Attachment-Based Family Therapy: Theory, Clinical Model, Outcomes, and Process Research

Abstract

Attachment-based family therapy (ABFT; G. S. Diamond et al., 2014) is an empirically supported treatment designed to capitalize on the innate, biologically based, caregiving instinct and adolescent need for attachment security. This therapy is grounded in attachment and emotional processing theory and provides an interpersonal, process-oriented, trauma-informed approach to treating adolescents struggling with suicide and associated problems such as depression and trauma. ABFT offers a clear structure and road map to help therapists quickly address the attachment ruptures that lie at the core of family conflict, which can fuel adolescent distress. Several clinical trials and process studies have demonstrated empirical support for the model and its proposed mechanisms of change. In this paper, we provide an overview of the theories underlying the model, the clinical strategies that guide the treatment, the outcome research that demonstrates efficacy, and the process research that explores the proposed mechanisms of change.

Suicide is the second leading cause of death among youth ages 15–24 (Centers for Disease Control and Prevention [CDC], 2019). An estimated 12.1% of adolescents contemplate suicide, 4.0% make a plan, and 4.1% make an attempt. In 2011, 4,688 adolescents in the US died by suicide (Nock et al., 2013; Hoyert & Xu, 2012). Even adolescent suicidal ideation can be frightening for families, a worrisome burden to medical and behavioral health providers and a driver of cost to medical systems. Given the heavy burden of suicide and suicide risk to individuals, families, and society, the identification of effective prevention and interventions for this population is critical (CDC, 2019).

The factors contributing to suicidal risk for adolescents are multifaceted. Extensive research has investigated cognitive, biological, environmental, and social factors (Frey & Cerel, 2015; Cha et al, 2019). One important risk factor consistently associated with adolescent suicidality is the quality of adolescent–parent relationships. Family risk factors, such as high conflict, low cohesion, poor attachment, and ineffective parenting, are associated with suicidality and depression in adolescents (G. Diamond et al., 2021; Wagner et al., 2003). In one study, Brent et al. (1988) found family conflict preceded 20% of suicides and 50% of nonfatal suicidal acts.

At the same time, families can serve as a tremendous protective factor. Studies have shown that a positive family environment can buffer against the stressors that contribute to youth suicidal risk and depression (Bilsen, 2018). Family cohesion, emotional support, and caregiver supervision are protective factors in reducing risk for suicide (Wagner et al., 2003). After controlling for depression and stressful life events, adolescents who described their family members as mutually involved and supportive, were three to five times less likely to be suicidal than their peers from less cohesive families (Rubenstein et al., 1989, 1998).

Given the powerful role of family relationships as a risk and protective factor, involving parents in the treatment of youth at risk for suicide may have several advantages. First and foremost, treatments that include parents have shown the most promise for treating suicidal youth (Asarnow et al. 2011, 2015; Harrington et al., 1998; Huey et al., 2004; Diamond G.S et al., 2010, 2016; Esposito-Smythers et al., 2011; Rotheram-Borus et al., 2000; Rossouw & Fonagy, 2012; Pineda & Dadds, 2013). In fact, in a recent review, Brent et al. (2013) concluded that interventions that included a family component tended to be more effective for treatment engagement and the reduction of suicidal ideation and self-harm. Second, parents serve as a safety net for these patients, often providing daily (if not hourly) suicide monitoring. If prepared, they can be the first to recognize changes in their adolescent’s mood or other indications of distress. Parents can also help adolescents apply new coping skills (e.g., safety plan, mindfulness). Third, parental are critical in determining adolescents’ attendance in treatment. Parents provide scheduling, transportation, payment, and the motivation to attend. Indeed, Shelef et al. (2005) found that although the therapeutic alliance with adolescents predicted treatment outcomes, the therapeutic alliance with parents predicted treatment retention.

Given these benefits, it is not surprising that most adolescent suicide treatments now include some form of parent component (G. Diamond, Asarnow, Berk, 2014). Interestingly, the structure, role, and purpose of family involvement varies greatly across treatment modalities. Some modalities are primarily individual therapy (e.g., CBT), with a primary focus on adolescent coping skills and psychological functioning (Brent et al., 2009; Brown, et al., 2014; Wells & Albano, 2005). These modalities often add a family component to augment or potentiate treatment. Generally, these family interventions are psycho-educational programs focused on improving parenting skills and risk management. Recently, some treatments have elevated the role of families to a more clinical focus but still view improving individual functioning as the primary mechanisms of change. These treatments include I-CBT (Esposito-Smythers et al., 2019), SAFETY (Asarnow et al., 2015), mentalization therapy (Rossouw & Fonagy, 2012), and dialectical behavioral therapy for adolescents (McCauley et al., 2018).

On the opposite side of the spectrum, a more traditional family therapy modality focuses on family dynamics as the primary treatment target. Suicide assessment, monitoring, and management are critical, but the primary initial clinical target is reducing family conflict and increasing warmth and structure. In some families, interpersonal dynamics are major causes of suicide-related distress (e.g., a difficult divorce, marital conflict/violence, parental psychopathology, neglect or abuse). In other families, adolescent distress may be driven by forces outside the home (e.g., bullying, school or social failure, gender or sexual discrimination). Unfortunately many suicidal adolescents cannot turn to their parents for comfort and protection because of family conflict or poor communication. This lack of a secure family base can exacerbate distress. Our family therapy model aims to initially address this family context to decrease discord and increase support. This process also affords an in vivo opportunity for everyone to learn and practice new emotion regulation and communication skills.

