Dependency includes universal personality traits, expressed in different ways to some degree over the life span. Like other mammals, humans start out being very dependent upon adults for care and protection, then evolve through maturation and learning into more self-regulating and autonomous individuals. In familial, interpersonal, and organizational settings, healthy expressions of dependency are characterized by adaptive interdependency, where individuals negotiate helping and being helped. In settings in which dependency and autonomy are either excessively encouraged, discouraged, ignored, or punished, dependency may become increasingly pronounced or pathological in its expression. When dependent behaviors are pervasive, frequent, and associated with impairment, an individual may be diagnosed with dependent personality disorder (DPD). Many individuals with DPD manage their lives by forming relationships with dominant spouses, friends, relatives, and bosses or coworkers, who in turn respond to dependency. A stable, if precarious, homeostasis in such relationships may allow the individual to function well to the outside observer. However, individuals with DPD may become symptomatic when dependent relationships are disturbed, threatened, or broken off (Perry and Vaillant, 1989), or when their own needs and feelings are increasingly ignored or punished, or failures at achievement occur (Zaretsky et al., 1997). In these instances, the individual with DPD may seek help, often precipitated by an Axis I disorder, or a painful life event. Dependency is commonly overlooked until the individual becomes symptomatic or overwhelmed with a life situation.
The psychotherapy of DPD can be quite successful, or quite lengthy and challenging depending on the patient, the therapist, their goals and alliance, as well as the technical approach employed. Although dependency issues are relevant in the treatment of many psychiatric disorders, this review is limited largely to those reports that are most relevant to the treatment of the PD.
Clinical interest in dependent personality traits began with Abraham's (1924, or reprinted in Abraham, 1954) description of the oral character. The PD first appeared in a War Department Technical Bulletin in 1945 (US War Department, 1945) and later in the first edition of the Diagnostic and statistical manual (DSM) (1952) as a subtype of passive-aggressive PD. Since then, a large number of studies have upheld the descriptive validity of dependent personality traits, viewed as submissiveness (Presley and Walton, 1973), oral character traits (Gottheil and Stone, 1968; Kline and Storey, 1977), oral dependence (Lazare et al., 1966, 1970; van den Berg and Helstone, 1975), or passive dependence (Tyrer and Alexander, 1979), or as a constellation of both pathological and adaptive traits under the term dependency (Hirschfeld et al., 1991; Bornstein, 1992, 1998). DSM-IV (American Psychiatric Association, 1994) emphasizes two sets of traits: dependency (criteria 1–5), and insecure attachment (criteria 6–8). In a study diagnosing PDs by both DSM-IV and ICD-10 (Ottosson et al., 2002), there was moderate agreement across the two systems in diagnosing DPD (kappa = 0.75), although ICD diagnosed almost 45% more cases, and excellent agreement between their dimensional scales (r = 0.94).
Dependent personality is common in the general population—the Midtown Manhattan Study found it in 2.5% of the entire sample (Langer and Michael, 1963), while a recent Norwegian survey found it in 1.5%, with the prevalence in women twice that in men (Torgersen et al., 2001). In clinical settings, DPD often co-occurs with other PDs, especially borderline, histrionic, and avoidant types (Hirschfeld et al., 1991; Bornstein, 1995a; Zanarini et al., 1998) and, although less frequently studied, with self-defeating, passive-aggressive, compulsive, schizotypal, and paranoid types (Bornstein, 1995a; Reich, 1996; Skodol et al., 1996), and, in the author's own research, depressive PD. Treatment should be modified accordingly. Patterns of Axis I and II comorbidity vary widely depending on sample source, reason for selection (e.g., major depression), and assessment method.
Managing the dependency that often accompanies chronic major psychiatric syndromes such as schizophrenia or unremitting depression (Bornstein, 1992; Kool et al., 2003) may have similarities with treating DPD. However, noting that Axis I disorders such as depression often increase dependency, Skodol et al. (1996) suggest that maladaptive dependency might become the focus of treatment in its own right, if it does not improve after the symptomatic disorder improves.
A factor-analytic study suggested that dependency is best characterized by three related dimensions (Hirschfeld et al., 1977). The first involves strong emotional reliance on close attachments and others. Livesley et al. (1990) labeled this dimension insecure attachment, after Bowlby's description. However, Borenstein (1997) has argued that insecure attachment is not a core aspect of DPD. Individuals with this dimension of dependency are prone to separation anxiety and will remain in relationships, even with those who mistreat them, to avoid the resurgence of feeling alone and helpless. They may act in ingratiating ways, doing whatever is asked of them in order to be liked or to secure succor. Whenever hospitalized, these individuals may transfer their attachment needs to the hospital. Prior to discharge, separation anxiety re-emerges and their presenting symptoms may recur, possibly delaying discharge (Sarwrer-Foner and Kealey, 1981). This is less likely to occur whenever the patient has an already established, good, supportive relationship outside the hospital.
