DEPRESSIVE AND POSTTRAUMATIC SYMPTOMS AMONG WOMEN SEEKING PROTECTION ORDERS AGAINST INTIMATE PARTNERS: RELATIONS TO COPING STRATEGIES AND PERCEIVED RESPONSES TO ABUSE DISCLOSURE

Abstract

This investigation examined the relationship of abuse-specific coping strategies and perceived responses to abuse disclosure to symptoms of depression and posttraumatic stress among 131 women seeking a protection order against an intimate partner. Disengagement, denial, and self-blame coping strategies, as well as blaming of the participant by others, were associated with greater depressive and posttraumatic symptoms. None of the strategies of coping or responses to abuse disclosure were negatively related to depressive or posttraumatic stress symptoms. Findings suggest that mental health providers may find it useful to address these negative styles of coping while public education campaigns should target victim-blaming.

Each year, approximately 1.3 million women in the United States experience intimate partner violence (IPV; Tjaden & Thoennes, 2000). Almost two-thirds of these women will meet criteria for Posttraumatic Stress Disorder (PTSD) while almost half will experience depression (Golding, 1999), symptoms of which may play a role in the maintenance of abusive relationships. For instance, the amotivation and cognitive deficits associated with depression could prevent abused women from reaching out to services that could facilitate the end of the abuse (Sato & Heiby, 1992; Walker, 1989). Coping style and social support have been identified as potential buffers of the psychologically harmful effects of IPV (e.g., Calvete, Corral, & Estevez; 2008; Coker et al., 2002). However, we know little about which particular coping strategies and which particular aspects of social support are most closely related to abused women's mental health. Improving our understanding of the factors related to these negative mental health consequences will inform the design of mental health interventions for this population. This study examined the relationships between coping strategies, perceived responses to abuse disclosure, and depressive and posttraumatic stress symptoms among 131 women seeking Protection Orders (POs) against intimate partners. This is an especially important group of individuals to consider given earlier research documenting their high mental health burdens (e.g., Logan, Shannon, Cole, & Walker, 2006) and lack of connectedness to mental health services (Shannon, Logan, Cole, & Medley, 2006). Additionally, court processes have been known to exacerbate symptoms among those seeking legal intervention (Herman, 2003; Osenbach, Stubbs, Wang, Russo, & Zatzick, 2009).

Coping Among Women Experiencing IPV

Coping involves the use of cognitive and/or behavioral strategies to preserve psychological and physical well-being during stress. Strategies for coping have been variously categorized (e.g., problem-focused, or attempting to change the problem causing the distress, versus emotion-focused coping, or attempting to manage the emotional response to the problem; engagement coping, or strategies that increase proximity to the stressor, versus disengagement coping, or withdrawing from the problem and the associated emotional response). Previous studies of coping strategies used by victims of IPV have produced inconsistent findings: problem-focused coping has been found to be negatively related to depression (Clements & Sawhney, 2000; Mitchell & Hodson, 1983), positively related to depression and posttraumatic stress symptoms (Kocot & Goodman, 2003), and unrelated to posttraumatic stress symptoms (Arias & Pope, 1999). Emotion-focused coping has been found to be negatively related to depression (Mitchell & Hodson, 1983) and positively related to posttraumatic symptoms (Arias & Pope, 1999). Fewer studies have examined engagement versus disengagement coping among battered women, but these have also produced conflicting findings. One recent study found that engagement coping predicted better mental health and disengagement coping predicted worse mental health over time (Taft, Resick, Panuzio, Vogt, & Mechanic, 2007), while an earlier study found no association between engagement coping and posttraumatic symptoms (Kemp, Green, Hovanitz, & Rawlings, 1995).