Attachment-Based Family Therapy

Attachment-based family therapy (ABFT; G. S. Diamond et al., 2014) is an empirically supported treatment designed to capitalize on the innate, biological desire for meaningful and secure relationships. ABFT is grounded in attachment theory and provides an interpersonal, trauma-informed approach to treating adolescent depression, suicidality, and trauma. Although process oriented in nature, ABFT offers a clear structure and road map to help therapists quickly address the attachment ruptures that lie at the core of family conflict. Therapy does not start with behavioral management. Instead, we work to uncover those experiences (e.g., abuse) and relational processes (e.g., high conflict, low warmth) that prevent adolescents from turning to parents for help when feeling suicidal. Solving these problems becomes the context for adolescents to learn new cognitive and emotional coping skills and caregivers to improve their parenting practices. In this paper, we begin with a review of attachment and emotional processing theories and their application to our model. Then, we review the outcome research that demonstrates the efficacy of ABFT, as well as work focused on adapting ABFT to other populations. We end with a review of the extensive process research that we have conducted over the past 15 years, primarily led by Gary Diamond at Ben-Gurion University.

Theoretical Foundation

Attachment

Attachment theory provides the primary theoretical and clinical framework of ABFT (Bowlby, 1969, 2008). Attachment theory proposes that children who experience their caregivers as sensitive, responsive, and available develop confident expectations of relational security. They feel the world is a safe place, and they are worthy of being loved and protected. These safe and supportive parental relationships become internalized working models of what children expect from future relationships. When children do not receive responsive parenting, they develop defensive strategies that protect them from being hurt or disappointed. Some children develop a dismissive attachment style where they stop expecting or relying on caregivers to help them cope with life’s stressors. Some children develop a preoccupied style where they seek their caregivers’ attention yet consistently fear rejection. Other children develop a more disorganized attachment orientation that is characterized by wanting to be loved and cared for but holding a deep fear, mistrust, or inner conflict about caregivers. These are often children who have experienced abuse. Bowlby (1969) believed that these negative relational experiences shape children’s internal working models of self (e.g., “I am not worthy of being loved”) and of others (e.g., “I cannot trust anyone”).

Attachment security also affects the learning of emotion regulation and self-reflection skills. For example, when feeling scared, securely attached children turn to their caregivers for comfort and soothing. Over time, repeated experiences of a caregiver helping a child downregulate fears and become internalized as a self-regulation skill. Caregiver responsiveness also frees children to be open to live development rather than being psychologically constrained by self-protection. Instead, they have the “epistemic space” and a safe environment to reflect on and tolerate vulnerable emotions and troubling thoughts and memories (Kobak & Cole, 1994). This enhances reflective functioning, which allows children (and then adults) to think about others’ point of view, perspective, or internal experiences (Fonagy et al., 1991; Slade, 2005). For Bowlby (1969, 2008), secure-based family relationships become the context in which children learn essential relational and psychological skills.

The role of secure attachment is no less important in adolescence. Rather than separation and individuation (e.g., Erikson, 1968), the central task of adolescence is to maintain attachment while negotiating autonomy (Steinberg, 1990). Adolescents who keep this balance do better in school, have fewer deviant peer relationships, and exhibit better health outcomes (Allen et al., 1998; Kobak et al., 2006; Lynch & Cicchetti, 1991; Rosenstein & Horowitz, 1996). When adolescents do not have a safe and protective parenting environment, they are less skilled in interpersonal problem-solving and emotion regulation. Lacking these skills puts them at risk for a number of psychiatric disorders, including depression and suicidality (Cicchetti & Toth, 1995; G. Diamond et al., 2021; Sheeber et al., 2001).

Attachment theory also provides a model for understanding therapeutic change. Because interpersonal interactions (e.g., family relationships) shape our internal working models or schemas of self and other, improvements in these relational experiences can revise the internal model. A good marriage or psychotherapy offer relationships that are supportive and encouraging, resulting in improved self-reflection (Kobak et al., 2015). Given this transactional model of change, many individual and family therapists have turned to attachment theory to describe the therapeutic process (G. S. Diamond et al., 2014; Fosha, 2000; Hughes, 2007; Johnson, 2004; Lieberman, 2004; Solomon & Seigel, 2003; Young et al., 2003). In individual therapy, the therapist serves as the secure base; the “good parent” who resuscitates the patients’ trust in others and confidence in themselves. However, the therapist’s empathic and validating responses (e.g., “That must have been hard for you”) pale in comparison to the validation children receive when their parents say, “It was not your fault. I am sorry.”

In family therapy, the adolescent sits in sessions with the actual family members with whom they have relational disappointments. Rather than emotionally withdrawing or acting out, we help the adolescent directly share these vulnerable thoughts and feelings with their parents about loss and felt injustices (e.g., neglect, abuse, abandonment, high criticism, low warmth, over-control). We help parents listen nonreactively, even if, initially, the adolescent’s recollection is not fully accurate or not the entire story. We use conversations to teach adolescents how to identify, articulate, and regulate primary, vulnerable feelings. These feelings may or may not fuel the depression, but, certainly, expressing them and feeling heard and supported can alleviate the depression and make adolescents feel less alone in their pain.

In this way, ABFT uses family conversation as an in vivo, experiential learning opportunity. At one level, these conversations help family members address avoided or ignored events and processes that fuel conflict and distrust (i.e., the divorce). At another level, these conversations provide an opportunity for adolescents and parents to practice new emotion-regulation and conflict-resolution skills learned in individual preparation sessions: learning to listen, learning to put feelings into words, learning to tolerate vulnerable emotions. At a third level, this conversation enacts (in vivo) a corrective attachment experience: Adolescents express painful thoughts and feelings, and parents remain available, responsive, and emotionally attuned. Engineering these attachment-promoting conversations experientially challenges the views and expectations of self and others (“Oh, maybe my parents can listen to me”). When adolescents are more direct, honest, and regulated, it revises caregivers’ view of them as autonomous people needing respect, love, and support. Therefore, from an attachment perspective, ABFT aims to restore or refurbish healthy, trustworthy, reliable, and emotionally sensitive caregiver–child relationships (G. S. Diamond et al., 2014).