A second dependent dimension is the lack of self-confidence in social situations, often accompanied by submissive behavior (Hirschfeld et al., 1977), which Livesley et al. (1990) considered the core dependency dimension. This includes having difficulty asserting oneself, agreeing with others despite believing that others are incorrect, and fearing self-expression of one's own anger, criticism, or wishes and needs. The individual may remain passive when events call for an active response. Despite this, the individual may be surprisingly able to confront anxiety-provoking situations courageously to help or protect those dependent on him or herself. In other situations, dependent individuals can be quite assertive, even aggressive, whenever striving to obtain or maintain a supportive relationship (Bornstein, 1995b).
The third dimension is the avoidance of (versus desire for) autonomy (Hirschfeld et al., 1977). Those who avoid autonomy want others to make decisions for them; otherwise, they are indecisive and have difficulty initiating or completing activities on their own. They often seek guidance and direction and thereby subordinate their freedom of choice to the will of others. The extreme opposite is often called counterdependency, in which individuals strive to be independent at all points. When ill or under extreme stress, counterdependent individuals may revert to very dependent behaviors, often accompanied by an intense sense of shame.
Dependency is moderately stable from childhood onward. Kagan and Moss (1960) found a high correlation between passive and dependent behaviors at 6–10 years of age and their continuation into young adulthood. While this was across a broader range of behaviors for women than men, cultural influences may discourage certain dependent behaviors in men. Twin studies indicate that some of the stability is due to genetic influences while others are due to specific environmental differences unique to each child (O'Neill and Kendler, 1998).
Dependent individuals experience excessive self-doubt and view themselves as incompetent and less worthy or deserving than others. They may appear overtly optimistic (Kline and Storey, 1977), but have a covert pessimistic view of their chances for self-initiated social and occupational achievement. They may be prone to ruminate on their fearful attitudes and phobic anxieties about self-assertion, social activities, independence, and abandonment (A. T. Beck et al., 1990).
A. T. Beck et al. (1990) proposed a cognitive conceptualization of DPD suggesting that the individual believes two key assumptions. First, the individual believes him or herself to be inadequate and helpless and the world to be cold, lonely, and dangerous. Second, he or she assumes that the best strategy is to find someone who is capable of dealing with the world and protecting him or her. Submissiveness and relinquishing independent decision making are considered acceptable tradeoffs for maintaining such protective relationships. Furthering this, Judith Beck (1997) suggested that the core belief in DPD about the self is that ‘I am helpless’ while the core belief about others is that ‘others should take care of me’. Life situations that stimulate these core beliefs then trigger assumptions such as ‘if I rely on myself I'll fail’ and ‘if I depend on others I'll survive’. These assumptions then lead to behavioral strategies of relying on others.
Some evidence supporting this was demonstrated by strong associations between DPD and specific dependent beliefs on the Personality Belief Questionnaire (A. T. Beck et al., 2001).
Whitman et al. (1954) suggested that dependent individuals may become passive whenever dependent needs are stimulated if the person finds these needs unacceptable in the situation, due either to a neurotic sense of guilt or to external frustration. As a secondary effect of frustration, the individual may become demanding in minor ways. Millon (1981) suggested that oversolicitous, controlling parents who discourage seeking rewards outside the family may discourage independence. Because dependent individuals have had a relatively good relationship with at least one parent, anxiety experienced in situations requiring independent action is counterbalanced by the expectation that someone will help. The expectation of criticism for making independent decisions, taking action, or venturing to new activities further stifles independence. Instead of channeling hostile feelings into assertive behavior, dependent individuals often smooth over troubles by acting in an especially friendly, helpful, and concerned manner. In a study of the family environment, Head et al. (1991) found some support for Millon's hypotheses in that the individuals with DPD reported that their families were low in expressiveness and high in control. Baker et al. (1996) found that DPD individuals reported early family environments that were lower in encouraging independence and higher in control over the subject than normal controls, while being lower in achievement and intellectual-cultural orientation than the environments of individuals with histrionic PD.
In a study of addicts living with their families of origin, Alexander and Dibb (1977) found that compared with control subjects, both the addicts and their parents perceived the addicts as passive, dependent, and incapable of autonomy and success. Neither the addict nor the overindulgent parent encouraged self-reliance.
Perry et al. (1989) and Waska (1997) noted that dependent individuals often act in a submissive, compliant way in order to earn others’ gratitude. This ingratiating behavior entitles them in fantasy to maintain their important attachments and protects them from abandonment and the development of separation anxiety, and entitles them to being soothed and taken care of (Waska, 1997). Despite this, they can be quite aggressive toward others when they think doing so will ingratiate themselves with authority figures or secure care or help (Bornstein, 1995b), or protect those under their care.
