|Home | About | Journals | Submit | Contact Us | Français|
Exposure to reminders of trauma underlies the theory and practice of most treatments for posttraumatic stress disorder (PTSD), yet exposure may not be the sole important treatment mechanism. Interpersonal features of PTSD influence its onset, chronicity, and possibly its treatment. The authors review interpersonal factors in PTSD, including the critical but underrecognized role of social support as both protective posttrauma and as a mechanism of recovery. They discuss interpersonal psychotherapy (IPT) as an alternative treatment for PTSD and present encouraging findings from two initial studies. Highlighting the potential importance of attachment and interpersonal relationships, the authors propose a mechanism to explain why improving relationships may ameliorate PTSD symptoms.
Interpersonal factors are extremely important in psychotherapy, regardless of the modality. A special contribution of interpersonal psychotherapy has been a primary focus on this dimension and the role of attachment throughout human development and distress. This guest psychotherapy column contributes an extensive literature review and the results of trials of interpersonal psychotherapy in ameliorating posttraumatic stress disorder.
Norman A. Clemens, MD
Confronting avoidance through exposure has constituted the central evidence-based paradigm for treating posttraumatic stress disorder (PTSD). Cognitive behavioral psychotherapies, which have demonstrated efficacy in treating PTSD and other anxiety disorders,1 reduce PTSD symptoms through repeated imaginal and in vivo exposure to memories and environmental reminders of traumatic situations. This process promotes habituation to trauma, thereby reducing activation of fear networks that drive symptoms.2 With repeated exposure, the patient organizes a more coherent narrative of the traumatic event. Research on conditioned responses, habituation, and extinction, and the replicated success of such therapies in outcome studies have made exposure the dominant explanatory model in PTSD therapeutics.
Almost all empirically tested psychotherapies for PTSD have relied on exposure techniques, dating back to Kardiner and Spiegel’s treatment of World War II veterans.3 Prolonged exposure (PE), perhaps the best established evidence-based treatment for PTSD, has shown efficacy in multiple trials.4–6 It requires extensive emotional processing and narrative reconstruction of traumatic events in each therapy session as well as in daily home review of audiotaped sessions. Other treatments have demonstrated efficacy, including other cognitive behavioral therapies4,6–11 and eye movement desensitization and reprocessing (EMDR).12 Both EMDR and Horowitz’ information-processing psychodynamic therapy13 are hybrid treatments incorporating cognitive behavioral techniques. All focus patients on consciously reviewing painful emotional details of their traumatic experiences. Consensus guidelines endorse PE, cognitive-behavioral therapy, and anxiety management as first-line psychotherapies for PTSD.14
Not since the heyday of psychoanalysis has a single theory so dominated a field as exposure-based theory does the psychotherapy of anxiety disorders today. Theoretical models of PTSD emphasize avoidance as a core symptom that prevents adaptive processing of trauma Theoretically; such processing should preclude development of chronic PTSD symptoms.15,16 The goal of exposure-based therapies is to reverse avoidance. Through desensitization, modeling, reciprocal inhibition, or flooding, exposure-based therapies facilitate a focus on traumatic memories, and thereby promote cognitive and emotional integration of trauma. Foa and Kozak2 noted that exposure as an anxiolytic treatment dates back to Freud and has emerged as “a common principle for…treatment…across schools of psychotherapy” (p. 20) for anxiety disorders. Recent American Psychiatric Association treatment guidelines for PTSD reinforce this focus: “The shared element [across efficacious treatments] of controlled exposure may be the critical intervention.” (p. 5).17 The Institute of Medicine concurred.18
An unfortunate consequence of the success of the exposure-based therapy model has been the neglect of other potentially useful treatment paradigms for PTSD. To date, interpersonal factors have played a marginal role in PTSD treatment models. Yet core features of PTSD are intrinsically interpersonal. PTSD has deleterious effects on interpersonal relationships and psychosocial functioning. Interpersonal variables strongly influence the chronicity of PTSD.19,20 This article reviews the importance of interpersonal factors and social support in PTSD and speculates on an interpersonal treatment model that has shown initial promise.21,22
PTSD has interpersonal aspects. In the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) PTSD diagnostic criteria (p. 468), the avoidance symptom cluster includes avoiding “conversations associated with the trauma” and “people” who evoke traumatic memories. Individuals may have a “feeliing of detachment or estrangement from others,” show a “restricted range of affect,” or have “irritability or outbursts of anger” in interpersonal contexts.23 “Diminished interest or participation in significant activities” typically occurs in social settings. The criteria also include a “sense of a foreshortened future” with social aspects of life given as prime examples: “career, marriage, children.”