Although these discrepancies may, in part, reflect sample or methodological differences, a review of the literature regarding coping among women experiencing IPV suggests that inconsistencies are at least partly attributable to different conceptualizations and measurements of coping styles across studies (Waldrop & Resick, 2004). In fact, the utility of examining broad distinctions in coping, such as between emotion and problem-focused or engagement and disengagement coping, has itself been questioned because a single coping strategy may serve both problem and emotion-focused functions (Skinner, Edge, Altman, & Sherwood, 2003). Moreover, each category is comprised of a diverse array of strategies (Skinner et al., 2003). A more fine-grained analysis that examines how specific coping strategies (such as planning, positive reframe, acceptance, humor, religion, using emotional support, using instrumental support, venting, disengagement, self-blame, self-distraction, active coping, denial, and substance use), rather than broad styles, relate to mental health among abused women may provide clarity on this subject.

Perceived Responses to Abuse Disclosure

Social support, or the provision of emotional, psychological, and physical resources by family, friends, and others, has been associated with well-being across a variety of populations and has been studied extensively as a buffer to negative outcomes for abused women (Coker, Smith, Thompson, McKeown, Bethea, & Davis, 2002). Controlling for IPV frequency, social support has been found to be negatively correlated with PTSD, depression, and other mental health problems in battered women (Anderson, Saunders, Yoshihama, Bybee, & Sullivan, 2003; Campbell, Kub, Belknap, & Templin, 1997; Coker, Watkins, Smith, & Brandt, 2003; Kaslow et al., 1998; Nurius et al., 2003). Thus, social support has been consistently related to mental health outcomes of abused women. Although deconstructing the concept of social support has rarely occurred in studies of abused women, it is likely that certain aspects of social support are more closely tied to mental health outcomes than others (Ullman, 1999).

One aspect of social support that may be closely tied to mental health of abused women is the responses of one's social support system to a disclosure of abuse. For various reasons, otherwise supportive individuals may be unable to provide support related to the abuse. In addition to abuser-imposed isolation, abused women may encounter family and friends who do not believe their reports of abuse, who blame them for the abuse, who are too frightened to help, or who become frustrated over time by the abused woman's failure to leave the relationship (Goodkind, Gillum, Bybee, & Sullivan, 2003). These individuals may also have their own relationships with the abusers and, as a result, prefer to refrain from “choosing sides.” Alternatively, friends and family may listen to the woman's story, empathize with her experience, or extend offers of help. Learning how specific responses to abuse disclosure correlate with distress is important to creating environments that foster recovery from IPV.

Studies of women who have been sexually assaulted suggest that the responses they receive from individuals in their social support system are particularly important to their mental health (Davis, Brickman, & Baker, 1991; Ullman, 1996). For instance, unsupportive responses to sexual assault disclosure, such as victim blaming, from women's support systems are negatively related to the rape survivor's well-being, whereas supportive behavior appears to have little relation to well-being (e.g., Davis, Brickman, & Baker, 1991; Ullman, 1996). In addition to demonstrating the relationship of perceived responses to abuse disclosure to mental health, these findings also highlight the distinct relationships of positive and negative social support to psychological distress (e.g., Gray & Keith, 2003; Lincoln, Chatters, & Taylor, 2003) and the importance of examining positive and negative responses to abuse disclosure as separate constructs.

Although one might expect similar findings for abused women, few examinations of social support among IPV victims have examined positive and negative social support separately (Nurius et al., 2003). Trotter and Allen (2009) documented that abused women regularly experience both negative and mixed reactions from their informal social networks and identified a need to explore the link between these negative responses and abused women's psychological health. However, to our knowledge, only Goodkind et al., (2003) has examined the effects of positive and negative responses on abused women's well-being separately. They found that family and friends’ negative reactions held the most explanatory power when predicting women's (lower) quality of life, while offers of a place to stay was the best (negative) predictor of depression. Emotional support was not a predictor of either well-being or depression. However, this sample was derived from a shelter population and composed entirely of women with low incomes. It is possible that a sample derived from a PO-seeking population may respond differently (e.g., offers of a place to stay may be less strongly related to symptoms of depression).