Emotional Processing

The role of emotional processing in this attachment process cannot be underestimated. According to emotion theory (Greenberg, 2012; Greenberg & Safran, 1987; Pascual Leone & Greenberg, 2007), emotional processing occurs by first activating an old, maladaptive response to a meaningful event and then accessing what are often avoided, more painful, vulnerable emotions. Productive emotional processing involves accessing, connecting to, differentiating between, and expressing adaptive and often vulnerable emotions. Learning to express these more vulnerable feelings directly frees adolescents from the constrained and guarded insecure attachment strategies they have developed to protect themselves. When adolescents express vulnerability in family sessions, it activates caregiving instincts. As parents connect to their worry or fear, as opposed to their frustration and anger, they respond in a softer, more caring, and attuned manner. The therapist might say to an angry father, “Dad, I know you are angry for all the trouble Johnny is causing, but I imagine that you must have been scared when you took him to the hospital. Did you worry he was going to die?” Accessing these primary, vulnerable, adaptive emotions provides adolescents and parents with better information about their needs and activates healthier, more effective, interpersonal exchanges: the father stops shaming the son for his depression and begins to provide more support.

Attachment-Based Family Therapy Treatment Tasks

In ABFT, the restoration of healthy caregiver–child relationships is accomplished via the “corrective attachment experience” and subsequent autonomy building conversations. These processes are engineered using five distinct treatment tasks. Tasks do not equate with sessions. Instead, a task is a set of procedures, processes, and goals related to resolving issues or accomplishing specific aims in therapy (e.g., establishing alliance; Rice and Greenberg, 1984). Task I focuses on an essential family therapy goal: getting family members to engage in a therapy focused on relationship building, rather than on behavioral management. To achieve this, the therapist focuses on resuscitating the adolescent’s desire for protection and support and the caregivers’ longing to provide love and protection to their child. The therapist promotes the caregivers as the medicine (not the problem) to help adolescents cope with and recover from depression and suicidal ideation.

Task II consists of individual sessions with adolescents. The therapist aims to help adolescents identify and articulate their perceived experiences of caregivers’ attachment failures and prepare them to discuss these felt injustices in Task IV. Task III consists of individual sessions with the caregivers. The therapist helps caregivers consider how their own life stressors and intergenerational legacies of attachment affect their ability to connect better with their child. This self-reflection helps caregivers develop greater empathy for themselves and then their adolescents. With this insight, caregivers become more motivated to learn new emotion-coaching parenting skills and be more present, open, and supportive of their child’s struggles.

Next, in Task IV, the therapist brings the adolescents and caregivers back together to discuss how these relational disappointments have damaged trust in the relationship. As adolescents share these thoughts, feelings, and memories and receive acknowledgment and empathy from their caregivers, they become more willing to consider their own contributions to family conflicts. As caregivers acknowledge adolescents’ experiences, adolescents become more emotionally regulated, open, and collaborative. Although these conversations may neither address nor resolve all relational problems, this mutually respectful and often emotionally profound dialogue serves as a “corrective attachment experience,” which can set in motion a renewed sense of trust and commitment.

In this new emotional climate, caregivers become a resource and secure base for their adolescents. Task V then focuses on using caregivers to support adolescents’ exploration of competency and autonomy. Adolescents begin to seek comfort, advice, support, and encouragement from their caregivers while exploring new opportunities and managing life stressors. Topics in this task focus on rebuilding the life the depression or suicide attempts have destroyed or on identity development, with a focus on topics such as social relationships, racial or sexual discrimination, religious identity, or future hopes and dreams.

Attachment-Based Family Therapy Research

Based on the results from 20 years of research, ABFT is listed on several national and international registries for evidence-based practices (Diamond, Levey, Russon, 2016). Registries include the California Evidence-Based Clearinghouse for Child Welfare (promising research evidence for adolescent depression), SAMHSA’s evidence-based practice guidelines (2020; evidence-based practices for suicidal ideation, self-harm, and suicide attempts), Promising Practice Network (proven practice), Swedish Guidelines for the Treatment of Depression, and Children and Young People’s Improving Access to Psychological Therapies in England. These registry catalogue treatments have received some level of empirical evidence and are ready for dissemination. The majority of clinical trials on ABFT have been conducted with low-income, minority populations at the Center for Family Intervention Science, located at Drexel University and previously at the University of Pennsylvania, Department of Psychiatry. A majority of the process studies have been led by Gary Diamond at both our center and at Ben-Gurion University in Israel. The outcome studies are briefly summarized below and in Table 1.

Table 1.