The PTSD clinical literature describes a range of associated interpersonal features. Patients have dysregulated affect and become withdrawn, mistrustful, and “interpersonally hypervigilant”21 in social circumstances. PTSD compromises intimacy with others,24 burdens significant others,25 and impairs social and marital functioning.26–29 Moreover, individuals with PTSD may fall into maladaptive interpersonal patterns that increase the risk of revictimization.26,27,30
Anecdotal evidence suggests that interpersonal distortions and functional deficits are typical foci in community treatment of patients with severe childhood trauma and neglect. Such treatment models appear throughout the psychodynamic literature,31–34 suggesting a widespread if unresearched approach.
Although most people (50%–90%) encounter trauma during their lifetimes,35,36 only about 8% develop full PTSD.35 Vulnerability to PTSD involves interaction of a biological diathesis, early childhood developmental experiences, and trauma severity. Predictor models have consistently found that childhood trauma, chronic adversity, and familial stressors increase the risk for PTSD and for its biological risk markers after a traumatic event in adulthood.37–39 The effect of childhood trauma, which is poorly understood, may reflect both traumatic experiences and attachment problems.40–42
Research demonstrates that prior trauma increases vulnerability to anxiety and other disorders,43 but there is no consensus on explanatory mechanisms. Shared diatheses probably interact with both interpersonal trauma and neglect in childhood and with temperamental factors that influence behavior.37,44–46 Separation anxiety disorder and inhibited temperament in childhood specifically predict adult anxiety disorders.46,47 Several longitudinal studies have linked the diagnosis of separation anxiety and behavioral inhibition in early childhood to various anxiety disorders later in life.47–51 In one study, 30% of behaviorally-inhibited children subsequently developed anxiety disorders.52
Another dyadic (interpersonal) variable, insecure mother-child attachment, contributes to children’s development of clinical anxiety syndromes. In one study, 65% of 20 children aged 18 to 59 months, whose mothers had anxiety disorders, had behavioral inhibition (BI).53 Of children with BI, 80% had insecure maternal attachments. Secure attachment was associated with less anxiety, a striking finding given the high rate of BI in this group. These results imply that insecure mother-child attachment contributed more than inhibited temperament to high risk for anxiety disorders.53,54 A longitudinal study of 172 mothers with anxiety disorders and their infants found a link between anxious/resistant attachment assessed in infancy and diagnosable DSM-IV anxiety disorders in adolescence.55
Mothers of anxious children display more intrusive involvement and negative interactions (overinvolved parenting style) than mothers of children without psychiatric disorders.56 Fonagy has elaborated the complex interactions between genetic predisposition and effects of anxious parenting in the first 3 years of life on subsequent development of anxiety disorders.57 Vanderwerker et al. found that a retrospective diagnosis of childhood separation anxiety was associated with developing complicated grief, a disorder related to PTSD, in 283 recently bereaved individuals.58
Thus research suggests that parenting styles and primary childhood relationships increase later risk of developing anxiety disorders, including PTSD. We speculate that premorbid insecure attachment and difficulty in establishing basic trust,42 due to biological predisposition and/or early life experiences, including trauma, leave individuals vulnerable to subsequent interpersonal disruptions after traumatic experiences, and hence to developing PTSD. Differing attachment styles will necessarily affect how individuals seek and benefit from social support, which is perhaps the strongest predictor to date of whether an individual shows resilience after a traumatic experience or develops PTSD.59,60
Bowlby explored the developmental importance of social supports for adult psychopathology, studying the effects of secure and insecure childhood attachment on depressive vulnerability.61,62 Bowlby postulated an evolutionarily determined, instinctual human drive to form attachments that enhanced infants’ survival by eliciting protective, caretaking behaviors from adults. Ethological research now supports this theory in species-specific behaviors.63 As they develop, children begin to explore their environment, gradually decamping from the “secure base” of their attachment figure.61
Whereas secure attachment provides confidence both for exploration and eliciting support from relationships, conflicts in this early caregiving connection produce an insecure attachment style associated with adult vulnerability, such as fearful avoidance. Hence, loss of one’s mother in the first decade of life is a risk factor for subsequent depression.64,65 Bowlby’s social attachment theories resonate with Erikson’s maturational phase of basic trust versus mistrust.66 With regard to PTSD, secure attachment would bolster and insecure attachment would compromise the ability to confidently explore the environment—to risk “exposure” to trauma reminders and use available social supports to overcome initial fearful reactions.
