Introduction

Midway through her third year of therapy, Alice,Footnote 1 a tall young woman with a cherubic face, began a session by telling me about her experiences on a recent date. Her words and the energy in her soft voice confirmed that the severe depression that had brought her to my office had lifted. She had been reporting feeling better for some time now, and her account that day reflected that she was happier, enjoying her new job, dating again, and seeing friends. Her appearance had changed as well. Previously neatly groomed in dark or muted colors, she was now wearing brighter ones, and the way she walked and gestured showed the same vigor as her voice. I liked Alice, but for reasons I did not yet understand she was someone with whom I frequently had to work to stay alert. However, in this session she made a comment that captured my attention completely: “I guess it’s time for me to talk to you about my bulimia.”

I asked her to tell me about it, my mind filled with questions. How could she have told me so many private and personal things but kept her bulimia hidden? Why had she waited two-and-a-half years to share such important information? Why was she telling me about it now? I wondered if she was afraid that I, like numerous people in her life, would withdraw my support from her now that she was feeling better. I briefly considered that she might be fabricating the symptom in order to keep me from discharging her from therapy, but she did not tend to falsify, and her symptom picture was actually similar to that of many individuals with bulimia with whom I had worked over the years. Her fear of my ending the therapy now that she was feeling better might have contributed to her revelation, but this simply returned me to the questions of how and why she had kept such a secret from me for all that time. I also started doubting both myself and our work. How had I missed this aspect of her makeup? Had I failed her as a therapist? And if so, could I repair this rupture, or was the work damaged permanently?

In the 1980s, when I worked at a special clinic for eating disorders, secrecy was the norm. Mail from the program was sent in unmarked envelopes and phone messages were left with no identifying information. Despite the widespread recognition of eating disorders today, some clients are still too ashamed of their symptoms to share them even with their therapists. Siebold (this journal) has noted that shame motivates clients to hide aspects of themselves even in a therapeutic relationship. When any symptoms do come to light, therapists often experience some of the concerns I had with Alice. They worry that a client may be more disturbed than they thought, or that they, the therapists, did something wrong. Although a therapeutic rupture can be part of the picture, these disclosures often reflect neither a therapist’s failure nor a more severe problem in a client. Instead, they often reflect something positive that has been quietly happening in the therapy: a client’s growing trust in the therapist’s presence and ability to help with feelings that have been, until now, dealt with through the eating behaviors themselves.

In my attempts to understand what happens before and after an individual shares a hidden eating disorder with a therapist, I have found helpful ideas in attachment theory, in particular those that explore links among attachment, affect regulation, and self-reflection. Many authors have described links between eating disorders and problems with attachment, loss, and separation (e.g., Armstrong and Roth 1989; Crowell et al. 1999; Dozier et al. 1999; Fonagy 1999; Hesse 1999; O’Kearney 1996). Others (e.g., Christenson et al. 1994; Grotstein 1991; McElroy et al. 1995; McDougall 1989; Schlosser et al. 1994; Schore 1994; Taylor et al. 1999) have explored links between eating disorders and difficulties in managing affects. Fonagy et al. (2002) point out that the capacity to regulate affect is directly tied to attachment experiences, an idea that has been underscored by contemporary research in neuroscience (Damasio 1999; Schore 1994, 2002; Siegel 1999). Demos (1993) notes that difficulties with attachment and affect regulation make it hard for clients to be self-reflective or use insight productively. This article reflects my attempts to integrate these ideas when working with clients who reveal an eating disorder after the therapeutic work has already begun.

Although each of these concepts has received significant attention in recent years, they remain complex and sometimes confusing. I will therefore specifically describe “eating disorders” as used in this article. I will also summarize some pertinent ideas about attachment theory and affect regulation and their pertinence to therapists attempting to understand and work with the complicated psychodynamics of eating disorders and the need to conceal them. An example from Alice’s therapy will illustrate the silent work of “test passing” (Weiss 1990) that helps a client feel secure enough to reveal an eating disorder. I will also describe how “detailed inquiry” (Sullivan 1954) can promote a sense of safety and help build initial skills for affect management and self-reflection that make such a revelation possible. As characteristic attachment styles, including resistance to attachments (inside and out of therapy) unfold, detailed inquiry helps clients bring these and other dynamics safely into the therapeutic dialog.

What is an Eating Disorder?

The spectrum of behaviors that come under the heading of eating disorders is surprisingly wide and complicated. Unfortunately neither the DSM IV (1994) nor the media do justice to the variety of symptom pictures or the complexity and diversity of the psychodynamics underlying these behaviors. Eating disorder symptoms can appear in “normal,” relatively short-lived adolescent experimentation with dieting, binging, purging, and/or overexercising. They can manifest as long-term, chronic and dangerous restriction, compulsive and/or chronic overeating with or without purging, and/or excessive amounts of physical activity. Binges can range from ingesting huge amounts of food to eating a half a muffin. Chewing food and spitting it out without swallowing it, going on severely restrictive diets, and undertaking punitive exercise regimens are also symptoms. Purging behaviors include compensatory and compulsive use of laxatives, vomiting, diuretics, and overexercise.

