“It was Pretty Scary”: The Theme of Fear in Young Adult Women’s Descriptions of a History of Adolescent Dating Abuse

The mental health impact of abusive adolescent dating rela- tionships has not been well described, but fear related to abuse has been reported. We elaborate the theme of fear in women’s descriptions of a history of adolescent dating abuse. A sample of community-based women, ages 19–34, who experienced an abu- sive dating relationship during adolescence (ages 11–20) was used. Data were analyzed via thematic analysis. Fear was a consistent and resonant theme. Three types of fear were identiﬁed: fear for self, fear for other relationships, and fearful expectation. These re- sults offer important insights into the impact of abusive adolescent relationships on women’s mental health.

types are highly valued, and the nursing profession has long prioritized relationships as an important aspect of comprehensive nursing care (Croom, Procter, & Couteur, 2000;Varcoe et al., 2004). This prioritization has been demonstrated in both mental health nursing practice and in nursing practice with women who are survivors of dating violence (Amar, 2007;Croom et al., 2000). Relationship-oriented nursing practices can thus play an important role in identifying and remediating the experience of fear among women affected by adolescent dating abuse (Renker, 2006).
Greater understanding of the effects of adolescent dating abuse is necessary to the provision of effective and comprehensive health care for adolescents and young adults (Ismail, Berman, & Ward-Griffin, 2007). These experiences have important implications for women's mental and physical health-in fact, some studies now suggest that mental health may drive physical health among abused women (Taft, Vogt, Mechanic, & Resick, 2007)-yet data on the impact of adolescent dating abuse on women's mental health is lacking. With this study, we sought to help remedy this gap in the nursing literature.

BACKGROUND
The value of relationships in adolescence cannot be overstated, and is a fundamental element in any examination of adolescent behaviors, motivations, thought processes, and experiences as they pertain to development and the transition to adulthood (Campbell, Torres, McKenna, Sheridan, & Landenburger, 2004;Hetherington & Stoppard, 2002;Wolfe, Jaffe, & Crooks, 2006). Dating relationships are no exception. Typically, the major source of socialization shifts from family to peer group in adolescence and adolescents discover relationships as a means of identifying themselves socially (Wiseman, 2002;Wolfe et al., 2006). Relationships can both indicate and convey social standing, and adolescents often incorporate the impressions of others in the formation of self-concept (Wolfe et al., 2006). The novelty of dating in adolescence is integral to developmental processes, and can cause young people to attend significantly to dating not only for themselves but also for others (Ismail et al., 2007). Dating and other types of relationships thus contribute to adolescents' social self-location, or what we have termed "sociolocation": self-identification and recognition of social integration through relationships.
Intimate relationships often become heavily weighted determinants of social standing-and components of sociolocationfor young women. Young women may value a dating relationship as an important indicator of attractiveness, likability, or gender role adherence-all potentially important factors in sociolocation and self-conceptualization. Further, stress related to the demands of socialization in adolescence has been shown to have a greater impact on young women than on young menrendering them more vulnerable to the pitfalls of decreased self-esteem, poorer self-concept, depression, and a decreased sense of coherence about the self (Banister & Schreiber, 2001;Connor, Steingard, Anderson, & Melloni, 2003;Hale, Van Der Valk, Engels, & Meeus, 2005;Räty, Larsson, Söderfeldt, & Larsson, 2005). The importance and influence of relationships is, as such, necessarily key to developing insights about young women's mental health (Fletcher, 2010;Wolfe et al., 2006).
Further, studies show that relationships are a high priority for young women, and that the desire for relationship stability can drive them to subsume their own health and other needs-even to the point of enduring abuse (Grych, Raynor, & Fosco, 2004;Hetherington & Stoppard, 2002;Ismail et al., 2007). Johnson and colleagues (2005) note that "developmental processes may make abuse a risk worth taking, from the adolescent perspective, to feel loved or proud of being attached" (p. 177). Structurally violent influences, such as gender bias and devaluation of adolescents' relationships, may also make it difficult for young women to report or seek help with abusive relationships (Banister & Schreiber, 2001;Chung, 2007;Elliott & Larson, 2004;Galtung, 2004;Ismail et al., 2007). Several studies suggest that young women are reluctant to disclose abuse even to those who may be best equipped to assist them, including parents, health care providers, clergy, or counselors, because of shame and fear of reprisals (Amar & Gennaro, 2005;Ashley & Foshee, 2005;Ismail et al., 2007;Johnson et al., 2005;Renker, 2006;Wolfe et al., 2003). The trauma of abusive dating experiences thus may come from the self-identified or anticipated identification of being an abused adolescent as well as from the actual experiences of victimization, and the fear that each of these engenders. This suggests the importance and potentially dangerous burden of fear in regard to adolescent dating abuse, as young women may fear both the abusive partner and the social ramifications of disclosing abuse.
The effect of fear as a predictor of adverse mental and physical health outcomes has not yet been explored in adult female survivors of adolescent dating abuse. It has, however, has been shown to be an important predictor of mental health symptoms and engagement in health-promoting behaviors in adolescent female populations (Raiford, Diclemente, & Wingood, 2009) and in populations of formerly abused adult women (Csoboth, Birkás, & Purebl, 2005). Greater understanding of the role, persistence, and effects of fear among young women who have experienced adolescent dating abuse will allow nurses to identify how and when these women may be experiencing fear related to such experiences and thus provide more comprehensive, supportive mental health care to these women.
Adolescent dating abuse likely has important implications for ongoing mental as well as physical health and as such should be considered by nurses seeking to provide mental health support to young women. The purpose of this study is to describe the experiences and health outcomes of young adult women who had previously experienced an abusive adolescent dating relationship. A specific aim is to discover how young women defined and described their experiences of abuse in these relationships and their health-both mental and physical-thereafter. Only young adult women who self-reported the past experience of an abusive relationship during their adolescence participated.