Epstein (1980) compared the social consequences of assertive, aggressive, passive-aggressive, and submissive behaviors. Submissive behavior (e.g., making a request accompanied by an indication that one will capitulate easily), consistently elicited high intentions to comply, low anger, and high sympathy from observers, generally equal to the levels obtained by assertive behavior. Thus, submissive behavior may meet with some success, depending on the responsiveness of others.
In a study of passive adolescents, Rosenheim and Gaoni (1977) postulated that a fear of having to mourn childhood fantasies about the future may result in a failure to make decisions, enter into personal commitments, and take independent action. Refusal to take an active stance in working toward any plan avoids having to set aside cherished, if overvalued or unrealistic, hopes for the future and avoids the sadness of mourning.
Andrews et al. (1978) suggested a biological hypothesis for dependency based on finding high levels of anxiety-proneness, emotionality, and easy fatigability in individuals with asthenic personality. This may encompass a constitutional predisposition to develop high anxiety levels under stress, often called neuroticism, which in turns disrupts learning.
The treatment literature is limited largely to case descriptions, uncontrolled studies, and some controlled treatment trials with admixtures of PDs, including DPD. Across all of these, there is an apparent consensus that the treatment of DPD is often successful. This is indirectly supported by the relative lack of articles that report failures or focus on difficulties in treatment, in contrast to the plethora of such reports for other PDs.
Systematic studies including DPD indicate that treatment on average leads to improvement. Virtually all studies indicate that psychotherapy produces sizable, positive effects in PDs (Perry, 1989; Perry and Bond, 2000; Leichsenring and Leibing, 2003). This is true for both dynamic and cognitive-behavioral therapy approaches (Leichsenring and Leibing, 2003). Two related meta-analyses found a number of studies with a median of 25% (range 10–33%) of individuals with DPD, and others with large proportions of unspecified Cluster C disorders (Perry and Bond, 2000). The therapies demonstrated medium to large positive effects (generally larger than 1.0) for individual (Winston et al., 1991, 1994; Hoglend, 1993; Hardy et al., 1995; Monsen et al., 1995; Patience et al., 1995), group (Budman et al., 1996), day (Karterud et al., 1992, 2003; Piper et al., 1993; Wilberg et al., 1998), and residential treatments (Dolan et al., 1997; Krawitz, 1997). This empirically supports the earlier conclusion reached by one expert panel that the treatment of DPD is generally successful (The Quality Assurance Project, 1991).
In the Treatment of Depression Collaborative Research Program, three active treatments (imipramine, interpersonal psychotherapy, and cognitive-behavior therapy) were compared with placebo plus clinical management over 16 weeks for the treatment of acute major depressive disorder (Shea et al., 1990). Significant improvement in depressive symptoms occurred in those both with and without PDs, although there were no clear-cut differential responses to type of treatment. However, complete remission was found in fewer of the anxious cluster (33%) than of those without a PD (49%). The former also showed worse social adjustment. Thus, patients with DPD who become depressed may respond to treatment, but upon return to baseline some symptoms and social adjustment problems remain, a conclusion confirmed by other studies (Diguer et al., 1993; Hardy et al., 1995; Patience et al., 1995).
Kool et al. (2003) conducted a randomized trial of patients with DSM-III-R major depression, half of whom also had a PD by a self-report method. Among those scoring in the PD range, DPD was the most prevalent (38%). Patients received either antidepressive medications alone or combined with 16 sessions of Short Psychodynamic Supportive Psychotherapy (SPSS) and the percentage recovered from depression at 40 weeks was similar (44% versus 51%). In both conditions, those who recovered showed improved personality traits. However, those receiving combined treatment showed such improvement even if not recovered from their depression. In fact, dependent and several other traits (e.g., avoidant, passive-aggressive) improved more with combined treatment than pharmacotherapy alone, most evident in Cluster C disorders. This strongly supports the additive effect of psychotherapy in combined treatment on dependent traits among the depressed.
Attrition from treatment may be lower in DPD than other PDs. Shea et al. (1990) found that patients with anxious-cluster PDs, including DPD, had a lower attrition rate (28%) than other PD groups. Katerud et al. (2003) found only 16% of those with DPD dropped out of an 18-week day treatment program, lower than all PD types, except for schizoid.