Summary of ABFT Intervention Studies

Study Title	Authors	Population	Design	Outcome Measure	Post Outcomes	Follow-up Outcomes
Attachment-based family therapy for depressed adolescents	G. S. Diamond et al. (2002)	N = 32, M age = 14.9, 78% female, 69% African American	Random assignment: 12 weeks ABFT or 6 weeks waitlist control	BDI, K-SADS	-No difference in mean BDI scores -More adolescents in treatment condition who endorsed nonclinical BDI scores -Increased family cohesion and perceived attachment	87% of treated sample no longer met MDD criteria and showed significant reductions in symptoms
Attachment-based family therapy for adolescents with suicidal ideation	G. S. Diamond et al. (2010)	N = 66 adolescents, M age = 15.2, 83% female, 74% African American	Random assignment: 16 weeks ABFT (M = 9.71 sessions) or EUC (M = 2.87 sessions)	Suicidal ideation and depression tracking	-ABFT group had larger/faster reductions in SI during treatment -ABFT group had greater rates of clinical recovery	-ABFT group differences still maintained, as were treatment benefits -87% had SI scores below clinical cutoff
Attachment-based family and nondirective supportive treatments	G. S. Diamond et al. (2019)	N = 129, M age = 14.87, 50% African American, 30% White	Urn random assignment: 16 weeks ABFT or NST	SIQ monthly BDI, conflict, cohesion	Significant reduction in suicidal ideation and depressive symptoms in both groups. Few suicide reattempts
Hospital aftercare program evaluation	G. S. Diamond et al. (in preparation)	N = 20, M age = 14.9, 80% female, 65% African American	Random assignment: 16 weeks ABFT or EUC	Suicidal ideation, family attachment	ABFT more effective in preventing attempts, reduced attachment avoidance
Feasibility of attachment-based family therapy for depressed clinic-referred Norwegian adolescents	Israel & Diamond (2013)	N = 20, M age = 15.6, 50% female	Random assignment: 12 weeks ABFT or TAU	HAM-D, K-SADS, BDI	ABFT patients had lower HAM-D ratings, more movement to nonclinical BDI scores	Implementation challenges discovered
Cognitive behavioral and attachment-based family therapy for anxious adolescents: Phase I and II studies	Siqueland et al. (2005)	Phase I N = 8, M age = 15.5 Phase II N = 11, M age = 14.9	Phase I 16 sessions CBT-ABFT Phase II Randomized CBT or CBT-ABFT	ADIS, HAM-A, HAM-D, BAI, BDI	-Significant decreases in anxiety and depressive symptoms -No between-treatment differences
Attachment-based family therapy for suicidal lesbian, gay, and bisexual adolescents	G. M. Diamond et al. (2012)	N = 10 suicidal LGB youth	12 weeks LGB-sensitive ABFT		Decreases in SI, depressive symptoms, maternal attachment-related anxiety, and avoidance
Relationship-focused therapy for sexual/gender minority youth adults and their parents	G. M. Diamond et al. (2019)	N = 31 SGM youth + caregivers	6 months adapted ABFT for SGM		Perceived parental rejection decreased, acceptance increased. Reduction in caregiver negative behavior
ABFT for transgendered suicidal youth	Russon et al. (2021)	N = 10, TGD or sexual/gender minority youth	Open trial, 10 weeks	SIQ, BDI	Decreases in SI. No significant decreases in depression	ABFT feasible and acceptable treatment for TGD youth
Attachment-based family therapy and emotion-focused therapy for unresolved anger: The role of productive emotional processing	G. M. Diamond et al. (2016)	N = 32 (presenting unresolved anger toward a parent)	10 weeks ABFT or EFT	Anger attachment psychological symptoms	Only ABFT group showed decrease in attachment avoidance. More productive emotional processing in EFT. Both groups showed decreased unresolved and state anger

Clinical Trials Research

After small open trial studies (G. S. Diamond et al., 2003), we launched a full treatment development project to manualize the treatment, develop adherence measures, and conduct a randomized pilot (G. S. Diamond et al., 2002). At that time, most studies for adolescent depression explored medication and CBT. Our study was also the first family therapy study to be conducted with depressed adolescents and one of the few studies that included a large African American sample. We randomly assigned patients to 12 weeks of ABFT or 6 weeks of a waitlist control condition.

Of the 16 treatment cases, 13 (81%) no longer met the criteria for MDD at posttreatment, and 7 (47%) of the 15 patients on the waitlist who completed the diagnostic interview no longer met the criteria for MDD postwaitlist (2 [1] = 4.05, p < .04). Overall, 62% of the adolescents treated with ABFT had a BDI of nine or less, compared to 19% on the waitlist condition. At 6-month follow-up, 87% of the treated sample no longer met the criteria for MDD and showed significant reductions in depression, anxiety, and negative family functioning (cohesion and conflict). Finally, in terms of treatment retention, 63% attended nine or more sessions, a high retention rate for a population typically characterized by high levels of early attrition (Mano et al., 2009).

Fully Powered Randomized Controlled Trial

In the first fully powered randomized trial (G. S. Diamond et al., 2010), 66 adolescents were randomized to 16 weeks of ABFT or enhanced usual care (EUC). EUC involved assistance in obtaining a therapist in the community, a safety plan, weekly tracking of depression and suicidal ideation, and access to a 24-hour crisis line. In all, 60% of the adolescents reported having made a suicide attempt in their lifetime.

Compared to EUC, youth treated with ABFT exhibited significantly greater and faster reductions in suicidal ideation during treatment. These differences persisted at follow-up, with a large effect size of .97. ABFT was also associated with greater rates of clinical recovery on suicidal ideation posttreatment. These findings were strengthened by the consistency across self-report and clinician ratings. ABFT was even effective with the most severe youth presenting with comorbid anxiety and a history of multiple suicide attempts. A secondary analysis showed that youth with a history of sexual abuse also responded well to treatment (G. S. Diamond et al., 2012), a finding not supported in several CBT or CBT + medication studies (Asarnow et al., 2009; Barbe et al., 2004; Beautrais et al., 1996; Shirk et al., 2009).

Results also indicated that ABFT was associated with greater rates of clinical recovery and treatment retention. At posttreatment, 87% of patients receiving ABFT reported suicidal ideation scores not only below the clinical cutoff but also in a range consistent with or below that of a nonclinical sample of similar demographics (Reynolds & Mazza, 1999). For EUC, only 51% achieved this level of recovery. Benefits were maintained at follow-up, with a strong effect size (OR = 4.41). One limitation of this study was the low dose of treatment received in EUC compared to ABFT (average of 3 vs. 11 sessions, respectively).