The Current Study

The current study examines abuse-specific coping strategies and perceived responses to abuse disclosure in a court-based sample of women seeking POs against intimate partners. It is important to note that, given the cross-sectional nature of this study, directions of causality could not be studied. Several questions and hypotheses were examined:

1. Which coping strategies are related to depressive and posttraumatic symptoms in women seeking POs against intimate partners? To resolve previous discrepant findings regarding the relation between coping and mental health among abused women, we conducted analyses at the level of specific coping strategies, rather than overall coping style (e.g., engaged vs. disengaged). To accomplish this, we chose to use the brief COPE scale (Carver, 1997), which includes 14 coping subscales that are not systematically subsumed under broader distinctions. Given previous discrepancies in the literature regarding the relation of coping styles to mental health, we refrained from offering hypotheses for these subscales.

2. Which perceived responses to abuse disclosure relate to depressive and posttraumatic symptoms? We hypothesized that offers of help (i.e., tangible support) would be negatively related to depressive and posttraumatic symptoms while blaming of the participant (i.e., negative support) would be positively related to these symptoms. Although we also viewed provision of emotional support as a positive response, results of previous studies (e.g., Davis, Brickman, & Baker, 1991; Ullman, 1996) suggested it would not be related to mental health symptoms. We did not offer any hypothesis about advice for the woman to leave or stay in the relationship, blaming of the partner, or maintenance of friendship with the abusive partner.

METHODS

Participants

Three hundred and eighty women petitioning for a PO against an intimate partner (defined as: married, divorced, or child in common) were approached for participation in this study. One hundred and ninety (50%) provided informed consent. Age and ethnicity were obtained from petition applications for all individuals approached. No significant differences were found between participants and non-participants on age or ethnicity. Of these 190, the 131 women (69%) who completed the instruments needed for these analyses comprise our sample. Of the 59 women who provided informed consent but were not included in our sample, 14 participated before the measures needed for the current study were added to the panel of instruments. The remaining women either did not return to the waiting room after their cases were heard or left before completing their measures as they decided not to complete the protection order process

Measures

Coping

The Brief COPE (Carver, 1997) is a 28-item self-report measure that includes 14 two-item subscales: planning (thinking about how to cope with a stressor), positive reframe (construing a stressful transaction in positive terms), acceptance (accepting the reality of a stressful situation), humor (making jokes about the stressful situation), religion (praying or finding comfort in one's religious beliefs), using emotional support (getting moral support, sympathy, or understanding), using instrumental support (seeking advice, assistance, or information), venting (ventilating feelings), disengagement (reducing one's effort to deal with the stressor, even giving up the attempt to attain goals with which the stressor is interfering), self-blame (blaming or criticizing oneself for the situation), self-distraction (using alternative activities to take one's mind off a problem), active coping (taking active steps to try to remove or circumvent the stressor or to ameliorate its effects), denial (refusal to believe that the stressor exists or of trying to act as though the stressor is not real), and substance use (using drugs or alcohol to feel better about the situation). Carver does not group the subscales into adaptive and maladaptive coping styles and instead recommends examining each scale separately to see identify its relation to other variables. Participants rate how often they used each response in trying to deal with the abuse using a 4 point Likert scale (1= I haven't been doing this at all; 4=I've been doing this a lot). Evidence of convergent and discriminant validity of the full scale was demonstrated by Carver, Scheier, and Weintraub (1989). In our sample, Cronbach's alpha was above .50 for 13 scales (α = .54 - .82), thus meeting the minimally acceptable reliability requirement established by Nunnally (1978). Because of low internal consistency (α = .44), we dropped the COPE Self-Distraction subscale from our analysis.