There is no firm data on the optimum duration of treatment. In an examination of the natural history of open-ended dynamic psychotherapy for adults with mood and/or PDs, the author has found that DPD and significant dependent traits were associated with a median length of treatment almost twice that of those with no dependent traits of DPD (129 sessions versus 66 sessions), which was statistically significant. Karterud et al. (2003) examined 18 weeks of day treatment on a large sample of PDs treated at eight sites. In general PDs improved, but most did not attain healthy level scores. Among DPD specifically, Global Assessment of Functioning (GAF) improved at termination and continued to improve over 1 year follow-up (mean of 47, 52, 56). The Global Severity Index (GSI) of the Symptom Checklist-90 (revised), and a quality of life measure improved at termination and were maintained over follow-up, whereas a measure of interpersonal functioning improved but later regressed. Employment was not significantly improved over follow-up. Thus while improvement is the rule, the duration of treatment required for recovery to full, healthy functioning remains to be studied.
As cultural factors influence what is considered normal dependency, the therapist should consider the cultural context of the patient. Having family members accompany the patient into a consultation might signal dependency issues in a northern American or European family, whereas the same would be absolutely the norm on the Indian subcontinent. Similarly, depending on the culture, women and men may express dependency differently, with the female stereotype showing more insecure attachment and submissiveness and the male stereotype submissiveness and especially avoidance of autonomy. This results in a tendency to overlook dependent traits in men, which may include the need to talk about every decision and seek reassurance and encouragement, or the failure to take action or move forward in a career. Failure to recognize these as dependency issues may lead to perplexing stalemates in treatment. Regardless of the treatment approach, it is important to identify the patient's specific dependent patterns.
In the dynamic psychotherapy literature there is apparent consensus about two central aspects in the therapy of DPD. The first is that the emergence of a dependent transference toward the therapist should be addressed in a way to promote emotional growth. The second is that therapist expectations and direct support should be used to promote self-expression, assertiveness, decision making, and independence. If both aspects are not addressed, treatment may be incomplete (Hill, 1970; Saul and Warner, 1975; Malinow, 1981; A. T. Beck et al., 1990).
At the outset of therapy, it is important to aid the development of a trusting relationship and allow the patient to begin to transfer dependent wishes on to the therapist. Hill (1970) suggested telling the patient that extra sessions may be allowed early on in therapy, especially around the patient's episodes of panic or distress. This assurance of readily available support helps the patient develop trust, and aids alliance formation. As therapy progresses, the therapist may help the patient find substitute ways of dealing with such feelings and limit extra sessions.
Alexander et al. (1968) found that dependency on the therapist increased from the beginning to the middle of short-term therapy and remained fairly high until termination. The high levels of dependency on the therapist necessitated working through transference issues right up until termination. In contrast, they found that the patient's dependency on outside relationships began to diminish from the middle of treatment until termination, which suggests a real effect of treatment on the resolution of dependency conflicts.
The hardest work of therapy occurs when a patient experiences increased dependency on the therapist and simultaneously has setbacks or losses in his or her outside life. Offering sympathy for the patient's distress is not helpful alone (Hill, 1970). The therapist should also encourage the patient to express real feelings and wishes and to bear the anxiety of making decisions, accepting pleasurable experiences, and dealing with episodes of anxiety. When the patient experiences frustration over his or her wish to have the therapist take a more directive role, the therapist should clarify and interpret the transference elements in addition to supporting the patient in finding more self-reliant ways to cope (Hill, 1970; Saul and Warner, 1975; Malinow, 1981; A. T. Beck et al., 1990). Leeman et al. (1975) limit attention to transference issues in favor of focusing on relationships outside of therapy.
At this stage, the therapist should avoid taking a directive role in the patient's life; otherwise a transference–countertransference fixation might develop that simply repeats patterns from the patient's other relationships (Leeman et al., 1975; Saul and Warner, 1975). This requires actively resisting the patient's repetitive requests for advice and attempts to have the therapist make decisions for the patient, something the patient expects from authority figures.
Saul and Warner (1975) described the following optimal circumstances for the therapist to give direct suggestions and encourage specific actions or solutions to problems. First, the treatment should have progressed long enough for the therapist to have a good understanding of the patient's dynamics. Second, the therapist should be aware of the state of the transference and his or her own reaction to it. Third, the patient should be at some impasse out of which a direct therapeutic intervention can mobilize the patient and prevent a repetition of feeling powerless. Given these circumstances, the therapist should help the patient conceptualize his or her own goals. If the goals are healthy, the therapist should discuss and support them. If there are conflicting goals, then it is helpful to discuss the consequences of each goal and to encourage the patient to bear the anxiety of making choices. While similar to cognitive therapy (A. T. Beck et al., 1990), this approach also makes use of previous insights about the patient's motivations. The therapist may then urge the patient to commit himself or herself to actions that are within the patient's reach (e.g., taking a job) or encourage perseverance despite the urge to give up (e.g., flunking out of school). The therapist must also ensure that he or she is using his or her influence in accordance with the patient's own values, not those of the therapist.