Second Fully Powered Randomized Controlled Trial

To control for the low treatment dose of the previous study, in the next study, we tested ABFT against a more active control group: nondirective supportive therapy (NST; Brent & Kolko, 1991). Additionally, to control for parental involvement, we added a four-session family psychoeducation component to the NST (Family-Enhanced NST; FE-NST). To control for therapist factors, we used a crossover design in which the same therapists provided both treatments (G. M. Diamond et al., 2019).

Adolescents in the ABFT condition showed significant reductions in suicidal ideation (t (127) = 12.61, p < .0001), with an effect size of d = 2.24. Adolescents in the FE-NST condition experienced a similar significant reduction (t (127) = 10.88, p < .0001), with an effect size of d = 1.93. Response rates (i.e., 50% or more reduction in suicide ideation symptoms from baseline) at posttreatment were 69.1% for ABFT versus 62.3% for FE-NST. Similar results were found for depressive symptoms. Both treatments produced substantial reductions in suicidal ideation and depressive symptoms, which were comparable to those reported in other treatment studies of more intensive, longer, multicomponent therapies (I-CBT, Esposito-Smythers et al., 2019; DBT-A, McCauley et al., 2018). The equivalent outcomes may be attributed to common treatment elements, different but unmeasured active mechanisms in the FE-NST condition (e.g., Ibrahim et al., 2021, or regression to the mean.

Secondary Outcomes, Moderators, and Mediators

Because ABFT focuses on family processes as the core mechanism to reduce suicidality, we also examined family-level outcomes, moderators, and mediators. In terms of secondary outcomes, G. S. Diamond et al. (2002) found subjects in ABFT reported significantly greater increases in family cohesion and perceived attachment than those adolescents in treatment-as-usual. In a pilot study of ABFT for lesbian, gay, and bisexual adolescents, the findings revealed decreases in youth-reported maternal attachment anxiety and avoidance (G. M. Diamond et al., 2012). Using data from the G. M. Diamond et al. (2019) study of ABFT and NST, Zisk et al. (2019) examined cooperative communication with caregivers as a moderator of treatment outcomes. The results indicated that youth who had more uncooperative communication with their caregivers at pretreatment had greater reductions in depression by the end of treatment. This finding was present regardless of the treatment condition.

Hospital Aftercare Program Evaluation

Another pilot study focused on patients who had recently made a suicide attempt (G. S. Diamond et al., 2019).We tested ABFT as an aftercare model for 20 adolescents hospitalized following a suicide attempt. We randomized families to 16 weeks of ABFT or EUC. To enhance continuity of care, we met with most cases before being discharged from the hospital. The results showed a nonsignificant trend at preventing future suicide attempts (0% ABFT; 16.7% EUC (1) = 3.60, p = 0.058, Fisher’s exact p = 0.206), reducing attachment-related avoidance for mothers (F(1,9) = 3.85, p = 0.08), and a significant impact on reducing attachment-related anxiety for fathers (F(1,3) = 12.33, p = 0.04).

Effectiveness Research Project

Our first dissemination study explored the feasibility of training therapists in a hospital-based, outpatient mental health clinic in Stavanger, Norway (Israel & Diamond, 2013). Adolescents were included in the study if they scored a 14 or greater on the Hamilton Depression Rating Scale (HAM-D; Hamilton, 1960) and met criteria for major depression based on the Kiddie-Schedule for Affective Disorders and Schizophrenia (K-SADS; Kaufman et al., 1997). We randomly assigned 20 adolescents to 12 weeks of ABFT or treatment as usual (TAU). At posttreatment, adolescents treated with ABFT had significantly lower ratings on the HAM-D (z = −2.05, p = 0.04) and moved from clinical to nonclinical scores on the BDI (χ2 [df1] = 2.88, p = 0.08). The primary benefit from the study, however, was an increased understanding of implementation challenges, as well as how to resolve them. Implementation challenges were also explored in Australia (G. S. Diamond et al., 2016), Belgium (Santens et al., 2016, 2017, 2020), and Sweden (Ringborg, 2016).

Adapting Attachment-Based Family Therapy to Other Populations

Lesbian, Gay, and Bisexual Suicidal and Depressed Adolescents

When working with youth at risk for suicide, there is a disproportionate number of youth self-identified as lesbian, gay, and bisexual (LGB). Often parental rejection of their sexual orientation contributes to the adolescent’s depression or suicidality. In response, Dr. Gary Diamond adapted ABFT to meet the unique needs of suicidal, openly LGB youth and their parents (Diamond G. M., et al., 2012). In phase I of the initial treatment development study, a group of experts met to modify ABFT so it was sensitive to the needs of this population. The treatment was adapted to (a) include more individual time working with caregivers to process their disappointments, pain, anger, and fears related to their adolescent’s minority sexual identity; (b) address the meaning, implications, and process of acceptance for both caregivers and adolescents; and (c) heighten caregivers’ awareness of subtle yet potent invalidating responses to their adolescents’ sexual identity. In a 12-week open trial, adolescents reported significant decreases in suicidal ideation (F(2,14) = 38.16, p = .001, d = 3.76), depressive symptoms (F(2, 14) = 3.78, p = .05, d = .67), maternal attachment–related anxiety (F(2, 8) = 10.89, p = .005, d = 1.25), and avoidance (F(2, 8) = 4.31, p = .05, d = 1.50).