There can be significant weight gain or loss in a relatively short period of time, ongoing “yo-yo-ing” patterns––or maintenance of a fairly stable weight anywhere on the continuum from obese to thin. While it is tempting to look for a common etiology at least in each category of eating disorders, doing so often closes off some of the individual meanings of the behavior and can interfere with some of the most significant therapeutic work. Even though issues of attachment and affect regulation regularly emerge in therapy with individuals with eating disorders, it would be a mistake to assume that all eating disorders are caused by early loss or trauma. In 30 years of doing this work, I have seen no family “type” or historical constellation that either predicts or explains all eating disorders. While it is not uncommon for a client with such a disorder to have family members who struggle with these and other addictive behaviors and have difficulties processing and managing emotions, even this is not always the case.

Eating disorders can serve concurrent adaptive and maladaptive purposes. Clients who hide symptoms from their therapists also frequently hide them from other important people in their lives. For example, a woman with anorexia may deny the severity of her restricting food intake and become enraged when family members worry about her, yet also long for someone to put a stop to the behavior. A teenaged boy who purges in order to make a lightweight category in his sport may lie to his parents to protect the coach who is encouraging him to keep his weight down. At the same time he may secretly wish that his parents would forbid him to continue to participate in the activity and he may unconsciously find a way to “sabotage” himself, for example by failing to keep his grades high enough to allow him to remain on the team.

Affect Regulation

Although obviously problematic behaviors, one way eating disorders are adaptive is their role in soothing intolerable and unmanageable affects (Barth 1994, 1998, 2001; Christenson et al. 1994; McElroy et al. 1995; Schlosser et al. 1994; Schore 1994; Siegel 1999; Taylor et al. 1999. Despite discomfort and self-criticism, many people describe a sense of calmness and well-being after a binge, binge/purge cycle, or period of excessive exercise, or even during a period of severe restriction of food intake. However, few have a sense of agency in these moments even when they feel they have made a conscious decision to engage in the behaviors. Impulses to binge, with or without purging, often feel as though they come from outside oneself, and therefore neither a matter of choice nor within one’s control. The restrictive intake and overexercising of an anorexic, often believed to be acts of “willpower,” take on a life of their own as they become compulsive mechanisms. It is often hard to grasp that despite not being under an individual’s control, these symptoms serve an important psychological and emotional purpose. As Demos (1993) notes, activities, such as eating disorders help us tolerate our feelings, which are crucial to survival. When therapists help clients recognize not only the painful, but also the adaptive components of the behaviors, they are helping clients develop a more integrated sense of self and thereby experience agency in relation to both the symptoms and their feelings.

Many people who struggle with eating disorders are extremely bright and articulate and speak freely of their feelings, so it is often not obvious that their words are not helping them process emotions. They may accurately describe feelings and explain both the current context and historical explanations for them, but they may suffer from what has been called alexithymia (Barth 1994, 2001; Krueger 2001; Krystal 1988; McDougall 1989; Schore 1994; Taylor et al. 1999), or an inability to use these thoughts or words to process emotions. The paradox of intelligent, insightful people who cannot use language manage their feelings can be confusing not only for their therapists, but for families, friends, colleagues, and clients themselves. A relationship with a therapist who is interested in finding ways to talk and think about feelings can stir both hope and anxiety in clients with these symptoms. Whether completely concealed as in Alice’s case, or veiled (e.g., an anorexic who says she is consuming more food but continues to lose weight, or a compulsive overeater who underestimates his actual food intake), a therapist’s ability to understand and communicate the role of the symptoms in managing intolerable or overwhelming feelings can help make the process feel less dangerous. Tosone (2006) has described similar relational dynamics with regards to lying, another form of secret-keeping.