METHODS
The multiple method approach was used in order to contribute to the existing body of evidence on adolescent dating abuse while also responding to the "call for qualitative approaches to capture the complexities and heterogeneity of partner violence" (Foshee, Bauman, Linder, Rice, & Wilcher, 2007, p. 500). We report here a striking finding that specifically emerged from the qualitative data, but the quantitative data have been reported at greater length elsewhere (Burton, Halpern-Felsher, Rehm, Rankin, & Humphreys, in press). Prior to implementation of any study procedures, approval was obtained from the appropriate institutional review boards.

Setting and Recruitment
The study was carried out in central Virginia, in a largely rural county. Recruitment strategies included posting fliers in community locations, advertising in local print and electronic media, and snowball sampling via enrolled participants. Interested women contacted the principal investigator (PI), who screened all participants via telephone using a modified version of the Abuse Assessment Screen (AAS; Soeken, McFarlane, Parker, & Lominack, 1998). The modifications allowed the PI to confirm that the relationship of interest occurred during adolescence and assess for a pattern of abusive behaviors with the same partner. Any potential participant who responded affirmatively to one or more items on the modified AAS was invited to participate. Participants were enrolled from September 2008 through September 2010.

Sample
Participants were English-speaking, community-based young adult women, between the ages of 18 and 35 who identified as having a past experience of an abusive dating relationship while they were between the ages of 11 and 20. On average, the women had experienced the abusive adolescent relationship ten years prior to their participation in this study (M = 9.64; SD = 5.45 years; see Table 1). A total, convenience sample of women (N = 100) was used for the larger study. A purposively selected subpopulation of participants (n = 10) was invited participate in interviews.

Procedures
The PI talked with each interview participant at least three times: during screening, at completion of the questionnaires, and in the in-depth interview. The screening conversations lasted between 5 and 30 minutes, in the course of which the PI asked questions about and clarified details of the participants' experiences. Informed consent for the study was obtained at the first meeting with each participant. Participants completed a series of questionnaires on relationship history, abusive experiences, sexual experiences, risky behaviors, depressive symptoms, suicidal ideation, and demographic information. Demographic characteristics of the total sample and the subsample are provided in Table 1. Questionnaire completion took approximately one hour, and the PI discussed the questionnaires with the participant, and provided clarification and assistance as needed.