Covert dependency on the therapist, in which the patient experiences the therapist as a benign, powerful parent figure (Goldman, 1956), can facilitate therapeutic change. The therapist's sincere interest, attention, and reliable presence may increase the patient's belief in the benevolent power of the therapist. This affects the patient's self-esteem in several ways. First, the patient may identify with the therapist and wish to be like him or her (Offenkrantz and Tobin, 1974). Idealization leads to a temporary rise in self-esteem. Second, the patient may accept and increasingly use the therapist's exploratory attitude toward his or her emotional life. Third, whenever the patient remembers or experiences hitherto unacceptable feelings for the first time, the therapist should be comprehending and accepting. This will enhance the patient's self-esteem, because the patient can identify with the more benevolent attitudes and responses of the therapist as an authority figure, rather than react according to his or her old prohibitions and ideals. This rise in self-esteem is only temporary as long as it relies largely on the reassuring presence of the therapist. However, if the patient can channel this increased self-esteem to risk trying new behaviors outside the office, he or she may experience other rewards, including approval from others. It is important for the therapist both to communicate genuine pleasure when these outside efforts succeed and to accept failures that inevitably occur. This helps the patient to shift self-perception from dependency toward social self-confidence.
Attending to the patient's defenses can inform the therapist about conflict areas requiring attention. In particular, reaction formation against feelings such as anger toward dominant others may be masked as concern. Similarly displacement may frequently divert attention away from the patient's problems to those of the people around him or her. When confronted with situations necessitating more autonomous and self-assertive functioning, the patient may lapse into help-rejecting complaining, preferring the safer experience of failure while also covertly criticizing others for their lack of care and material help. The therapist should help the patient explore the meaning of such experiences and return to the underlying feelings such as anger, disappointment, and shame. Understanding what makes such affects distressing can then lead to better tolerance of them and finally point towards more effective functioning.
During the final stage of therapy, the therapist gradually increases the level of expectations for autonomous decision making and action and for socially effective responses (Leeman and Mulvey, 1975). This includes reinforcing the individual's increasing ability to handle crises without extra sessions, to manage anxiety/panic episodes by self-soothing rather than by seeking reassurance from others (Hill, 1970). The therapist must help the patient to resolve transference wishes to be dependent and fears of aloneness, powerlessness, and others’ intolerance for self-initiated expression and action, while accepting instead a more self-reliant position in relationships. Prior to termination, if the patient avoids mourning the therapeutic relationship, for instance, by fantasies that he or she was never really close to the therapist, or that the therapist will always be available, then termination will provide a crisis. The patient may feel betrayed that the therapist is after all not available, and begin to deteriorate (Werbart, 1997).
The consensus of the literature is that dynamic psychotherapy is usually helpful for the patient with DPD. Hill (1970) noted that only two of 50 cases treated showed no observable improvement. Treatment required several months to more than 2 years. Leeman and Mulvey (1975) noted that short-term (3–7 months’ duration), focused psychotherapy was successful in five of six patients, although one patient required a second course of treatment. Hoglend (1993) found that more than 30 sessions were needed. Most authors used weekly sessions.
The comparative efficacy of short-term versus long-term treatment has not been adequately addressed. In general, short-term psychotherapies are most likely to succeed when a circumscribed, dynamic conflict or focus is present, the patient can form a therapeutic or working alliance rapidly, and the tendency to regress to severe dependency or acting out is limited (Malan, 1976; Davanloo, 1978; Horowitz et al., 1984; Luborsky, 1984; Strupp and Binder, 1984; Winston et al., 1991; Hoglend, 1993). Unfortunately, many patients with DPD will not meet these criteria. Short-term dynamic therapies usually require once-weekly sessions over 3–9 months.
Hoglend (1993) found that among patients with PDs, the length of treatment was more essential for long-term dynamic improvement than were patient characteristics, such as suitability, cluster category, or initial global functioning. Significant long-term dynamic changes did not appear before 30 sessions, and the amount of change correlated with the number of sessions, a finding not obtained in those without PDs. Many patients do better in longer-term, dynamic psychotherapies or psychoanalysis. These include those who have failed to improve in short-term treatments, have multifocal conflicts or histories of significant emotional neglect or abuse. These treatments generally require two to four sessions per week over a period of several years to work through the dependent transference.