A gradual process of naming, talking about and accepting tolerable feelings leads to a sense of feeling understood (Fonagy et al. 2002) and contributes to attachment security. Damasio (1999), Siegel (1999), and Schore (1994, 2002) describe findings that talk therapy, which brings an opportunity for a new attachment experience and growing self-reflection, builds new neural pathways that open up the possibility for new relationships and new ways of managing and processing feelings. This process has often been quietly happening in a therapy in which a previously hidden eating disorder is suddenly revealed. When the symptoms join the therapeutic conversation, it is important that a therapist maintain the sense of safety that has been developing. A discussion about the details of a client’s eating behavior can be useful if a client can tolerate it, but interpretations of symbolic meaning (e.g., a woman with bulimia who seems to be both consuming and rejecting her mother) can be counterproductive. Even linking the behavior to a specific feeling can feel like criticism and lead to more shame (“now I should change, because I know why I’m like this,” as one client put it) and a variety of other “indigestible” feelings. More useful can be a therapist’s straightforward and repeated recognition of the ways that the behaviors are both addictive and serving a necessary purpose. Simple cognitive techniques for identifying and managing feelings, without expecting the symptoms to disappear or even diminish, are helpful, as is a therapist’s curiosity about the small details of an individual’s life (Sullivan 1954), which I will describe more fully in the clinical section. As individuals manage a wider range of feelings and become more self-reflective, their eating disorders gradually become less necessary to their survival and emotional well-being. However, as we will see, this process does not occur quickly, easily or without conflict and anxiety.

Attachment Theory

Numerous authors (Bowlby 1981; Fonagy et al. 2002; Eagle and Wolitzky 2008; Holmes 1996; Mitchell 1999; Siebold 2002) have noted that attachment theory and psychoanalysis, once considered incompatible theories, have been integrated in recent years. As Mitchell wrote, the practice of “psychoanalysis has always been centrally concerned with human relatedness” (p. 85), even though most early psychoanalytic and psychodynamic theorists did not put attachment needs in the foreground. Today, Bowlby’s once controversial idea that secure attachment is crucial to healthy human development seems almost self-evident. However, attachment theory has taken several different paths in recent years. As Holmes (1996) wrote, “Attachment theory … is not so much a single theory as an overall framework for thinking about relationships, or more accurately, about those aspects of relationships that are shaped by threat and the need for security” (p. 3). A variety of theorists over the years have considered the significance of connection and separation (e.g., Benjamin 1988; Blass and Blatt 1992; Mahler et al. 1975), the desire for relationship and defenses against it (e.g., Fairbairn 1946), and the role of others in both affect regulation (e.g., Demos 1988, Tomkins 1981, Winnicott 1958) and self-development (e.g., Kohut 1977; Stolorow and Atwood 1992). While many attachment theorists focus on issues of early loss and separation (Lyons-Ruth 1991), some also look at the normal use of attachment in the development of coping skills, including the capacity to be self-reflective, to have empathy for others, and to regulate affects (Fonagy et al. 2002; Reeves, personal communication, May 12, 2007; Schore 1994, 2002; Siegel 1999). It is this broader aspect of attachment theory, as well as its recognition that connection far more than separation promotes healthy emotional development (Lyons-Ruth), that I find useful with clients who have eating disorders.

Reinforcing Bowlby’s (1960) definition with evidence from neuroscience research, Siegel (1999) describes attachment as “an inborn system in the brain that evolves in ways that influence and organize motivational, emotional, and memory processes with respect to significant caregiving figures” and adds that “attachment establishes an interpersonal relationship that helps the immature brain use the mature functions of the parent’s brain to organize its own processes” (p. 67). Thus, it is a crucial component of emotional and cognitive development. Fonagy et al. (2002) link attachment and affect regulation:

None of us is born with the capacity to regulate our own emotional reactions. A dyadic regulatory system evolves where the infant’s signals of moment-to-moment changes in his state are understood and responded to by the caregiver, thereby achieving their regulation. The infant learns that arousal in the presence of the caregiver will not lead to disorganization beyond his coping capabilities … In states of uncontrollable arousal, the infant will come to seek physical proximity to the caregiver in the hope of soothing and the recovery of homeostasis. (p. 37)

Because many individuals with eating disorders have difficulties with separation (Armstrong and Roth 1989; O’Kearney 1996), it is tempting to assume a history of early loss and/or traumatic separation in all clients with these symptoms. Yet Dozier et al. (1999) state that the “evidence linking attachment behavioral strategies in infancy with adult psychopathology is limited to two studies” (p. 498). In her research, Demos (personal communication, May 4, 2007) has found repeatedly that not only are adult memories often inaccurate, but perhaps more surprisingly that clinicians who are asked to match material describing a group of adults with information about the same individuals as children frequently cannot match older people with their childhood stories. Furthermore, as most clinicians have discovered at some time or another, a client’s historical narrative can be unwittingly organized for self-protection rather than accuracy (Holmes 1999). Thus, Safran and Muran (2000) note, “Formulations based upon the patient’s narratives about past or present events are at best crude hypotheses” (p. 86).