Purposive Subsampling
All participants were asked at the questionnaire completion meeting if they were willing to be re-contacted for an interview; all replied affirmatively. In order to more fully explore the complexity of adolescent dating abuse and its impact on subsequent life experiences, we conducted a purposive selection of participants for interview. This helped ensure maximal sample diversity given the research setting (Creswell & Plano Clark, 2011). The selection process involved iteratively reviewing the questionnaires and identifying women who reflected the changing demographics of the total sample as it was obtained. We then applied criterion sampling based on types of abuse reported, age, income, and ethnicity to minimize non-randomization bias as described by Collins (2010). Similar techniques have been successfully used elsewhere with populations of young women (Haglund, 2003;Ismail et al., 2007). Included in the subsample (n = 10) were those identified through this process, re-contacted within two weeks of questionnaire completion, and who then accepted the invitation to interview. The final subsample included participants from a diverse set of socioeconomic backgrounds and ages to enhance potential for completeness in descriptions of abusive relationships and health outcomes.

Qualitative Data Collection
Since completion of questionnaires preceded interviewing, the PI was able to discuss the relationship of interest with each participant before her interview and ascertain some specifics of each case. We did not directly review surveys with interview participants, but did ask them to reflect upon what they reported therein during the interview. The initial request of each interviewee was, "Please tell me about the relationship you mentioned when we spoke on the phone and referred to in the surveys." Interviews further employed a semi-structured interview guide to elicit narratives and contextual details of the relationship deemed by each participant to be abusive. Participants were interviewed for a minimum of one hour, during which they were asked to talk about their experience in an abusive relationship, their lives as adolescents and as adults, and their current perspectives on the experience. Open-ended and clarifying questions also covered events surrounding the relationship, termination of the relationship, and participants' views of the relationship with regard to their own health and behavior. Rather than ask participants to cite examples of mental symptoms or health-related problems, we asked if and how they thought the abusive relationship had affected their health. All interviews were digitally recorded and transcribed verbatim, under confidentiality agreement, by a professional transcriptionist.

Analysis
Data were analyzed via thematic analysis using ATLAS.ti TM software from Cleverbridge, Inc. Thematic analysis was particularly appropriate given the mixed methods design of the larger study because it has been described as a "bridging or translating" method between qualitative and quantitative research tactics (Boyatzis, 1998, p. vi). This inductive approach enables the investigator to acknowledge the individual's role in assimilating events and acclimatizing to changed personal environments as a result of the studied phenomenon (Braun & Clarke, 2006). In this study, the phenomenon was young adult women's past experience of adolescent dating abuse, and we sought to discover not only the historical events that comprised participants' abusive relationships but also the social and personal contexts in which these occurred.
Analysis began simultaneously with data collection in that the PI's field notes served to highlight striking descriptions or repeated phrases. The PI also reviewed each recording in order to screen for salient features such as changes in tone or emotional emphasis that might not be recorded textually. This review allowed for attention to prosody and paralanguage in participants' speech, so that we captured differences in emotional and intentional content of speech pattern and emphasis, and coded appropriately (Dupuis & Pichora-Fuller, 2010). Member checking also was used during each interaction, as well as with subsequent participants to establish ongoing, interpretive interaction with the evolving thematic threads. This involved the use of reflective and inductive statements, as well as textual comparisons of transcriptions, to assure social currency throughout data collection and analysis (Sandelowski, 1993).

Development of Focal Theme
The theme of fear emerged from the data through the iterative processes of inference and extraction, consistent with the definitional criteria for themes as established by DeSantis and Ugarriza who wrote that a theme is "an abstract entity that brings meaning and identity to a recurrent experience and its variant manifestations" (2000, p. 362). Since this was a descriptive study, the coding process was data-driven, and themes developed based on the data itself rather than pre-defined theoretical constructs (Braun & Clarke, 2006).
Initially, open coding identified incidents in which participants made direct statements of being afraid of someone or something, such as in this quote from a participant who experienced a physically and emotionally abusive relationship at age 15: "I was so scared, I really couldn't talk to anyone about it, because nobody knew, it was just my friends." These statements were used as in vivo codes, as shown in Table 2. The in vivo codes were gradually defined conceptually and used to generate a basic vocabulary describing the phenomenon, with illustrative examples from the data further establishing the code's meaning (Boyatzis, 1998).
Transcripts were then reviewed again to identify similar incidences in which fear was described rather than explicitly stated, such as when one participant described her fear of labeling her ex-boyfriend abusive, even after their breakup: "I felt like if I called it an abusive relationship, then people might think-I was still really into what other people thought, my image and how people perceived me, so I always thought that people would see that as me being vindictive for a boyfriend who broke up with me." This process thus identified latent thematic threads by integrating the in vivo code structure into the interpretive analysis, so that codes and their definitions were increasingly specified.
Next, codes were clustered according to what type of feeling was described, and clusters grouped into families that identified the type of fear. Each family was defined on the basis of the clustered codes, and our extensive conceptual memoing. This process linked important aspects of the data across interviews, thereby creating larger and more abstract thematic families describing participants' experiences (DeSantis & Ugarriza, 2000). The three families comprising the theme of fear are Fear for Self, Fear for Other Relationships, and Fearful Expectation. Initial codes, code clusters, and thematic families are shown in Table 2.