Recent studies of attachment style and PTSD support this model. Stovall-McClough and Cloitre reported that unresolved childhood trauma increased the risk of a PTSD diagnosis seven-fold and was associated more with avoidant than dissociative PTSD symptoms.67 Declercq and Palmans found that both adult attachment style and perceived social support moderated development of PTSD in 544 subjects exposed to a critical incident.68 Fraley et al. found that adult survivors (N = 45) with secure attachment who were highly exposed to the events of September 11, 2001 described fewer PTSD and depressive symptoms on follow-up than those with insecure attachment.69 Twaite and Rodriguez-Srednicki, studying 284 New Yorkers affected by 9/11, found that a history of childhood sexual or physical abuse increased the likelihood while secure attachment decreased the likelihood, of PTSD symptoms. Thus secure attachment protected against and potentially mediated the development of PTSD.70
A major meta-analytic study showed that social factors play a key role in PTSD. Based on a careful meta-analysis of research on risk factors, Brewin et al. reported that “lack of social support” was the largest single predictor of developing PTSD after a traumatic event (r = 0.40).59 Life stress (r = 0.32) and trauma severity (r = 0.23) were the next largest predictors. Gender, age, race, socioeconomic status, education, intelligence, psychiatric and family history, childhood abuse, and previous trauma were less important predictors. Ozer et al. replicated this finding in another meta-analysis: they found that posttraumatic perceived social support had a larger weighted average correlation (weighted r = −0.28, from 11 studies with a combined N = 3,537) with PTSD than five of the other six predictors they examined (excepting only peritraumatic dissociation, r = 0.35). The association between lack of perceived social support and PTSD strengthened as time elapsed between the traumatic event and the assessment of PTSD.60 These findings suggest interpersonal factors, such as the ability to build and elicit support from social networks, promote adaptive responses to traumatic experiences. Conversely, lack of social support is a major risk factor for developing PTSD.
A large literature links social support to the onset of major depressive disorder (MDD) and schizophrenia. Absence of social support increases the risk of developing MDD, whereas its presence reduces this risk.64,65 Having a confidant has also been found to reduce the risk of depression.65 Life stressors, including complicated bereavement, struggles in relationships (role disputes), upheavals in marital status, changes in housing, job status, or physical health (role transitions), and social isolation (interpersonal deficits), also increase the risk for MDD.64,71 Events involving interpersonal loss occur in at least half of the cases of new onset panic disorder.72 A dysregulated or hostile interpersonal environment (high “expressed emotion”)73 predicts relapse in schizophrenia,74 bipolar disorder,75,76 and MDD.73 Thus social support may mediate several psychiatric disorders, including PTSD.
Some patients with PTSD refuse exposure-based therapies, an unsurprising consequence of their anxious avoidance. PTSD treatment guidelines state: “Some trauma survivors are reluctant to confront trauma reminders and to tolerate the high anxiety and temporarily increased symptoms that sometimes accompany exposure. Thus, not everyone may be a candidate for exposure-based treatment”(p. 324).1 A non-exposure-based, interpersonal model might effectively engage such patients in treatment. Little has been written, and and only limited research has been done, on interpersonal approaches to the treatment of PTSD.32,33 Cloitre et al. developed a manualized, two-phase treatment that addresses interpersonal deficits in victims of childhood sexual abuse, with promising pilot findings.77
Interpersonal psychotherapy (IPT) is a time-limited treatment developed for the treatment of major depressive disorder.64,71 Its demonstrated efficacy for MDD has led to its adaptation and testing for other psychiatric diagnoses.64,71 IPT has been shown to build social skills.64,78 Focusing on an interpersonal crisis, IPT aims to relieve symptoms by enhancing social functioning and recruiting social supports. Emphasizing the recognition and repair of interpersonal difficulties, IPT provides opportunities to reverse interpersonal avoidance, increase social support, undergo corrective emotional experiences with others that potentially modulate trauma-related interpersonal distortions, and improve demoralization and helplessness that inhibit motivation to overcome trauma-related fears.