Sexual and Gender Minority Young Adults and Their Nonaccepting Parents

Gary Diamond, working at Ben-Gurion University, focused his efforts on sexual and gender minority (SGM) young adults and their nonaccepting parents (G. M. Diamond et al., 2019). Instead of targeting depression and suicide, the young adults and their parents recruited for this study came because they longed for more meaningful, mutually accepting and respectful, and loving relationships. The treatment focuses on helping caregivers work through their shame, anger, loss, fear, and grief associated with their adult child’s minority identity and, gradually, become more empathic, accepting, validating, and supportive. At the same time, therapists help adult children better articulate their adaptive emotions and unmet attachment and identity needs.

Results from an initial open trial with 31 SGM young adults receiving up to 26 weeks of ABFT-SGM showed that both young adults and their parents independently reported increases in parents’ acceptance of the adult child’s same-sex orientation or noncisgender identity. In addition, young adults reported decreases in parents’ levels of rejection. Additionally, mothers, but not fathers, reported decreases in their own levels of rejection. Finally, young adults reported a decrease in attachment avoidance in their relationships with both mothers and fathers, but not a decrease in attachment anxiety. Importantly, these treatment gains were maintained up to three months after the end of treatment (Diamond G. M., et al., under review). Moreover, reductions in caregivers’ in-session negative behaviors (e.g., criticism and invalidation) over the course of therapy predicted decreases in young adults’ perceived caregiver rejection of their identity. (Boruchovitz-Zamir & Diamond, 2019).

Transgender and Gender Diverse Youth and the Settings That Serve Them

We have begun work to adapt ABFT to meet the unique needs of transgender and gender diverse (TGD) youth and the settings that serve them (Russon et. al., in press, 2012). We conducted a study implementing ABFT for the unique ecologies of LGBTQ+ community settings (e.g., affirmative care centers, youth centers with counseling departments; Russon et al., 2021). LGBTQ+ youth prefer care in community organizations who specialize in their needs. Using the mental health systems ecological model (Schoenwald & Hoagwood, 2001), we worked with stakeholders (i.e., administrators, providers, caregivers, and patients) to generate ABFT delivery modifications that fit the culture of LGTBQ+ community settings (Russon et al., 2021).

We conducted an open pilot study of 10 adolescents identifying as TGD (80%) or sexual and gender minorities (70%). Given the population, we had to pay special attention to gender dysphoria, as well as suicidal ideation (Russon, Smithee, et al., in press). After 16 weeks of ABFT, youth had significant decreases in suicidal ideation (β = −12.16, t (10) = −3.14, p < .01), though decreases in depression were not significant (β = −1.83, t (9.11) = −.88, p = .40).

Young Adults With Unresolved Anger Toward Parents

Young adults often present for therapy with unresolved anger toward their parents. When such anger persists, it can undermine the very fabric of the attachment relationship. In a study with young adults, Gary Diamond et al. (2016) compared 10 weeks of ABFT to individual emotion-focused therapy (EFT; Greenberg, 2010). The results showed that, although both treatments led to significant and equivalent decreases in anger resolution, state anger, attachment anxiety, and psychological symptoms, only ABFT was associated with a decrease in attachment avoidance (G. M. Diamond et al., 2016).

Eating Disorders

Family-based treatment (FBT; Lock & Le Grange, 2013) has proven to be quite effective for anorexia, and some data support its effectiveness with bulimia (Stiles-Shields et al., 2012). The FBT model, however, predicates itself on putting the caregivers in charge of feeding the adolescent back to health. However, when family relationships are dominated by power struggles and interpersonal ruptures, children resist parental guidance (Wagner et al., 2016). This is particularly common in adolescents with bulimia. These adolescents look more like depressed suicidal teens (i.e., emotionally dysregulated, high family conflict, and low parental warmth). Both the Maudsley Group in England (Baudinet et al., 2020) and the Well Center at Drexel University (Manasse et al., 2021) are pilot testing the ABFT model with this population, trying to reduce the focus on control and increase the focus on repairing attachment.

Adherence Research

One important question in both efficacy and effectiveness research is whether therapists are adherent to a model’s theoretical principles and prescribed interventions, as well as whether their interventions are distinguishable from interventions prescribed in comparison treatments. To study this question, we developed the observer rated Therapist Behavior Rating Scale - 3rd version (G. M. Diamond et al., 1996). The results of one adherence study found that ABFT therapists extensively employed essential ABFT interventions and did not use standard CBT interventions (G. M. Diamond et al., 2007).

In a recent study using a global measure rating the essential treatment tasks of ABFT (Ibrahim et al., 2021), we found that higher adherence to ABFT interventions was linked to positive patient outcomes when adolescents reported a high alliance with the therapist. Adherence was not associated with treatment outcomes in individual nondirective therapy. These findings suggest that alliances alone, as important as they are, might not be enough to effect change. In contrast, manualized treatments work better when also attending to the therapist–client relationship.

Process Research

In contrast to outcome research, which asks whether a treatment worked and whether it worked better than waitlist or another active treatment, process research asks how the treatment works. Process research examines what in-session micro-processes (e.g., therapist intervention) promote other essential in-session processes (e.g., client productive emotional processing) thought to be linked to outcome (i.e., process–process research). In other cases, the question is whether the completion of a given process or task leads to better outcomes (i.e., process–outcome research). Process–outcome research is often used to identify the mechanisms through which the treatment works. In this way, process research aims to test the theoretical assumptions that guide clinical models and better understand how to facilitate them.

Improving Parenting

In one qualitative study of interviews with caregivers participating in the initial open trial of ABFT, the researchers examined caregivers’ problem constructions and attitudes toward their adolescents (G. S. Diamond et al., 2003). The results revealed that caregivers were deeply touched when hearing their adolescents’ pain related to feeling disconnected and alone, and how the rupture in their relationship with their caregivers was linked to their depression. Adolescents’ expressions of vulnerability opened caregivers’ hearts and motivated them to commit to repairing the relationship.