Assuming specific categorizations of attachment style based on eating symptoms can limit exploration of an individual’s actual experience and dynamics. Studies from which conclusions about categories have been drawn are small and the conclusions sometimes flawed. For example, Hesse (1999) cites a small study of 36 individuals suffering from eating disorders. “Few were secure, and … anorexia was associated with the dismissing category, while bulimic subjects tended to be preoccupied” (p. 413). Dozier et al. (1999) cite a study in which the majority of individuals suffering from eating disorders were viewed as preoccupied, but could have been considered as dismissing if those who were depressed were removed from the data (as they apparently were in the study cited by Hesse). Yet given the evidence cited earlier that eating disorders are directly tied to affect and mood disorders, it seems illogical and arbitrary to separate out those who are depressed from so-called “pure” eating disorders.

Slade (1999) writes, “It simply does not make sense to think of patients in terms of a single, mutually exclusive attachment classifications [sic.] that presumably remain stable within the clinical situation” (pp. 584–585). Rather than focus on specific attachment styles or assumptions about history, when a client reveals an eating disorder it is more helpful to think about the specifics of an individual’s current attachment behavior or “attachment state of mind” (Dozier et al. 1999). Safran and Muran’s (2000) suggestion that therapeutic impasses offer useful data for understanding a “patient’s characteristic style of construing and acting as it unfolds in the present” (p. 86), Kohut’s (1971, 1977) ideas about “experience near” interpretations and Sullivan’s (1954) concept of detailed inquiry into a specific current experience are useful ways of thinking about and responding to the material.

Creating a “Safe Haven”

Over the past quarter of a century, many psychoanalytically oriented psychotherapists have come to believe that much of the work is done in a relationship between therapist and client. Individuals who hide an eating disorder, even those who make a positive connection early in the work, often need to do this work silently. Fear of being disappointed, abandoned, ashamed, hurt and angry as a therapist gets to know more about an individual’s symptoms leads some clients to hide them until a therapist has passed a number of (often silent) “tests” (Weiss 1990). Many therapeutic relationships include what Kohut (1971) calls a “silent idealizing transference” in which a therapist’s powers and knowledge are idealized and unquestioned. Such a transference can be damaged if discussed directly (for an example, see Barth 1988). Eagle and Wolitzky (2008) similarly suggest that before self-exploration is possible a therapist must sometimes serve as a “safe haven” to which a client turns for help managing difficult feelings.

When a client does reveal symptoms to a therapist, it can be a signal that therapy now feels like a place in which it is safe for clients to bring parts of themselves that they generally keep outside of their relationships. Far from being negative, these revelations often reflect that positive work has been occurring in therapy. It is crucial at this point that a therapist respect this show of trust by understanding that revelation does not mean that a client is ready to discontinue or even work on these previously hidden behaviors. As I have noted, for many clients food and eating behaviors function to soothe and restabilize in the face of disruptive emotions. Even clients who come to therapy specifically for help with eating symptoms may, either consciously or unconsciously, fear that they will be in emotional danger if the symptoms are removed. Numerous clients have told me, long after their symptoms have diminished, that as much as they longed to get rid of their painful behaviors, they also feared life without them. Thus even after symptoms are revealed therapists continue to have to pass more silent “tests,” including understanding that the behaviors help clients maintain psychic equilibrium and cannot be brought to a close until other mechanisms are in place.

As I have described elsewhere (Barth 1998, 2001) and will illustrate in the clinical section of this article, a detailed, nonjudgmental inquiry (Sullivan 1954) into the small, apparently insignificant aspects of a client’s daily life along with a genuine acceptance of a client’s anxiety about trusting the therapist can be an important part of the development of this sense of safe haven and a step towards a client’s gradual ability to tolerate and manage a range of feelings. Through this interest a therapist communicates interest in and readiness to accept a client’s experience no matter how difficult, conflicted or confusing it may be. By showing that he or she can tolerate and help a client learn to think about and gradually cope with “small” feelings about apparently insignificant moments, a therapist provides a frame for exploring and handling larger issues that arise both in and outside of the therapeutic relationship. Within this frame, and over a long period of time, an individual who has hidden an eating disorder can gradually develop what Demos (1993) calls “the capacity for self-analysis,” which

Requires the capacity to generate and sustain interest in one’s own inner experiences and to experience excitement and enjoyment in the process of exploring and understanding oneself. This capacity … does not just happen…It must be nurtured and fostered by the caregiver and by the analyst. (Demos 1993, p. 19)