Perceived responses to abuse disclosure

Participants reported how frequently (never, rarely, sometimes, often) the person they talked to most often regarding the abuse (1) offered emotional support, (2) advised you to leave the relationship, (3) advised you to stay in the relationship, (4) blamed you or said you deserved the abuse, (5) blamed your partner, (6) offered help such as a place to stay or financial assistance, or (7) continued to spend time or stay friends with your abusive partner (adapted from Coker et al., 2002). Although Coker et al. (2002) combined these items (reverse scoring as appropriate) and conducted a median split to identify high and low disclosure support, we wanted to examine positive and negative responses separately, as we expected these items to distinctly relate to mental health. We were also interested in the relation of specific responses to depressive and posttraumatic symptoms. Thus, although we recognize the validity issues related to using single item responses, we decided that examining each item separately would be the more informative approach.

Depression

Women rated their depressive symptoms over the past week on the Center for Epidemiological Studies – Depression Scale (CES-D; Radloff, 1977). This 20-item scale has been used extensively with community samples, including battered women in the court system (Dutton, Goodman, & Bennett, 1999). Higher scores indicate greater depressive symptoms. A score of 16 has been used as the cutoff for high depressive symptoms (Radloff, 1977). Research on the CES-D among battered women in the court system has yielded evidence for adequate reliability (Dutton et al., 1999). In our sample, Cronbach's alpha was 0.81.

Posttraumatic Stress

The Modified Posttraumatic Stress Disorder Symptom Scale-Self Report (Falsetti, Resnick, Resick, & Kilpatrick, 1993) is a 17-item self-report measure adapted from the PTSD Symptom Scale (Foa, Riggs, Dancu, & Rothbaum, 1993) that assesses both frequency and severity of PTSD symptoms. The scale is validated for community samples with a wide variety of traumatic events and has demonstrated excellent internal consistency (Falsetti et al., 1993). We used the frequency subscale which yields a continuous score. Scores over 15 are suggestive of PTSD in community samples. In our sample, Cronbach's alpha was 0.92.

Variables examined as Potential Covariates

Intimate partner violence

Because abuse severity has been found to predict PTSD and depressive symptoms (Cascardi, O'Leary, & Schlee, 1999), social support (Coker et al., 2003), as well as coping (Waldrop & Resick, 2004), we evaluated this variable for use as a covariate. IPV victimization was assessed using the Conflict Tactics Scale – 2 Short Form (CTS2S; Straus & Douglas, 2004). Physical assault, injury, sexual coercion and psychological aggression were rated by the respondent on an 8-point likert scale from “this has never happened” to “this has happened more than 20 times in the past year” and summed to create an overall score. Items from the negotiation subscale, which measures nonaggressive means of resolving disagreement, were not included in this sum. Correlations of the short form with the full CTS2 scales (Straus, Hamby, Boney-McCoy, & Sugarman, 1996) range from .67 to .94. In our sample, Cronbach's alpha was 0.66.

Socioeconomic status (SES)

Given the consistently demonstrated relationship between socioeconomic status and mental health (e.g., Kessler, Price, & Wortman, 1985), we evaluated socioeconomic status as a possible covariate. To establish SES, we used neighborhood socioeconomic conditions derived from census data (U.S. Census, 2000), which have been shown to relate to rates of police-reported domestic violence (Pearlman, Zierler, Gjelsvik, & Verhoek-Oftedahl, 2003). The percent of people living below poverty in the area covered by each participant's zip code was used as the measure of SES.

Procedure

Women were approached in a secure waiting room of a family court in upstate New York as they waited to be seen by the judge regarding an OP. The PI and research assistants described the study to potential participants and conducted brief screens to determine eligibility based on the following criteria: (1) the woman was at least 18 years old, (2) she had experienced IPV by a current or former spouse or a man with whom she had a child in common (other types of intimate partners were not eligible to seek POs in family court in New York at the time of recruitment), (3) she spoke and read English, and (4) she showed no signs of significantly impaired mental status. Interested women completed the informed consent process and began a paper and pencil assessment. Women who could not complete the assessment packet during the initial session were scheduled to finish the procedure when they returned for their next court appointment (within a week). Women who completed the assessment were given a $20 gift card to a local supermarket.