RESULTS
The subsample of participants interviewed (n = 10) included four who identified as Caucasian and six who identified as African-American. They ranged in age from 19 to 34 years, and had experienced the abusive relationship in adolescence an average of 10.46 (SD = 5.29) years prior to enrolling in the study. Demographic characteristics of the subsample are provided in Table 1. The Theme of Fear All participants in the subsample described episodes of fear related to their abusive adolescent dating relationships.

Fear for Self
Direct statements of fearing injury or loss during the abusive relationship were identified thematically as representing fear for self. This label reflects the direct relationship between a participant's recall or description of her feelings and fear or feeling afraid. Descriptions of fear for self were the most readily identifiable incidences of fear, and provided important, primary insights into how participants thought and talked about their experiences of feeling fear. In addition, less direct statements describing feelings of uncertainty, discomfort, or alarm were included if the interpretive context suggested that they were associated with feeling fear in some way. Fear for self sometimes related to potential for violence in the relationship, but also to concerns about how others' responses could affect the individual, or to alarming social situations-such as being threatened in public or having to lie to friends and family-that occurred because of the relationship.
The main source of this fear for self was the abusive partner, whether because of physical abuse, threats, or other frightening interactions. One participant explained that she feared staying in the relationship, but was also afraid to leave her partner because he scared her: Let's see, after [my child] was born, I think we split up after [that], and throughout the year, I kept coming back to him, because he used to scare . . . me. He used to call and say that he was going to pour kerosene around the house . . . and that he'd light it on fire while I was sleeping in it, and he'd constantly follow me around and tag, following us, seeing what we were doing and making threatening phone calls, so I was scared.
For many of the participants, the abusive relationship represented a first or early dating experience, and they were uncertain how to evaluate their feelings with regard to the relationship-even when they could identify feeling afraid. One participant explained that she felt fear both because of the violence she experienced in her relationship as well as because she found herself in an unfamiliar situation: "So probably after a month or two of whatever it was, if it was dating or whatever, he started to get very jealous and always wanted to know where I was, and it was pretty scary because I had never really dealt with anyone like that before. You know, I was very young, I'd never had a serious relationship . . . because I was a kid." Other participants described efforts to account for partners' behaviors that frightened or intimidated them by ascribing those feelings to their own inexperience: I attributed everything to the fact that he loved me. Everything. Even the stuff that was not cool. When he told me not to be friends with [other males], I, in my mind I said, oh, it's because he cares so much about me that he doesn't want to see me get hurt by anybody. . . . I just assumed that everything was, I literally assumed that this was how relationships worked, because I would see my parents fight all the time, and so I assumed that, while my parents had been married for like 20 something years, so if they fight, it's natural that [he] and I fought a lot. I never, I never saw that I was always apologizing; I never put two and two together. I didn't want to look at it too hard, I just didn't.
In this passage, the participant talked about her sense that her boyfriend's behaviors were "not cool," suggesting her own discomfort or worry, but because she believed "that he loved me," she sought ways of reassuring herself, and avoided having to "look at it too hard," such that she might have had to face feeling afraid. This passage demonstrates latent evidence of fear, in that the speaker does not specifically state that she felt fear-although she did in other portions of the interview-but she does reference feelings of uncertainty, discomfort, and worry about what was happening in her relationship. Taken together, these descriptions suggest a fear of something that was "not cool" in her relationship.
Not all incidences of fear for self involved the abusive partner. Participants described living consistently with distinct fear of others' responses to them or to the abusive relationship, particularly parents and friends. As one woman said, "I was so scared of my parents finding out, my dad, and I did not want them to find out anything. . . . My dad and my mom did not know . . . I was just scared to death of them, and I didn't feel like I could talk to them about anything. So they were my number one fear." Although she did not specify what would have happened if her parents found out about the abusive relationship, this participant clearly described fear of the possibilities and characterized that fear as "number one." Another participant described similar fears of others' reactions or responses, noting that, "Whenever like, his family's around, or my friends, I just wanted to hide. Because of the way he was keeping me under his thumb, and everyone . . . seeing it." In these examples, fear came from others' perceptions of the abuse as well as from the abuse itself.

