psychopathology. Journal of Personality and Social Psychology, 60, Kobak, R. (1994). Adult attachment: A Personality or relationship construct? Psychological Inquiry, 5, 42&ndash44. Lyons-Ruth, K. (1996). Attachment relationships among children with aggressive behavior problems: The role of disorganized early attachment patterns. Journal of Consulting and Clinical Psychology, 64, Lytton, H., & Romney, D. M. (1991). Parents differential socialization of boys and girls: A meta-analysis. Psychological Bulletin, 109, Malinosky-Rummell, R., & Hansen, D. (1993). Long-term consequences of childhood physical abuse. Psychological Bulletin, 114, Nolen-Hoeksema, S. (1990). Sex differences in depression. Stanford, CA: Stanford University Press. Pearson, J., Cohn, D., Cowan, P., & Cowan, C. (1994). Earned and continuous security in adult attachment: Relation to depressive symptomatology and parenting style. Development & Psychopathology, 6, Pettem, O., West, M., Mahoney, A., & Keller, A. (1993). Depression and attachment problems. Journal of Psychiatry and Neuroscience, 18, Phares, V., & Compas, B. (1992). The role of fathers in child and adolescent psychopathology: Make room for daddy. Psychological Bulletin, 111, Pittman, F. (1993). Man enough. New York: Putnam s. Rapee, R. (1997). Potential role of childrearing practices in the development of anxiety and depression. Clinical Psychology Review, 17, Real, T. (1997). I don t want to talk about it: Overcoming the secret legacy of male depression. New York: Scribner. Roberts, J., Gotlib, I., & Kassel, J. (1996). Adult attachment security and symptoms of depression: The mediating roles of dysfunctional attitudes and low self-esteem. Journal of Personality and Social Psychology, 70, Rohner, R. (1976). They love me, they love me not. New Haven, CT: Human Relations Area Files Press. Rohner, R. (1986). The warmth dimension: Foundations of parental acceptance rejection theory. Newbury Park, CA: Sage. Rohner, R., Bourque, S., & Elordi, C. (1996). Children s perceptions of corporal punishment, caretaker acceptance, and psychological adjustment in a poor, biracial southern community. Journal of Marriage and the Family, 58, Rohner, R., & Cournoyer, D. (1994). Universals in youths perceptions of parental acceptance and rejection: Evidence from factor analyses within eight sociocultural groups worldwide. Cross-Cultural Research: The Journal of Comparative Social Science, 28, Spangler, G., & Grossman, K. E. (1993). Biobehavioral organization in securely and insecurely attached infants. Child Development, 64,

15 22S 03oliver (ds) 17/7/03 2:36 pm Page 635 Oliver & Whiffen: Depressive symptoms in men 635 Straus, M., Hamby, S., & Sugarman, D. (1996). The revised Conflict Tactics Scales. Journal of Family Issues, 17, Thompson, J. M., Whiffen, V. E., & Aube, J. A. (2001). Does self-silencing link relationships and depressive symptoms? Journal of Social and Personal Relationships, 18, Waters, E., Merrick, S., Treboux, D., Crowell, J., & Albersheim, L. (2000). Attachment Security in infancy and early adulthood: A twenty-year longtitudinal study. Child Development, 71, Waters, E., Hamilton, C., & Weinfield, N. (2000). The stability of attachment security from infancy to adolescence and early adulthood: General introduction. Child Development, 71, Whiffen, V. E., & Aube, J. A. (1999). Personality, interpersonal context, and depression in couples. Journal of Social and Personal Relationships, 16, Whiffen, V. E., Kallos-Lilly, A. V., & MacDonald, B. J. (2001). Depression and attachment in couples. Cognitive Therapy and Research, 25, Whiffen, V. E., Oliver, L. E., Emery, P., Varshney, N., & MacDonald, B. J. (2001). Distal and proximal predictors of clinical depression. Manuscript submitted for publication. Widom, C., & Shepard, R. (1996). Accuracy of adult recollections of child victimization. Psychological Assessment, 8,

16 22S 03oliver (ds) 17/7/03 2:36 pm Page 636