Turkat and Carlson (1984) reported two successive behavioral treatments of a patient with DPD. The patient had initially been treated with behavioral techniques for anxiety-related complaints but had relapsed immediately after termination. The authors then reformulated the case, focusing on the dependency constructs of excessive reliance on others and deficient autonomous behavior, which they posited resulted from long-standing anxiety over independent decision making. The therapist and patient constructed a hierarchy of situations with which the patient had little experience but about which the patient was required to make independent decisions. The therapist emphasized previously taught anxiety management skills. Treatment proceeded every other week for 2 months. As therapy progressed, the patient showed decreasing levels of self-rated anxiety, and less avoidance of situations requiring independent decisions. The gains were maintained at 1-year follow-up.
A. T. Beck et al. (1990) and J. S. Beck (1997) have described cognitive-behavioral treatment for DPD. As in dynamic therapy, they view the patient–therapist relationship as a microcosm of the patient's dependent beliefs and behaviors. The therapist must foster the therapeutic alliance early and adjust the therapeutic approach somewhat to maintain it. For instance, some patients need to begin a session telling the therapist whatever is on their mind, in order to cooperate subsequently with more directed or structured tasks. The therapist formulates the case and then chooses each technique to foster accurate self-appraisal and independent decision making and behavior. The patient's dependent behavior is initially accepted, but the therapist encourages self-reflection and agenda setting for sessions.
Independence is first encouraged by helping the patient set goals for treatment. Using a Socratic method avoids directing the patient's agenda. The therapist continually challenges the patient's dichotomous thinking (e.g., ‘If I am not fully successful, then I'm inadequate’) to improve self-evaluation. Successful graded exposure to anxiety-provoking situations in real life challenges the patient's belief about being incompetent. Patient diaries can be used to monitor the patient's automatic thoughts, especially of inadequacy, highlighting their negative consequences. The therapist can challenge the patient to select healthier responses that aid the development of positive schemas. Relaxation training may aid in the reduction of anxiety surrounding independent reflection and decision making. Assertiveness training and role playing may help counter submissive behavior whenever real skill deficits exist.
J. S. Beck (1997) recommends a session format that includes checking the patient's mood, providing a bridge between sessions, setting an agenda for the session, reviewing any homework, discussing the items on the agenda, and then summarizing the session and giving and obtaining feedback. Patients are given work sheets that can help them combine previous work and current situations to prepare for the next session. The formulation or ‘cognitive profile’ plays a crucial part in helping the patient understand connections between early experiences, core beliefs, and compensatory strategies as well as reactions to current situations. Once therapist and patient have identified maladaptive core beliefs, the patient can fill out a ‘core belief worksheet’ each session that contrasts the old maladaptive belief with disconfirming experiences and substitutes new more flexible and adaptive beliefs. The therapist can use a variety of different techniques to help the patient discover and shape new ways of thinking and behaving, such as proposing a behavioral experiment to test a belief.
Whenever resistance to change develops, the therapist must help the patient think through ambivalence about changing, with the goal of finding constructive substitutes for the loss of old dependent habits. As treatment progresses, the dependent transference can be reduced by the addition of group therapy. Toward termination, tapering the frequency of sessions will allow the patient to feel increasingly competent without frequent visits. At termination, the fear of losing the therapist may be mitigated by offering booster sessions at infrequent intervals. Specific guidelines regarding the optimal number of sessions have not yet been developed or tested.
Marchand and Wapler (1993) conducted a retrospective study of cognitive-behavioral treatment for panic disorder with agoraphobia. A chart review diagnosis of DPD, compared with nondependent patients, was not associated with any worse response to treatment.
Overall, treatment based on a cognitive-behavioral formulation of the mechanisms for a variety of dependent features shares many features with that based on psychodynamic formulation, although the treatments differ on some specific techniques. Further case studies and treatment trials are needed to differentiate the advantages of each approach.
Several reports suggest that group psychotherapy can be successful for the treatment of DPD. Montgomery (1971) used group therapy for dependent patients who used medications for chronic complaints such as insomnia and nervousness. All but three of 30 patients eventually discontinued medications and began to confront their anger at being dependent on the therapist. In an inpatient treatment setting for alcoholism, Poldrugo et al. (1988) found group therapy most beneficial for patients with DPD.
Sadoff and Collins (1968) employed weekly group psychotherapy for 22 patients who stuttered, most of whom had passive-dependent traits. Although the dropout rate was high, the authors found that the interpretation of passive-dependent behavior and attitudes (e.g., asking for help, believing that others are responsible for helping them) as a defense against recognizing and expressing anger proved helpful. Both stuttering and passive dependency improved in two patients who became angry and were able to confront their anger.
Torgersen (1980) studied college students who attended a weekend-long encounter group. On follow-up several weeks later, individuals who initially scored high on dependent traits had mixed responses. While the group experience left them feeling disturbed and anxious, they also reported becoming more accepting of their own feelings and opinions. No other changes were found.