Data Analysis

Prior to conducting the primary analyses, the participant's age, socioeconomic status, relationship to the perpetrator (married, divorced, child-in-common), ethnicity, and IPV severity were examined as possible covariates. Because bivariate correlations indicated no significant relationships with depression or PTSD, these variables were not included as covariates and were subsequently excluded from the analyses.

Bivariate relationships between coping strategies and depressive and posttraumatic symptoms were then examined, with p < .01 as the cutoff for significance to adjust for the high number of analyses. Next, bivariate relationships between social support responses to abuse disclosure and depressive and posttraumatic symptoms were examined, again using p < .01 as the cutoff for significance.

RESULTS

Participant Characteristics

Respondents ranged in age from 19 to 82 years, with a mean age of 34.0 years (SD = 10.1). Fifty-nine percent were employed. The sample self-identified as White or Caucasian (53%), African-American or Black (34%), and Hispanic or Latina (13%). Regarding their relationship to the perpetrator of the abuse, 36% were legally married (including separated), 4% were legally divorced, and 60% had a child in common but were neither married nor divorced. Participants’ scores show elevated depressive (M = 28.9; SD= 12.8) and posttraumatic symptoms (M=22.9; SD=13.6), compared to community samples.

Data Screening

There were no outliers (± 3 standard deviations from the mean) identified for any variables in the study. A number of variables in the study exhibited non-normal distributions. However, analyses using transformed and untransformed scores yielded identical patterns of significant results. Analyses using untransformed scores are reported for ease of interpretation. Cases that had missing data on specific variables were excluded only on analyses involving those variables.

Coping Strategies and Depressive and Posttraumatic Stress Symptoms

As shown in Table 1, none of the Brief COPE subscales were significantly negatively correlated with symptoms of depression or PTSD. However, , the disengagement (r = .37, .39, self-blame (r = .66, .47), and denial (r = .38, .47) subscales were all positively related (p < .001) to depression and PTSD symptoms, respectively.

Table 1.

Correlations between Study Variables

	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.	11.	12.	13.	14.	15.	16.	17.
1. IPV severity
2. SES	−.06
3. Depression	−.08	.00
4. PTSD	−.16	.18	.66**
5. Planning	.08	−.04	.03	.17
6. Positive Reframe	−.02	.02	−.01	.16	.42**
7. Acceptance	−.03	−.09	.08	.15	.52**	.42**
8. Humor	−.04	−.10	.16	.14	.01	.15	.23**
9. Religion	−.03	.25**	−.02	.13	.26**	.16	.36**	.16
10. Using emotional Support	.01	−.17	−.08	.04	.51**	.19*	.37**	.13	.05
11. Using Instrumental Support	.00	−.04	−.09	.02	.59**	.29**	.50**	.12	.16	.77**
12. Venting	−.03	.03	.15	.23*	.48**	.23*	.49**	.29**	.20*	.36**	.48**
13. Disengagement	−.18	.04	.37**	.39**	.01	.05	.10	.30**	−.06	.03	−.02	.21*
14. Self-blame	−.04	−.11	.66**	.47**	.18*	.16	.24**	.16	−.01	.11	.08	.28**	.36**
15. Active coping	.14	−.00	−.02	.06	.64**	.40**	.43**	.00	.26**	.48**	.58**	.43**	−.14	.06
16. Denial	−.23*	.13	.38**	.47**	.03	.18*	.12	.17	.09	.08	.09	.40**	.48**	.37**	.12
17. Substance use	.26**	−.04	.06	.13	.01	.14	.12	.29**	.05	.26**	.21*	.24**	.24**	.06	.11	.29**
18. “Offered emotional support”	.08	−.18	−.10	−.15
19. “Advised you to leave relationship”	−.07	−.22*	−.07	−.02
20. “Advised you to stay in relationship”	.08	.01	.06	.09
21. “Blamed you”	.01	.21*	27**	27**
22. “Blamed partner”	.05	−.06	.15	.17
23. “Offered help”	.05	−.10	−.12	−.04
24. “Continued to spend time with partner”	.00	−.00	.21*	.17