A second study examined the immediate effect of therapists’ relational reframe interventions on how parents defined the problem at hand. Relational reframes are interventions designed to temporarily shift caregivers’ attention away from their adolescents’ depression and behavioral problems and toward the rupture in their relationship. Microanalyses of five early therapy sessions with caregivers of depressed adolescents revealed that relational reframe interventions led caregivers to define problems in relational terms in at least two of their six immediately subsequent speech turns. There was partial support for the hypothesis that reframes led to shifts in caregivers’ problem constructions, from intrapersonal to interpersonal. In addition, in good but not poor alliance sessions, caregivers’ interpersonal problem constructions led therapists to use relational reframes (Moran et al., 2005).

A third study examined whether therapists’ relationship-facilitating interventions (e.g., empathy and positive regard) and attachment-oriented interventions (e.g., relational reframes and addressing core relational themes) affected the valence (i.e., positivity–negativity) of caregivers’ attitudes toward their depressed adolescents (Moran & Diamond, 2008). Lag sequential analyses of 13 individual sessions with parents revealed that, in good alliance sessions, both relationship-facilitating and attachment-oriented interventions were associated with caregivers’ nonnegative attitudes toward their adolescent in the five speech turns subsequent to the intervention. No such effects were evident for comparison interventions.

A fourth study examined whether ABFT was associated with decreases in maternal psychological control and increases in maternal psychological autonomy granting. We wanted to explore whether such changes were associated with changes in adolescents’ attachment schema and psychological symptoms. Overall, 18 suicidal adolescents and their mothers received 12 weeks of ABFT. Independent raters observed parents’ in-session behaviors, and adolescents reported perceived maternal care and control, attachment-related anxiety and avoidance, and depressive symptoms at baseline, 6, 12, and 36 weeks. The results indicated that, across the first four sessions, observed maternal psychological control decreased and observed maternal psychological autonomy granting increased. Moreover, increases in mothers’ autonomy granting were associated with decreases in adolescents’ self-reported attachment-related anxiety and avoidance. These adolescent changes were associated with reductions in adolescents’ depressive symptoms at discharge (Shpigel et al., 2012). Thus, this study tested a core premise of the ABFT model, that improving parenting practices would create a family context that would help reduce depression.

Alliance

Another domain of study has been the therapeutic alliance. A strong therapeutic alliance is characterized by a strong bond between each family member and the therapist and agreement on the goals and tasks of treatment (Horvath, 2001). The therapeutic alliance has been found to be a robust predictor of treatment retention and outcome in family-based therapies (Friedlander et al., 2011).

In one study, we dismantled Task III into its component parts then codified if the therapist did that component and how well it went. Then we aimed to determine whether therapist adherence to the model of Task III was associated with an overall increase in alliance and better treatment retention (DeLuca & Diamond, 2014). The results revealed that the quality of the intergenerational component was associated with improvements in the parent–therapist alliance from the first session to the end of the alliance task (p = 0.54). The commitment component was associated with improved alliance at the halfway point (p = .01). In contrast, the preparation component (p = .024) and the total of all task components (p = .040) were positively correlated with treatment retention.

In the second study, we examined whether the strength of the alliance with caregivers, established during individual alliance-building sessions (i.e., Task III), predicted the degree to which caregivers exhibited attachment-promoting behaviors in subsequent conjoint attachment sessions (i.e., Task IV). The results showed that the stronger the caregiver–therapist alliance was during individual sessions with the caregiver, the more caregivers encouraged their adolescents to share emotionally significant events and were empathic and less critical and defensive during conjoint reattachment sessions. However, caregivers’ attachment-promoting behaviors during the attachment task did not predict the posttreatment levels of adolescents’ depressive or suicidal symptoms (Feder & Diamond, 2016).

Emotional Processing

Two studies have examined which therapist interventions promote adolescents’ and young adults’ productive emotional processing in ABFT. Two other studies have examined the link between productive emotional processing and treatment outcome.

One study, conducted by Tsvieli et al. (2020), examined the effect of ABFT therapist interventions aimed at promoting the productive emotional processing of primary adaptive emotions on a sample of 30 depressed and suicidal adolescents who had participated in one of our randomized clinical trials. Sequential analyses revealed that relational reframes and therapists’ focus on primary adaptive emotions were associated with the subsequent initiation of adolescents’ productive emotional processing of primary adaptive emotions. In contrast, interventions not intended to promote productive emotional processing, including interpretations, reassurances, and therapists’ focus on adolescents’ rejecting anger toward their parents, led to the discontinuation of adolescents’ emotional processing, which had already begun. Finally, therapists’ general encouragement of affect and focus on adolescents’ unmet attachment/identity needs were associated with both the initiation of adolescents’ productive emotional processing and the discontinuation of such processing once it had already begun.

In a second study, Tsvieli and Diamond (2018) examined the effects of these same interventions on a sample of 15 young adults receiving ABFT for unresolved anger toward a parent. They also studied individual alliance-building sessions with the young adult. Similar to the findings from the study on adolescents, the results indicated that young adults’ productive emotional processing occurred at a rate significantly greater than chance following therapists’ focus on vulnerable emotions and attachment needs and empty-chair interventions. In contrast, therapists’ focus on clients’ rejecting anger preceded the discontinuation of such processing at rates significantly greater than chance.