Fear for Other Relationships
In addition to fearing others' reactions, participants described fearing what might happen to other relationships if anyone discovered that they were being abused. This thematic group includes those instances in which participants described fearing an outcome that had to do with social status or relationship to others-how they potentially sociolocate themselves. This fear differed from the instances of fear for self that involved others in that fear for other relationships specifically involved potential changes in the timbre of those relationships. Fears in this group included losing social standing, being punished by parents or peer groups, and fear of what might result from either. The latter usually involved an anticipated outcome, with its possibility engendering the feeling of fear. Clusters included in this type of fear were "managing the situation" and "status changes." Managing the situation described efforts to prevent others from finding out the relationship was abusive, maintain positive relations with other people, make the partner happy, and cope with personal responses and feelings. One of the participants described the work she did to maintain the image of her relationship as a good one as "a very strenuous job" while another participant referred to similar efforts as creating "a lot of stress." In describing her struggle to be "the perfect girlfriend," to make her partner happy and avoid the possible consequences of him being unhappy, she said: I figured that the closer I could get . . . the happier [he] would be, and perfect meant not making him mad, and if I did, saying I'm sorry. . . . Literally always being there for him. Literally. Being happy and excited when he would surprise me at lunch, never being disappointed if he surprised me at lunch and that meant I couldn't do something else, never showed that disappointment. You're always happy to see your boyfriend, always. It's a lot of stress, I can tell you that. This passage clearly describes the participant's fear of her boyfriend's unhappiness or anger, as well as her fear of letting the image of their relationship as "perfect" slip in front of others who might judge her. Other participants also described seeking to preserve their relationship by preventing family and friends from discovering the abusive relationship, or by simply keeping quiet about their feelings. Many of the participants described defending either their partners or the relationship to others, or seeking to provide acceptable explanations of partners' behavior. This was seen as a means of preventing difficulty in other relationships, as well as avoiding strife in the dating relationship. One participant deflected her mother's questioning about bruises on her neck by saying, "he's just an enthusiastic kisser." While managing the situation included fears having to do with others learning about things between the couple, status changes delineated fears related to the potential loss of identity or social location markers of being in a particular relationship or being someone's girlfriend. Status changes encompassed fears related to ending or disclosing problems in the relationship, shifts in social group or status, or otherwise experiencing a change in status relative to others. This might be as simple as a shift in the dynamics of a parent-child relationship or as complex as losing membership in a peer group. One woman described trying to maintain the "couple image" with her abusive boyfriend because she did not want to lose her social standing: Here I am, all these girls in my class are skinny and they want to date this boy but he doesn't like her, and I'm like, here I am, I've got this 21-year-old guy who loves me, and he's in college, he's going places, or so I thought, so it was kind of like, it put me on their level. Because previously I had been down below a little bit, you know . . . and this kind of like brought me up into their world, it was like, oh, you know . . . she's dating a college guy. . . . It was one of the best feelings ever. Being 16, being in love and having someone love you. It was like a trifecta of happiness right there. The only thing better would have been getting a car, but [he] had a car, so I didn't really need one. For a while there, I had it made. I was the "it girl." Although this participant did not necessarily enter into the relationship because it afforded her particular status, it is clear that she valued the status she acquired as a result, enough that she feared its loss and what might then become of her socially. Another participant explained that she never told her friends why she broke off her relationship, because she feared she would be either be blamed "for not being able to see it coming," or for ending a relationship over something "people would think . . . wasn't a big deal." In this case, the participant's fear was of what others might think of her for being involved in an abusive situation, and what that might do to her social situation. Interestingly, several participants also mentioned that their ex-partners were now dating people they knew, and that they feared for the safety of these other women, but felt they could not approach them or that if they did, they would not be heard. This, too, illustrates the importance of social structures and interactions for these women-they did not want to disrupt social dynamics even if it meant ignoring risk of abuse to other women.