Perceived Responses to Abuse Disclosure and Depressive and Posttraumatic Stress Symptoms

Table 2 presents the frequency and percentage of women who report each response to abuse disclosure. Overall, women were more likely to report supportive responses than nonsupportive responses. Contrary to our hypotheses, “offered to help” was not significantly negatively correlated with symptoms of depression or PTSD (Table 1). However, corroborating our hypotheses about detrimental forms of social support, blaming of the participant (r = .27, .27) was significantly related to depressive and posttraumatic symptoms (p < .01), respectively.

Table 2.

Frequency and Percentage of Participants who Report Responses to Abuse Disclosure

	Never	Rarely	Sometimes	Often
Emotional Support	6 (4.8%)	13 (10.5%)	45 (36.3%)	60 (48.3%)
Advise to leave relationship	9 (7.3%)	7 (5.6%)	35 (28.2%)	73 (58.9%)
Advise to keep relationship	66 (53.2%)	34 (27.2%)	18 (14.5%)	6 (4.8%)
Blame you	98 (79.0%)	13 (10.5%)	10 (8.1%)	3 (2.4%)
Blame partner	17 (13.7%)	10 (8.1%)	42 (33.9%)	55 (44.4%)
Offer help	20 (16.1%)	12 (9.7%)	46 (37.1%)	46 (37.1%)
Remained friends with partner	77 (62.1%)	23 (18.5%)	17 (13.7%)	7 (5.6%)

DISCUSSION

This investigation examined the relationship of coping strategies and perceived responses to abuse disclosure to symptoms of depression and PTSD among women seeking POs against intimate partners. Our findings indicate that denial, self-blame, and disengagement strategies of coping, as well as blaming of the victim by others, are associated with higher levels of depressive and posttraumatic symptoms. By contrast, none of the strategies of coping or responses to abuse disclosure were negatively related to depressive or posttraumatic stress symptoms.

Coping

Although a longitudinal study would provide more definitive evidence regarding coping strategies as actual buffers to the negative mental health impact of IPV, this study did not find that any coping strategies were associated with relatively lower levels of mental health symptoms. In contrast, three coping strategies were found to be significantly related to greater distress: women who used disengagement, denial, and self-blame strategies of coping were more likely to be experiencing symptoms of depression and PTSD. Although self-blame has been proposed as an adaptive coping strategy, serving to restore one's sense of control after interpersonal victimization (Janoff-Bulman, 1979), prior empirical studies have identified self-blame as a common response that is strongly related to poor mental health (e.g., O'Neill & Kerig, 2000; Weaver & Clum, 1995). The current study replicates this finding.

Although Carver (1997) conceptualized denial and disengagement styles of coping as distinct concepts and the two demonstrate only a moderate correlation within this sample, previous studies of abused women do not appear to have examined these variables as separate constructs. The two terms have, at times, been used interchangeably or as components of the same coping factor. In fact, in their review of the coping literature and recommendation of best practices in the categorizations of coping, Skinner et al. (2003) propose that both denial and disengagement compose a higher order family of coping, which they labeled “escape” coping. Given the similar results for the two constructs in this study and the overlap in the discussion of these two constructs in the literature, denial and disengagement styles of coping will be discussed together below.”