Attrition may be higher in group than individual therapy for PDs (Perry and Bond, 2000), although may be less of a problem for individuals with DPD. Budman et al. (1996) demonstrated moderate improvements after an 18-month group for PDs (10% with DPD), although some changes were not evident until 6 months.
These reports suggest the usefulness of group psychotherapy in the treatment of DPD. Most clinicians employ weekly sessions of 1–11/2 hours duration. Sessions may be more frequent when group therapy is used as a major treatment modality in a day or residential treatment setting (Piper et al., 1993). Outpatient group therapy generally lasts 6 months to several years.
Both of these modalities are useful when patients require a higher level of support and treatment intensity than is available in most outpatient therapies. Such patients often have comorbid Axis I and II disorders, and a history of refractoriness to previous treatments (Karterud et al., 1992, 2003; Piper et al., 1993; Wilberg et al., 1998). Such therapeutic approaches usually employ mixtures of individual and group therapies along with additional services, such as occupational therapy, expressive therapies, guided work experiences or counseling, and so forth. Controlled (Piper et al., 1993) and uncontrolled studies (Krawitz, 1997; Karterud et al., 1992, 2003; Wilberg et al., 1998), including Cluster C PDs, generally demonstrate large effects. Temple et al. (1997) found that interpretive group therapy was specifically helpful enabling most of a group of patients who were very dependent on day hospital improve enough for discharge to outpatient care. Day treatment duration ranges from about 18 weeks to more than a year, although a naturalistic comparison of different day treatment centers in Norway found no differences in effectiveness for PDs between longer and shorter treatment durations (Karterud et al., 2003). Residential treatment is specifically useful for those patients who have failed to improve or deteriorated with outpatient therapy, while living alone or with family. Such patients usually require several months to a year or longer to progress to the point of living independently and benefiting from further outpatient therapy.
Some patients with DPD may live with family members who exert great degrees of influence over issues of support and autonomy. The family may view the patient as needing to be cared for, and the family reward and punishment contingencies maintain the patient in a dependent status. Increasing autonomy by the patient, which may include the threat of leaving home, is covertly experienced as threatening to the family. In such cases, family therapy, or periodic family meetings adjunctive to individual therapy, may help. The therapist's task is first to identify the functional relationships in the family that encourage dependency and discourage normal autonomy. The therapist must then help the family members initially develop a consensus on some modest goals for increased autonomy for the patient. As the patient begins to reach some early goals, the therapist can help the family revise the consensus. The therapist must point out discrepancies between attitudes of helping the patient and behaviors that undermine this goal. However, the alliance with the family members may become strained if the therapist takes too directive a stance. Family meetings range from once per week to once every few months, when adjunctive to individual therapy. There are no studies on the sole use of family therapy for DPD.
One common but unpredictable occurrence in the therapy of DPD arises when the patient experiences a significant separation, loss, or diminution in personal or financial support. Such stressors often overwhelm the ability to employ newly acquired attitudes and skills, resulting in a regression in defensive functioning and an increase in dependent wishes, requests, and behaviors. This may be further exacerbated by recurrence of panic, general anxiety, somatic symptoms, or a major depressive episode. Some regression to earlier more dependent functioning is common. This may strain the therapeutic alliance if the patient perceives the therapist as insensitive to his or her emotional reactions, disappointed, impatient, or too demanding of progress. The therapist must find a balance between listening, being supportive, offering suggestions and some direction, which the patient will find helpful, while temporarily accepting this interruption in the tasks of growth. In fact, if the therapist negotiates such crises well, the alliance will be strengthened and the patient, feeling supported and understood, may return sooner than imagined to working on issues of autonomy, separation sensitivity, effective coping, and self-esteem enhancement.
The challenges in treating individuals with DPD often arise in the therapeutic relationship itself in the form of transference and countertransference problems. Five such patterns often arise in the treatment of individuals with DPD (Perry, 2001).
In the first instance, the patient entering therapy may make many demands or requests of the therapist for advice, succor, or concrete help, which the therapist is unable to meet. In one study, such patients often terminated early in therapy and were rated as having had unsuccessful outcomes (Alexander and Abeles, 1968). The therapist should give special attention to helping modulate these patients’ demands early in treatment to prevent overwhelming disappointment and dropout. These patients also invite a countertransference response of emotional withdrawal and disengagement, which in turn reinforces neurotic guilt about needs.
A second problem may occur when the patient repeatedly attempts to put the therapist in the role of a dominant other who will both take responsibility for all decisions and tell the patient how to run his or her life (Hill, 1970; Saul and Warner, 1975). If the therapist assumes this directing countertransference role, he or she may become an external substitute for the patient's own will. Some therapists do this out of a sense of exasperation at the patient's protestations of helplessness or because of a personal wish to assume an idealized role as wise and all-knowing. This reinforces the patient's emotional reliance on the therapist without challenging him or her to learn more independent ways of coping. Directive approaches may have a useful, but limited, role during crisis interventions, but even cognitive-behavioral therapies require the therapist to foster the patient's independent decision making (A. T. Beck et al., 1990).