Fearful Expectation
A third thematic category, fearful expectation, developed to describe types of fear that were related less to things that had occurred than to those which might yet occur. Participants described this type of fear with words like "paranoid," "worry," or "not knowing what's going to happen." Included in fearful expectation were the "being on high alert" and "what might happen." Being on high alert describes how participants expressed fear in terms of vigilance. All participants described some degree of heightened alertness in the wake of the abusive relationship, whether in terms of fearing future interactions with the abusive partner, carefully checking out potential dating partners for themselves and others, fearing for their children, or questioning their judgment in subsequent romantic relationships. One participant felt that she had failed "to see the signs" of abuse from her partner, and that she "felt really bad, that I really should have seen this coming, and I didn't." Since then, she noted, "I just won't tolerate it. And if I'm ever with a friend and her boyfriend or something, I'll feel them out to see how they are, just because I don't think anyone should tolerate that." She also stated that she believed it was generally better to end a relationship prematurely than risk an abusive experience. In this case, the fear seemed to arise from a combination of feeling that the abuse could have been avoided and wanting to prevent it from happening again.
Other participants related instances of fearful expectations that affected their daily lives. One woman who had lived with and had children with her abusive partner gave a compelling description of continued vigilance: And to this day I still don't like to be around drama, or fights or anything else, anything loud. It really gets me, my stomach goes into knots. . . . My [most recent, not abusive] ex-boyfriend, who I just broke up with in January, he used to get so mad at me because he'd go . . . to brush his hair back, and I'd duck. And he used to get so mad, because he'd say, "I'm not going to hit you," and I'd be like, "I know that," it's just instant, you see that flicker out of your eye, and if you've trained yourself for four and a half years, you know, to protect, it's just this instinct, if I catch a motion out of my eye, whether it be you or my girlfriend, I'm still going to duck. It's not just you, I do it with anybody, even if I'm on the bus, and someone moves real quick or something, that's what I do.
She also described fears about the effects of the abusive relationship on her children, with particular concern for whether her sons might one day behave like their father. Another participant described being frightened for her daughters, given her own parents' lack of awareness during her abusive relationship: "Which scares me, too, that a 14-year-old can outsmart her parents. I've got three children, so . . . sometimes I think of all this as a blessing, to keep me, you know, [aware] of what's going on with my children, because I'm really going to make an effort to find out what's going on in their lives at that point, especially their teenage years." In all of these examples, fear led to a constant state of awareness and focus on preventing abusive experiences for oneself or others.
Similarly, the cluster "what might happen" was comprised of codes indicating fear of something that might be preventable, but for most of the incidents in which participants described these fears there was a less identifiable outcome. For example, one participant described fears that surfaced when she "got paranoid" that a more recent boyfriend was lying to her or cheating on her: "and I would say, you know, well, you came back at 2 o'clock in the morning; that means you were with somebody. I went off the deep end." Others described being afraid that their feelings might be disregarded in other types of relationships, such as with colleagues: And the thing that I find now is that I don't want anyone physically abusing me, or taking advantage of me, and I find now that even at work, if someone asks, "Oh, can you do this," I get so defensive because I feel like I'm being taken advantage of, that maybe I get a little too upset. "No," you know, "you're not going to do this to me." . . . I'm just so defensive about everything that's happened to me in the past, I don't want anyone walking on me now. It really affects me. I go through stages when I'm really upset with people at work for teeny, tiny things, but really, it's just because I don't want anyone treating me like I was treated when I was young. . . . it had a huge effect on me.
In this case, the participant felt strongly that her fears stemmed from her abusive experiences that had left her with a sense of needing to be constantly attentive to her situation-lest another relationship, even a work relationship, take a similar turn.
For others, loss provided a source of fear in some situations. While most of the participants stated that in the course of their abusive adolescent relationships they feared losing their partner, some were left with ongoing fear of having something valued turn out to be harmful-in a sense, "losing" it. In describing her abusive relationship as "a big old learning phase," one woman noted that she felt it was always possible to "lose this very important part" of herself, and that she viewed the physical scars left by her partner as a reminder of that. Elsewhere, the same participant explained that her most recent relationship had lasted several years and resulted in the birth of a child, but that she had chosen not to marry her partner because of her fears. Another participant described a serious adult relationship in which she realized that she had "always felt like I was living for something, for art, for [my abusive ex-boyfriend], for something, and in this amplified way I felt that all this 'for' had crumbled, that I didn't have anything, and I felt so purposeless, so empty and void. And I freaked out." Here, possible loss worked in tandem with fear in the new relationship, leaving this participant wondering if her past relationship had caused her to repeat a pattern that left her "empty and void."