Consistent with the current findings, a positive relationship between disengagement/avoidant/denial strategies of coping and psychological distress has been noted by Clements and Sawhney (2000; depression), Calvete et al. (2008; depression and anxiety), Krause, Kaltman, Goodman, and Dutton (2008; PTSD), and Schnider, Elhai, and Gray (2007; Posttraumatic stress). Given that leaving the abuser may put the victim at risk for escalating violence (Campbell et al., 2003), it may be safer for some women and their children if they remain with the abuser. In these situations, disengagement or denial of the abuse may be adaptive with regard to their physical safety. On the other hand, these styles of coping prevent the use of more action-oriented coping strategies that could result in an end to the abuse or help the woman protect herself. Among the women in this sample, all of whom were seeking help to end the abuse, the use of denial and/or disengagement as a coping strategy related to greater symptomotology.

Perceived Responses to Abuse Disclosure

Contrary to hypotheses, and similar to the findings regarding coping strategies, none of the responses to abuse disclosure related to fewer depressive or posttraumatic symptoms. In contrast, blaming of the victim by her support network was significantly (positively) related to depressive and posttraumatic symptoms. These findings are consistent with several studies of victims of sexual assault (Davis et al., 1991; Ullman, 1996) that have also identified positive relationships, but failed to identify negative relationships, between social support responses to abuse disclosure and mental health symptomotology. This may indicate that, while it is important for friends and family to be supportive, it is even more important that they are not unsupportive. Blaming the victim may be particularly detrimental to the victim's mental health. Alternative interpretations also exist. Women who experience depression and posttraumatic stress symptoms may be especially sensitive to perceived blame from others or individuals with a negative response bias may be more likely to report both greater symptomotology and blame from others.

These findings stand in contrast to Carlson, McNutt, Choi, and Rose (2002), Coker et al. (2002), and Thompson et al. (2000) who found a buffering effect of social support on the mental health of battered women. These differences in findings likely result from differences in the underlying constructs measured. Both Carlson et al. and Thompson et al. examined more general perceptions of the availability of social support, as opposed to specific responses to abuse disclosure by a single individual, as examined in the current study. The relationship of specific responses to abuse disclosure to symptoms of depression or PTSD may vary, given the larger context of perceived availability of support. This suggests that both general social support and specific responses to abuse disclosure may be important to examine in relation to abused women's psychological health.

Alternatively, differences in findings may result from the current study's concurrent examination of both positive and negative aspects of social support. Most previous studies have examined only positive support or have reverse coded items measuring negative support and then combinedthem with items measuring positive support to form a single social support measure (e.g. Coker et al., 2002, which used the same response to abuse disclosure questions as used in this study, reverse coded and combined into one score). By not examining the positive and negative relationships between social support and mental health separately, their unique relationships can be obscured (Ullman, 1999). These findings highlight the importance of examining both aspects of social support in order to parse out the effects of each on women's well-being.

Limitations and Implications

Before discussing implications, it is important to describe limitations of this study. These data are cross-sectional, prohibiting any causal conclusions from being drawn. As mentioned above, it is possible that negative responses from a woman's support network lead to depression, or that depression leads to negative responses from the support network. Alternatively, a woman's depression may influence her perception of others’ responses to her disclosure of abuse. Furthermore, certain strategies of coping may be conflated with symptoms of depression or PTSD, such as self-blame and depression, or disengagement and denial and the avoidance features of PTSD. In addition, perceived responses to abuse disclosure were assessed only with regard to the person with whom the participant spoke most often about the abuse. It is possible that she received qualitatively different responses to the abuse disclosure from other individuals. It should also be noted that the response to abuse disclosure subscales consisted of one item each and the coping subscales consisted of two items each, resulting in some subscales having adequate, but lower than ideal, reliability.

It is possible that participants and nonparticipants had systematic differences (e.g., could not afford to miss work, more distressed by the court proceedings) that would limit the generalizability of our findings. Additionally, all subjects were drawn from a pool of petitioners seeking POs against intimate partners. Abused women who seek POs may differ from abused women who do not seek POs in their methods of coping (e.g., more problem-focused or engagement coping), in the responses of their support networks to their disclosures of abuse (e.g., more positive and less negative responses; more offers to help), or in having fewer practical impediments to ending the abusive relationship (e.g., women who seek POs are more likely to be employed full-time than women who do not seek POs; Wolf, Holt, Kernic, Rivara, 2000). That being said, this sample represents a group of highly distressed women seeking legal help to end IPV who may benefit from mental health services. This research informs appropriate targets for interventions with this group.