We developed a manualized modification of IPT for PTSD21,79 that focuses on how trauma (or PTSD) has compromised patients’ current interpersonal perspective and social functioning. The central tenet of treatment is that trauma impairs the individual’s ability to use the social environment to process environmental trauma, shattering perceived environmental safety and poisoning trust in interpersonal relationships. Hence individuals with PTSD develop “interpersonal hypervigilance,”21 withdraw from or distance themselves within relationships, and restrict social activities. Experiencing the environment and relationships as dangerous then triggers maladaptive social functioning that helps to perpetuate PTSD. PTSD symptoms, in turn, reinforce social detachment and dysfunction. Treatment counters the perceived helplessness and shamefulness of PTSD with a sense of interpersonal competence, and redirects the patient’s attention from inner preoccupation with past trauma to coping with the immediate interpersonal outer world.
We conducted an open trial of 16 subjects with chronic PTSD (7.7 ± 9.9 [mean ± SD] years) using our manualized adaptation of IPT (data on the first 14 subjects were reported in Bleiberg and Markowitz21). The study involved 14 weekly IPT sessions that focused on daily interactions with family members, co-workers, and others in the patients’ environment rather than on exposure. Of the 16 subjects with chronic PTSD, 15 (94%) completed treatment. Scores on the Clinician Administered PTSD Scale (CAPS),80 a benchmark PTSD scale, fell from 66.3 ± 16.0 to 23.5 ± 16.1 (p = 0.001), with improvement across symptom clusters. Rate of treatment response, defined a priori as ≥ 50% decrement in CAPS score, was 69%; and rate of remission, defined as CAPS score ≤ 20, was 44%. Depression and anger were also reduced.21 Interestingly, as patients improved, they spontaneously began to expose themselves to trauma reminders. Subjects also reported improved social functioning on the Social Adjustment Scale (SAS-SR),81 Inventory of Interpersonal Problems (IIP),82 and other measures.
These preliminary findings of this open trial require replication under controlled circumstances. With support from the National Institute of Mental Health, we are conducting a randomized controlled trial comparing three 14-week psychotherapies that each employ very different mechanisms for treating chronic PTSD:
In a separate, randomized trial, Krupnick et al. treated 48 low-income, multiply traumatized women with chronic PTSD recruited from public family planning and gynecology clinics.22 Patients received either 16 2-hour sessions of group IPT or a waiting list condition. Mean CAPS scores fell from 65.2 ± 20.9 to 40.6 ± 16.9 with group IPT, with gains maintained at 4 month follow-up. In contrast, mean CAPS scores in the wait list group decreased from 62.6 ± 16.6 to 56.8 ± 12.2 on the CAPS. As in our individual IPT therapy, scores on the Hamilton Rating Scale for Depression83 and on the IIP also improved.
Our current randomized controlled trial of IPT for PTSD administers instruments repeatedly over the course of psychotherapy to measure whether self-initiated exposure to trauma reminders or interpersonal change may mediate therapeutic improvement. Reflective Function,42,84 an assessment that measures how individuals judge emotional aspects of relationships, their own emotions, and feeling states of others—i.e., a proxy for attachment–may provide a novel gauge of IPT’s mechanism. We will also examine whether IPT increases Reflective Function (i.e., improves attachment).
Might secure childhood attachment and premorbid social support predict better posttraumatic social support, marking individuals with a better prognosis? Higher levels of perceived social support in adulthood may indicate better interpersonal skills, which might predict a more secure treatment alliance, a crucial outcome factor.85,86 Greater social support also suggests better integration into social networks, greater availability of help in the immediate environment—other presumably good prognostic signs. In contrast, insecure attachment increases the perceived risk in exploring one’s environment, a pattern with negative implications for individuals with PTSD. Interpersonal aspects of PTSD deserve further exploration, as do treatment approaches that address this aspect of PTSD symptomatology.
Supported in part by grant MH079078 from the National Institute of Mental Health
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.