DISCUSSION
Like adolescence itself, young women's experiences of relationships and dating during the adolescent years may be extremely complex. All of the interviewed participants in this study, now adults, described feeling fear related to their abusive adolescent dating relationships, and their descriptions of that fear were diverse. The details of these experiences suggest that there is an important role for mental health nurses in assisting women with a history of adolescent dating abuse to identify and address negative emotions related to their experiences of abuse.
Three typologies of fear emerged among the participants' accounts: Fear for Self, Fear for Other Relationships, and Fearful Expectation. Each category describes a particular set of fears as well as the settings in which fear emerged, and contributes to greater understanding of the complexity of young women's experience of adolescent dating abuse. Fear for Self described incidences where participants concretely or obliquely stated having felt afraid or scared. These statements usually pertained directly to fear felt during the abusive relationship, sometimes indicating fear of the abusive partner and sometimes indicating fear of other things. The striking descriptions of this type of fear demonstrated the strong emotional responses of participants to their abusive dating experiences, even long after the fact. On average, the women interviewed for this study were over ten years past the abusive adolescent relationship (see Table 1), yet they still recalled this type of fear in great detail. Their descriptions of feeling fear, being scared, afraid, or otherwise "freaked out" at the time of the relationship indicate that these young women were able to identify that something was wrong in their relationships, despite often being inexperienced and uncertain about expectations of dating partners. This was apparent even in cases where the fear stemmed from potential consequences, such as parents discovering the abusive relationship, rather than from events in the relationship itself. Renker (2002) studied 40 pregnant adolescents' stories of abuse in relationships, and noted the importance of specifically asking about fear when screening for abuse because it may be a means of discovering experiences not otherwise identified as abusive.
In addition to Fear for Self, the women in this study described feelings of fear related to the social location of the abusive relationship in their lives. Fear for Other Relationships was the theme assignment for fears involving some change in social status or necessitating strategies for managing the situation. These fears speak intimately to the temporal and developmental situation of being both an adolescent and an abused partner, in that the attached social situation was often a critical factor in how the participants evaluated their relationships. This is consistent with recent developmental literature that emphasizes the importance of relationships to young women, even if they perceive some of them as harmful or upsetting (Johnson et al., 2005). In fact, Martsolf, Draucker, Bednarz, and Lea (2011) found that adolescents at times intentionally listened to and disregarded the input of others regarding their dating relationships in order to maintain relationships with those others as well as with partners.
In developing the idea of sociolocation to describe how young women evaluate themselves through the timbre and nature of their relationships-dating or otherwise-we came to understand the poignancy of the risk these women associated with impact on their social situation and contacts. Banister, Jakubec, and Stein (2003) reported a similar dynamic in their focus group study of 40 girls aged 15 and 16. Their participants described confusion and conflict when they considered choosing between the isolation and danger of an abusive relationship and the isolation and unknown social status that would result from a break up. The researchers also noted the participants' preoccupation with pleasing their partners, or managing the role of the relationship in their lives-both of which were reiterated by participants in the present study.
Also in the present study, participants described Fear for Other Relationships related to the anticipated reactions of friends-either to terminating or staying in the abusive relationship, and some noted that they simply chose to keep silent rather than risk being judged. This is somewhat counter to Chung's (2007) inference that young women can be sources of support and intervention for each other with regard to abusive relationships. From a study of 25 young women aged 15 to 19, Chung reported on the importance of friends' opinions to those in abusive dating relationships-regardless of whether those opinions were positive or negative regarding the girl's actions. Both sets of findings, however, indicate the importance of relationships of all types to young women, and suggest that fear for other relationships may weigh as heavily on those in abusive relationships as does the fear associated with things that happen in the relationship itself. This weighted importance is echoed in current literature on adolescent development, which emphasizes the importance all types of relationships have for adolescents (Wolfe & Feiring, 2000;Wolfe et al., 2006). Adolescents develop a self-image through personal sociolocation, a means of framing self-concept based on with whom and in what ways they have relationships. As such, fear for other relationships represents an important component of the experience of abusive adolescent dating relationships for young women.