Understanding that symptoms are adaptive (e.g., both self-soothing and sometimes a way of maintaining attachments to and individuating from needed and loved, if problematic, objects) is important in this early work. Educational, cognitive, and behavioral techniques can help ease self-criticism and are often crucial to the development of the capacity for self-reflection. Empathic yet firm limit-setting—for example, insisting that a starving anorexic be under the supervision of a physician and follow the physician’s recommendations while also recognizing and understanding her fears of being forced to gain weight—can demonstrate that a therapist can tolerate and help a client manage emotions, such as fear, anger, and distrust. While they will not deter a client from an eating behavior, joint explorations of data about the effect of eating disorders on the body can be useful as part of the ongoing process once the symptoms have been brought into the conversation. (e.g., the online site for Multi Eating Disorder Association, Inc.) A therapist’s nonjudgmental curiosity about a client’s actions and choices works as both a model for and an entry into self-understanding and self-regulation. Material that has not been repressed or warded off, but also has not been processed because it did not feel manageable or safe, can now gradually and quietly emerge in the therapeutic work. Although it is outside the scope of this article, this process may also occur with other groups of symptoms that help manage affects. As anyone who has tried to stop smoking knows, it can be extremely difficult to give up the behavior even with insight and the knowledge that the behavior is harmful to oneself and to anyone nearby unless other affect-management techniques are put in place. Some clients with eating disorders need years before they can manage feelings without eating symptoms.

Attachment-Individuation

Kernberg (1995) points out that many individuals with eating disorders experience autonomy and mature dependency as mutually contradictory. In my experience this is often true of individuals who hide their symptoms early in therapy. One of the areas a client may unconsciously test is a therapist’s capacity to tolerate and modulate both separateness and connectedness. A client who reveals previously hidden symptoms is often already working on what Lyons-Ruth (1991) calls “attachment-individuation,” or the capacity to integrate dependency and autonomy. However, even at this point, difficulties maintaining a sense of connectedness with someone when they are also feeling conflict with that person lead many of these men and women to avoid any conflict with their therapists. Disclosure of eating behaviors can indicate a burgeoning hope that a therapist might be able not only to help a client manage conflicting and shameful feelings, but might also find ways to cope with their own conflicts when a client’s behavior arouses them.

In my work with Alice, this hope and fear emerged in relation to my ability to stay connected even when she withdrew from therapy. As she gradually began to trust in my ongoing presence in the face of a variety of conflicts and emotions, she began to integrate experiences of attachment and separateness, along with elements of Fairbairn’s (1946) mature dependency, Kohut’s (1971) mature self-object needs, and Schore’s (1994) recognition that the need for help with self-regulation occurs throughout life. Bowlby (1980), Kohut (1977), Schore (1994), and Stern (1985) have noted that both attachment and individuation are important components of healthy functioning. A therapist’s verbal affirmation of this idea is extremely useful for clients who feel shame about their dependency needs. Numerous authors have also suggested that rupture and repair is crucial to the therapeutic process (for a useful summary, see Safran and Muran 2000). However, as noted earlier, focus on the therapeutic relationship itself can be too frightening for many clients to manage in the early stages of the work (see Kohut 1977). A therapist’s willingness to listen to and credit a client’s point of view while not necessarily negating his or her own perspective can help this process.

Clinical Illustration

Alice started therapy with a debilitating sense of hopelessness after the break-up of a long-term relationship. Although she managed to go to work, afterwards she could do nothing but go home to bed. She spent weekends alone, sleeping or watching television. Medication helped ameliorate the severity of her depression, but her internal world was so depleted and her ability to be curious about her own experience so limited that it was often painful to even sit in the room with her. In the early days of our work, she was unable to provide much useful personal history. When she began to emerge from the depths of her depression and could give what seemed to be a relatively coherent and organized narrative of her life, her memories did not help her think about or understand her current experiences. Alice initially presented a history of a depressed, withholding mother and loving, supportive father whose death in her early teens had created a tremendous hole in her life. However, as is often true, her past became more complex and difficult to “pigeonhole” as therapy progressed.

Rather than focus on her history, which Alice experienced as “parent-blaming,” I asked her about seemingly insignificant minutiae of her daily life, something along the lines of Sullivan’s (1954) “detailed inquiry.” When Alice reported that she had “watched TV” for example, I asked for specifics about the programs she chose. She replied that she just watched what was on. I asked about programs she enjoyed and disliked. She resisted this line of inquiry initially, but gradually engaged in discussions about characters and their conflicts, plotlines, and even about the writers she liked and why. Like many clients, Alice did not see such minor aspects of her life as significant and could not imagine that I would be interested in them. However, as I continued to ask about minute facets of her daily experience, she began to trust not only that I was genuinely curious, but also that the material was meaningful. She began to bring in bits and pieces about her family, her job, and her coworkers, and we found some common themes, particularly in situations in which she felt uncomfortable, conflicted, and/or depressed.