Several tentative recommendations can be considered regarding service outreach and effective treatment of women who seek POs against intimate partners. First, these findings suggest that assisting abused women to increase their use of coping styles that are considered to be adaptive may be less fruitful than helping women avoid those coping styles (disengagement, denial, and self-blame) that are closely related to poor mental health. Treatment models that address these maladaptive coping strategies include cognitive techniques that challenge distorted thinking, behavioral techniques that strive to reduce avoidance of traumatic material, and interpersonal approaches that promote social engagement. However, further research on the use of these interventions with abused women is necessary to ascertain the effectiveness as well as the safety of these techniques. It is possible, as noted earlier, that disengagement, denial, and self-blame, while associated with negative mental health outcomes, may in some way help to keep abused women physically safe. Therapist and client pairs should not attempt to change these coping styles without thorough consideration of the client's safety.

The findings regarding response to abuse disclosure point to the potential utility of public education about IPV and ways to help an abused friend in need. Community mental health programs and IPV services could reach out to social support networks of abused women in an effort to enhance the social support available (Yoshioka, Gilbert, El-Bassel, & Baig-Amin, 2003). If women who report victim blaming from support figures are open to including these individuals in treatment, therapists may be able to elicit more supportive responses. Information about how to support a friend or family member in an abusive relationship could also be disseminated in community gathering spots, such as churches, community groups, adult education programs, and beauty salons (Fraser, McNutt, Clark, Williams-Muhammed, & Lee, 2002). These findings also highlight the importance of support groups for women who, for whatever reason, may not receive the support they need from those in their established social support networks.

In summary, women who report the use of disengagement, denial, and self-blaming strategies of coping and/or perceive blame by others are more likely to experience symptoms of depression and posttraumatic stress disorder. Although we also examined coping strategies and responses to abuse disclosure that might potentially confer protection from the detrimental mental health effects of abuse, none were identified. These findings suggest that future research concerning abused women would benefit from more fine-grained analyses of coping, the examination of perceived responses to abuse disclosure separate from the perceived availability of more general social support, and the concurrent examination of positive and negative aspects of social support, preferably with the use of more sophisticated multivariate statistical techniques.

Biography

Sharon M. Flicker, Ph.D., is a clinical psychologist and an assistant professor at the University of Maryland University College Europe. Her research program focuses on the intersection of culture, social relationships and mental health. Her recent research examines women's coping and helpseeking behaviors in response to intimate partner violence as well as the responses of others to women's disclosures of abuse. She is particularly interested in ethnic differences in women's experiences of and responses to abuse and the factors that may influence these differences.

Dr. Catherine Cerulli is the Director of the Laboratory of Interpersonal Violence and Victimization (LIVV), Associate Professor, with the Department of Psychiatry in the School of Medicine and Dentistry at the University of Rochester. Her program of research focuses on whether enhanced mental health enables intimate partner violence victims to better navigate safety. She is particularly interested in the intersection of suicide and intimate partner violence. She was formerly an Assistant District Attorney in Monroe County, New York, where she created a special misdemeanor domestic violence unit. She has been working on issues surrounding domestic violence and child abuse since 1983 in a variety of capacities.

Nancy L. Talbot, PhD, is a clinical psychologist and associate professor of Psychiatry, University of Rochester School of Medicine and Dentistry. Her research is focused on treatments for women with chronic, complicated depression and trauma histories. Dr. Talbot is currently conducting a clinical trial of Interpersonal Psychotherapy for depressed women with sexual abuse histories in the community mental health center within the Department of Psychiatry.