A third study, led by Lifshitz et al. (2020), examined the link between emotional processing and treatment outcomes among a sample of 39 suicidal adolescents who had received 16 weeks of ABFT as part of a randomized clinical trial. This study examined purported sequential pathways through which adolescents were thought to traverse as they moved from secondary global distress and rejecting anger to primary adaptive hurt, grief, and assertive anger. Across the entire sample, adolescents indeed moved from states of global distress to maladaptive shame, from maladaptive rejecting anger to adaptive assertive anger, and from adaptive assertive anger to adaptive grief/hurt—though these sequences were not unique to cases with effective treatment outcomes. However, adolescents who did not respond to treatment evidenced higher rates of maladaptive global distress.

In the fourth study, mentioned above in the section “Adapting Attachment-Based Family Therapy to Other Populations,” the researchers explored the link between the amount of young adults’ productive emotional processing and treatment outcomes in ABFT and EFT for young adults presenting with unresolved anger toward parents. Emotional processing was measured during either attachment sessions (in the ABFT condition) or empty-chair interventions (in the EFT condition). The results showed that, whereas the amount of young adults’ productive emotional processing predicted pre- to posttherapy changes in psychological symptoms in both treatments, it did not predict decreases in unresolved anger, state anger, or attachment anxiety and avoidance in either condition (G. M. Diamond et al., 2016).

Attachment Task

The final, and perhaps most critical, process that has been studied is that of the attachment task. The attachment task serves as the primary change mechanism in ABFT. The task involves a series of in-session enactments geared to helping adolescents or young adults communicate their adaptive and often vulnerable emotions associated with relational rupture and unmet attachment and identity needs. In turn, we help caregivers respond in a nondefensive, empathic, and validating manner. This iterative, positive interactional cycle serves to increase trust and security in the attachment relationship (Kobak & Bosmans, 2019) and results in adolescents and young adults feeling understood, worthwhile, connected, and loved, which, in turn, is thought to lead to decreases in depression, suicidality, anger, and anxiety.

In one study, G. S. Diamond and Stern (2003) conducted a task analysis of repairing parent–adolescent attachment in ABFT. One unsuccessful and two successful reattachment episodes were videotaped and coded, using the structural analysis of social behavior model (Benjamin, 1974). The results showed that successful reattachment was associated with high parent affiliation and autonomy. There was also partial support for a phase model of the task, in which successful reattachment episodes were associated with a parent disclosure phase that showed both independence and warmth.

In a recent study, Tsvieli et al. (2021) used THEME analyses to uncover sequences of therapist interventions, young adults’ emotional processing, and parental responses characterizing reattachment episodes in good versus average or poor outcome cases of ABFT for unresolved anger. The results revealed that only in good outcome cases were there four-step sequences hypothesized to facilitate attachment. One sequence began with the therapist focusing on the young adult’s primary adaptive emotions, followed by the young adult’s productive emotional processing of their vulnerable emotions, followed by the therapist empathizing with and validating the parent, followed by parents’ expressions of warmth toward their young adult. The second sequence began with the therapist focusing on the young adult’s unmet attachment needs, followed by the young adult productively processing their vulnerable emotions, followed by the parent expressing a willingness to fulfill their young adult’s attachment needs, followed by parents’ expressions of warmth toward their young adult.

Finally, in a study by Winley et al. (2018), we explored whether caregivers’ empathic validation increased over the course of the attachment task conversation. Decades of developmental research shows that caregivers’ emotional validation provides critical support for healthy adolescent development (Gottman et al., 1996; Kobak et al., 2006). DBT views caregiver validation as a core process, yet studies were unable to show that DBT treatment improved parental validation (Adrian et al., 2018). Using a family interaction-coding tool, Winley showed that caregivers’ use of empathic validation increased over the course of several sessions of the attachment task.

Conclusion

ABFT is an empirically supported treatment designed and developed specifically for repairing attachment ruptures that have damaged trust in the caregiver–adolescent relationship. ABFT focuses on family factors, such as caregiver rejection and criticism, low parental warmth, and adolescent–caregiver conflict, all of which are associated with a host of adolescent behavioral and mental health problems, especially depression and suicide. The model aims to help improve individuals’ (i.e., adolescents and caregivers) emotional regulation and interpersonal problem solving skills, as well as the quality of interacting between people processes. Improving the secure base of attachment helps adolescents improve their emotional regulation skills, a core feature of adolescent depression. ABFT has strong empirical support, demonstrating the efficacy of the model, as well as preliminary data regarding effectiveness. ABFT has also been adapted for several different populations, with good initial results. Additionally, ABFT has a long history of process research that has helped us better understand some of the micro-changes within the therapeutic conversation that contribute to treatment success or failure. This research has set the foundation for our training program, which supports the dissemination of ABFT in 17 countries (http://www.drexel.edu/familyintervention/abft-training-program/overview/). ABFT is a proven treatment for depressed and/or suicidal youth and their families. Additionally, ABFT has relevance for other presenting problems in a variety of clinical settings.

Is ABFT the right treatment for all adolescent struggling with suicide? Yes and no. Some self- harming adolescents come to treatment with family discord as a driving stressor. Here, a family focuse treatment may be indicated. Some self-harming adolescent struggle with stressor outside the home (e.g., bullining) or more endgenious depression, and may need other treatment elements. But, even in this situation, a suicide epidsoede destablizess even the best families. Helping the family recover from this event and helping the caregivers help the adolescent recover from this event seems clinical advantagues. In the modern landscape of empirically spported treatment, not including caregivers in the treatment of adolescent risk for suicide woud be unthinkable (Ougrin & Aswarnow, 2018) if not unethical. Empirical questions remain about the formate, dose, and duration of treatment. ABFT however, has proven effective at reducing suicide ideation and attempts in a low dose, one a week, outpatient therapy (Diamond, et al., 2019).

Highlight.

• Adolescences

• Depression

• Suicide

• Family treatment

• Outcomes research

• Process research