Over time we began to formulate organizing themes or narratives along the lines described by Holmes (1999), Lichtenberg et al. (1992), and Stern (1985). Topics that help give order to seemingly disparate and unconnected experiences offer a sense of internal cohesion and consistency for individuals whose internal world feels chaotic and confusing. Helping a client begin to “quiet all the noise inside,” as Alice put it years later, is one of the functions of a “safe haven” that will gradually become a secure attachment. With some skepticism, Alice answered my questions about things, such as what she did in the morning after she woke up, how she decided what to wear, and when she made her coffee, brushed her teeth, and took her shower. I also asked about her nighttime rituals. I have been surprised over the years by the number of clients (not only those with eating disorders) who fall asleep on the couch while watching television, still wearing their work clothes and contact lenses. Discussing these bedtime behaviors has led not only to eventual shifts in these routines, but also to increased curiosity about other automatic, often unregistered moments. My curiosity frequently expands their capacity for self-awareness, both in relation to behaviors known by the individual but hidden from the outside world, and to internal experiences that have never been put into words.

I often try to share some observations about what I am hearing. For example, early in the work Alice reported moments of overwhelming shame about herself followed by withdrawal. When I named this behavior and suggested that it might be a way she tried to protect herself, she was taken by surprise as though she did not recognize these qualities as her own. Material that is not formulated may seem self-evident to a therapist yet invisible to a client. Because shame about such aspects of self is common, a therapist needs to be tactful in making observations about it. However, even if not quite on target from the client’s point of view, statements about what a therapist sees can be confirmation the therapist is paying attention and trying to understand. Kohut (1977) believes that an analyst’s genuine attempt to understand an analysand is one of the most important components of psychoanalysis. From the perspective of attachment and affect regulation, there is a parallel to a parent’s effort to understand a crying baby. Is the baby hungry? Does the baby have a dirty diaper or want to be held? Even if a caregiver has not correctly diagnosed the problem, genuine efforts to soothe the infant are significant, and the eventual ability to take care of the problem (an experience repeated over time) leads to a sense of agency on the baby’s part (I can get my needs met) and to an internalized view of others as attuned and responsive. One of the valuable contributions of attachment theory is recognition that part of the therapeutic work in such cases is to provide both consistency in the relationship and consistently available attempts to help a client regulate affect over a long period of time.

Alice feared and therefore resisted this experience as much as she longed for it. As she described the details of her life, including interactions with her parents, siblings, friends, and colleagues, she showed evidence of what would be called from an attachment theory perspective a preoccupied (or ambivalent/resistant) attachment style. She had difficulty soothing herself, feared rejection, and compensated with compulsive caregiving and self-destructive self-soothing. She was also smart, thoughtful, and highly capable. Often she found herself in predicaments that she could not handle because she sounded and appeared far more sophisticated and self-confident than she was. She found it helpful when I framed her difficulties in terms of boundaries, and later when I told her about the concept of object constancy, or as she preferred to call it, “object permanence.” For example, she described not only upsetting phone conversations with her mother and sister and difficulties with her boss at work, but also a variety of “unimportant” dealings with the dry cleaner, superintendent of her building, and pharmacist who filled her prescriptions. Gradually we made links between these groups of experiences: Alice did not like to ask for anything, did not like to “put anyone out,” yet she had tremendous difficulty refusing anyone else’s requests no matter how much they inconvenienced her.

In general she seemed to have an open door policy, not only as to her apartment, but with her possessions, her time, and also her body. She responded with an observation about her relationship with her ex-boyfriend. If he called, she inevitably invited him over; and if he wanted sex (as he always did), she agreed, even though she knew that to do so would stir up painful feelings of need and loss once again. Articulate as always, Alice said that she wanted to be with him so badly that she would do anything, even if she knew she would hurt later. I said that it seemed that the pain of not being with him at all seemed worse than the pain of losing him after having him for a short moment in time. “You understand that?” she asked. “Everyone else gets mad at me for not getting rid of him.” I said it seemed that she felt that she could not tolerate not having him in her life. Yet what, I mused in the silence that followed, was the pain involved in telling the dry cleaner that they needed to find the belt they had lost? Or in telling her sister that she had to get off the phone after an hour of listening to complaints? Although she had been crying while talking about her boyfriend, she smiled at these questions. “The pain is that my sister won’t love me anymore, and that the dry cleaner or the guy at the drugstore won’t like me. And then I’ll feel like a bad person. I’ll hate myself. I’ll feel like I’m bad and deserve to suffer.”