The last thematic category of fear, Fearful Expectation, is the fear that participants described as applying to events that might be yet to come or to situations they encountered after the abusive relationship. Fearful expectation represents the component of fear that may have the greatest impact on young women's health and well-being once an abusive dating relationship has ended, because as participants explained, this type of fear continued to impinge on their lives even into the present-for most, many years later. The clusters "high alert" and "what might happen" indicate how this fear took shape. Participants' descriptions of being vigilant, worrying about possible future occurrences, and having physical reactions to situations that reminded them of their abusive experiences were included in the high alert cluster. These descriptions resonate with those proffered by some existing research on survivors of other types of intimate partner abuse. Becker-Blease and  review the far-reaching effects of the fear and betrayal inherent in experiences of family violence, noting that such traumatic experiences have been frequently correlated with significant health issues such as disordered sleep and depression. The instances included in these clusters also suggest the experience of betrayal trauma, as described by Freyd, in which the hurt done by a trusted other causes a psychological injury in addition to the injury caused by the abuse (Freyd, Klest, & Allard, 2005). From their abusive adolescent relationships, these women had learned what it was like to be betrayed and they no longer failed to anticipate betrayal. We have further described elsewhere how betrayal trauma might affect young women who experienced adolescent dating abuse (Burton, Halpern-Felsher, Rankin, Rehm, & Humphreys, 2011). In addition, it is worth noting that symptoms such as hypervigilance, intrusive memories of abuse, and feeling unable to care for other people are among those diagnostic for post-traumatic stress disorder (PTSD) among formerly abused women (Taft et al., 2007). Both betrayal trauma and PTSD have been posited as risk factors for revictimization of abused women (Arata, 2000;Becker-Blease & Freyd, 2005;Kimerling, Alvarez, Pavao, Kaminski, & Baumrind, 2007).
Finally, the cluster "what might happen" captures feelings of fear related to unknown possibilities or potential losses. Here, participants described feeling uncertain or especially nervous, even in the present, when faced with situations that evoked responses similar to those they had in their abusive relationships. Fear that something similar could happen again, whether to the participant or to someone in her life, was a common thread. This is consistent with other literature on women's experiences of abuse. In a study of women ages 15 to 24, Csoboth, Birkás, and Purebl (2005) demonstrated a relationship between experiencing fear in daily life and a history of abuse either directly, such as by a partner, or indirectly, such as seeing one parent abuse the other.
Additionally, in the present study fear and loss were often described together. Some participants described the fear of losing their sense of self or gains they had made since the abusive relationship, while others explained that they were fearful of having another important emotional connection betrayed. In these ways, fearful expectation indicates how fear associated with the abusive relationship may resonate even long after that relationship has ended. Given the results of a recent study that suggest an association between increased psychological appraisal of threats and decreased leukocyte telomere length-a key indicator of cellular stress and aging-this ongoing experience of fear may have significant implications for young women's longterm health (O'Donovan et al., 2012). As such, interventions by mental health nurses that target stress reduction in tandem with interventions to lessen the burden of feeling these fears have significant potential to improve the health of women across the lifespan.

LIMITATIONS AND CONCLUSION
Our findings demonstrate the value of qualitative exploration of women's experiences with adolescent dating abuse. By exploring how women think about and describe these experiences, mental health nurses may discover new means of providing help and support to those involved in abusive relationships. In particular, the ongoing effects of fear described by participants underscore the need for routine screening for experiences of intimate partner abuse by all providers, with all patients, at all encounters. Our study was limited by its small sample size, lack of ethnic and socioeconomic diversity, and by geographic constraints. Future research should seek to expand the cultural diversity of participants so that a variety of racial, ethnic, and sociocultural considerations regarding fear can be accounted for in the study of adolescent dating abuse and its effects on health. Additional research is also and especially needed to discover the concerns of adolescents experiencing abuse in same-sex relationships, as well as those of young men who report abusive experiences. Increased depth and breadth of understanding of adolescents' experiences of dating abuse will undoubtedly be of benefit to mental health nurses and other mental health professionals serving this vulnerable population.

Declaration of interest:
The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.