A third problem results when the patient avoids making real changes but stays in therapy to maintain the emotional attachment to the therapist (Leeman and Mulvey, 1975). The patient's compliant attitude toward the therapist may be mistaken for cooperation with the goals of therapy. Such individuals have tacitly refused to accept responsibility for making changes and may have their passivity reinforced if the therapist does not recognize and deal openly with this problem.
Case example in psychodynamic therapy. A 47-year-old accountant presented with feelings of alienation, insecurity, and needing to please others at work and in his marriage and family of origin. Being left at a boarding school in early childhood during a prolonged period of family dislocation left him feeling alone, emotionally neglected, and needy of others help. He eagerly participated in therapy and over 5 years made several advantageous career moves—usually prompted by an external event such as a layoff—and became more assertive in his personal relationships. Nonetheless, there were crucial areas where he appeared to repeat the same well-worn themes, always bringing in a series of complaints followed by discussion, which he invariably found helpful. Yet, there was no clear progression toward an ultimate sense of autonomy, satisfaction with marriage and family, or termination of treatment.
The therapist recognized that negative feelings expressed toward the therapist were minimal, and that the patient worked to keep the relationship comfortable, at the cost of continuing in the role of the dependent, needy one. The therapist began to interpret this pattern, that the patient reported problems followed by small successes, which he gave as gifts that served to make the therapist feel helpful, and thus maintain the relationship as it was. In fact they were all displacements and reaction formations against the harder themes of fear of abandonment, fear of hurting others and being seen as aggressive if he furthers his own wishes, envy toward those more successful, and disappointment in the therapist's limited power to change the patient's life. The therapist began to interrupt the weekly myriad of stories and point out their diversionary aim, sometimes revealing that they led him to day dream or even get sleepy, which kept them from really connecting. This mobilized the patient who responded with dysphoric feelings, but increased interest, and attention to his acceptance of the status quo. Focusing on these in-session phenomena led to an increase in relevant earlier memories juxtaposed with confronting changes not imagined possible. While the therapy became less comfortable for the patient, he reengaged with the more central rather than peripheral areas of conflict. The therapist also found the sessions more engaging, and progress reemerged.
A fourth problem may occur with patients who have unsatisfying, punitive relationships, commonly described as masochistic or self-defeating. The patient's repeated stories about mistreatment may evoke in the therapist a desire to control the patient's self-defeating pattern or even to punish the patient for not changing. Should the therapist challenge the patient to leave or to assert him or herself in the relationship, the patient may become extremely anxious, because of the strength of the emotional attachment or the realistic threat of a punitive response from the patient's partner, or fear of losing the therapist if he or she stays with the partner (Perry and Flannery, 1982, 1989). Such a challenge may make the patient feel trapped between pleasing the therapist and being punished by the patient's partner. It may result in panic or early termination. Instead the therapist must address the patient's fantasies that submission brings with it entitlement to be taken care of by dominant others (Waska, 1997). The patient may resist mourning this expectation that he or she is owed the right to be taken care of, as one may have trouble giving up on a debt not repaid (Perry et al., 1989).
A fifth problem occurs with the patient who avoids dealing with separation issues in therapy, which often involve mourning past losses or disappointments (Werbart, 1997). This may lead the patient to avoid anticipating the loss of the therapist at termination and mourning appropriately. The therapist may tacitly collude with this avoidance, because of a countertransference fantasy of always being available or fear of provoking separation panic or distress. Failure to confront the avoidance may result in a failure to make lasting dynamic changes, leaving the patient at risk for a sense of betrayal after termination, followed then by deterioration.
The psychotherapy of DPD is usually quite helpful. All modalities, individual, group, and residential treatment, report sizable treatment effects, as do the two major theoretical schools studied: dynamic and cognitive-behavioral. The effects on symptoms are generally large but there is less documentation in areas such as improved autonomy, healthy, nonsubmissive relationships, and successful employment. The required treatment duration leading to full recovery and healthy function still remains undetermined. This generally positive conclusion should be tempered by recognizing that there are a number of challenges in the psychotherapy of DPD that can allow improvement to plateau rather than proceed toward substantial improvement and a healthy termination. The next generation of studies, focusing on both process as well as outcome, should address these. Until then, therapists should pay particular attention to the potential for therapeutic stalemates after an initial period of improvement, and attend to particular patterns including the transference–countertransference, which may provide a key to addressing the stalemates and allow progress to resume.