Alice thought she would lose not only the love of others if she maintained her boundaries, but also a positive sense of herself as a good person. She would do almost anything to avoid the feelings that emerged in these situations. Our work involved both recognizing this fear and helping her find ways to manage the feelings. Demos (1988) notes that negative affect is not only inevitable, but also crucial to human survival. The therapeutic work was not to stop all of her unpleasant feelings, but to help her learn to process them. She found it extremely useful when I shared Demos’ (1988) belief that distress could occur with positive feelings like excitement and joy as well. Alice feared the loss of the assistance she needed for managing these affects when she seemed not to need them (i.e., set boundaries), so she gave others what they wanted and dealt with the inevitable subsequent intolerable (and generally unformulated) emotions with secret bingeing and purging behavior. As we teased out these dynamics, Alice gradually began to speak of historical material in ways that helped her think about and process current experience. Throughout the work we talked about the early loss of her father, who died when she was in her teens; her mother’s subsequent depression and withdrawal from all of the children; and her feeling that she needed to take care of everyone in order not only to make sure that they survived, but to ensure her own survival. However in the early years, our musings about her history did little to help her modulate either the behavior or the feelings. Bowlby (1981) made the somewhat astonishing (at least to those therapists who consider finding and delineating historical trauma to be key to the work) comment:

What a patient tells us about his childhood … (is) probably influenced as much or more by the analyst’s preconceptions as by anything the patient may in fact have said or done … The research strength of the therapeutic situation lies not in what it tells us about the patient’s past but in what it tells us about disturbances of personality functioning in the present. (p. 250)

We utilized historical reconstruction to help formulate what Alice was feeling in the present (see Holmes 1999). Over time, as we found language for representing a wide range of Alice’s feelings including both her longing for and fear of close relationships, her recollections became emotionally richer and meaningful. Still, Alice often found it more helpful when I focused on the process of learning to build and accept boundaries than when I tried to help her name or understand her feelings. Some of these boundary issues were played out in the therapeutic relationship, but while they were certainly important, they did not seem more significant than those that happened outside of therapy. Alice’s sense of safety and well-being grew as she became confident she would be understood as she tried to talk about these experiences in therapy. For example, as we recognized that she was conscientiously refusing to set limits despite our discussions of her need for boundaries, we also began to talk about how much she disliked it when other people set limits with her. Over the course of numerous discussions, we teased out her sense that such restrictions, whether set by her or the other person, impeded her access to those she needed for psychic sustenance. We also began to see that when people “kept her out,” she interpreted it as meaning they did not like her. Having developed a language for talking about these issues, we were able to examine them as they developed in her relationships both inside and outside of the therapy. An ongoing opportunity to explore these experiences was Alice’s habit of missing appointments without notifying me.

Because of her depression and her difficulty trusting that her connection to others remained when she was not in contact with them, I called her when she missed a session. If I did not reach her, I would leave a message saying I was concerned and asking her to call to let me know if she was alright. She did not always call back and sometimes would miss the next session as well. However, eventually she would return for a session, shamefaced and worried about my potential irritation. Although I did not yet know about the eating disorder, Alice’s behavior was consistent with many of the young women I had seen over the years with these symptoms. I had come to understand that ambivalence about and fear of the very connection they longed for was part of the therapeutic construct and its enactment was part of the therapeutic work (again along the lines of Weiss’s “test passing”). As part of the provision of a setting in which her feelings and needs were recognized and respected, I made it clear that it was her prerogative to miss sessions when she did not feel like coming in. However, I also indicated that I expected to be paid for those sessions, since I had put aside the time for her. Alice was compliant at first, only later admitting that it bothered her that I was charging her for time she was not using. We explored her feeling that I was being punitive, her fear that I was angry at her, and much later, her relief and recognition that if I took care of my own needs I could be more empathic to hers. Variations on this theme of self-care for both of us led to a growing and ever-richer understanding that some boundaries actually enhanced connection.

It was after one such period of missed sessions that Alice reported her eating disorder to me. As she put it later, “I guess I could admit something else bad about myself to you, because you’d gotten the missed appointments so clearly and hadn’t been mad at me or critical of me for it. And because you didn’t let me cross your boundaries––even though I didn’t like it, it was a relief that you charged me for those sessions. Somehow that made it easier for me to live with myself.” In retrospect, I also believed that she could bring the eating behavior into the therapy because she trusted me to help her manage both the intolerable affects that they helped her endure and the painful feelings related to revealing and talking about this aspect of herself. Thus, her acknowledgment of her eating disorder was part of the progression of the therapeutic work, although her eating disorder remained active for several years after this. In fact, after she told me about them, they got significantly worse for a period of time. I believe she was testing to see if I genuinely could continue to accept her “worst” parts and help her manage the feelings that emerged as she shared these qualities with me. The increased bulimia was both a defiant “this is my worst; do what you will!” and a way of soothing herself in the face of my potential rejection.

Gradually Alice learned to tolerate her feelings without binging and purging and discovered that she could expose more of the complex aspects of herself to other people without being rejected by them. Today she still uses food and exercise to cope with painful emotions and difficult situations; but she is no longer bulimic. Her depression, which recurs at times, is more easily managed. Not only does she accept the need for medication at these times without feeling that she is fatally flawed, but she also finds it useful to look at the emotions as cues that something is wrong and that she needs to understand what it is. Despite several personal tragedies that occurred during the course of our work together, her personal life and career have become reflective of her unique personality and are both deeply satisfying in many ways she could only long for when we began our work together.

Discussion

Although many therapists are distressed when clients with whom they have been working suddenly reveal that they have an eating disorder, such a revelation often indicates not a failure, but instead that therapy is actually succeeding. In these instances, a therapist has often successfully passed a number of silent tests (Weiss 1990) presented unconsciously by a client, and therapy has started to feel like a “safe haven” (Eagle and Wolitzky 2008) for that client.

Attachment theorists and neuropsychologists have cogently illustrated that capacities for both self-reflection and mentalization grow in the presence of an other who is both interested in one’s experience and able to reflect back an understanding of that experience. It is this aspect of attachment theory, as well as the recognition that the capacity for self-regulation and processing of affects develops within the context of a secure attachment that is often useful in working with individuals who reveal a hidden eating disorder after therapy has begun.

Despite being articulate and insightful, many clients with eating disorders have difficulties in the realms of “self-reflection” and “mentalization” (Fonagy et al. 2002) as well as affect regulation. An individual who hides an eating disorder from a therapist often hides it from other people as well, frequently in an attempt to maintain self-esteem and self-cohesion. However, hiding it often reinforces the very feelings that lead to the behaviors in the first place. When a client discloses the behavior to a therapist, it often means that he or she trusts that therapist will not judge the behavior or the client too harshly and that he or she trusts that the therapist can help manage some of the shameful and/or overwhelming feelings and characteristics associated with the behavior.

When a client makes such a revelation, it is extremely important that a therapist neither abandon them (e.g., because of feeling inadequate to deal with the symptoms) nor immediately begin to work to diminish the eating behavior. It is important for a therapist to take her or his cues from each individual as to whether or not the revelation brings with it a capacity to begin focusing on the symptoms. For some clients, it is enough that they have shared this hidden part of themselves, and they need time to live with the revelation. For others, it is important to offer some initial understanding of the symptoms and some specific advice for working on the behaviors. After an eating disorder is revealed, much therapeutic work can be done by focusing on specific details about the small and apparently insignificant aspects of a client’s daily life.

Therapists need to recognize adaptive and maladaptive aspects of the behavior. Respecting boundaries, setting limits, and providing reasonable flexibility are also part of the therapeutic work. Although much contemporary thinking about psychotherapy focuses on therapeutic relationships, clients who have eating disorders (openly revealed or hidden from their therapists) are frequently not able to explore the relationship until many years into the work. Because many clients with eating disorders are verbal and intelligent yet suffer from “alexithymia” and cannot use words to process their feelings (see Barth 1994, 2001; Krueger 2001; Krystal 1988; McDougall 1989; Schore 1994; Taylor, et al. 1999), they can spend many years in therapy, developing a strong positive transference to their therapists and learning about themselves cognitively, without developing a capacity for self-regulation, empathy, intimacy or self-regulation––components of what Fonagy etal. (2002) has called “mentalization.” A therapist’s interest in the small details of a client’s daily experience, along the lines of Sullivan’s (1954) “detailed inquiry,” provides a “safe haven” (Eagle and Wolitzky 2008), a place in which affect-regulation and self-exploration can gradually occur. Genuine interest in apparently unimportant details of the daily lives of clients, exploring meaning when and where possible, providing education (e.g., my teaching Alice about “object constancy” and what it meant), setting boundaries and maintaining a sensitive stance vis-à-vis a client’s feelings (e.g., no feeling is “wrong”) is not only what helps a client who has hidden an eating disorder to feel safe enough to bring it out into the open, but also what makes it possible for that client to move forward in therapy once the secret is out.

A final note: while attachment theory enhances our understanding of the dynamics of eating disorders and the need to hide them from a therapist, it is important not to assume that all individuals with eating disorders have the same attachment history or difficulties, or even that one individual has only one clearly defined attachment style. Therapists may feel more of a background presence than a subject of exploration for some time in these sorts of therapies. However, this does not mean that work is not going on. When a client reveals an eating disorder after a period of therapy, it is often a sign that a therapeutic process has occurred, even if the therapist is not aware of it having happened. Accepting and at times asking about resistance to attachment both in and outside of a therapeutic relationship while maintaining a respectful and consistent availability to the client makes possible new attachment experiences and a concomitant development of a capacity for self-analysis. Research in neuroscience (e.g., Damasio 1999; Schore 1994; Siegel 1999) confirms that talk therapy helps an individual develop areas of the brain linked to attachment, self-awareness, and self-regulation. While I do not agree with all of the ideas that fall under the heading of “attachment theory,” I have found it helpful in these cases to think of the therapy within the frame of a developing attachment.