Emotional Attachments in Abusive Relationships: A Test of Traumatic Bonding Theory

What This Research Found

Donald Dutton and Susan Painter’s 1993 study represents the first empirical test of traumatic bonding theory, providing scientific explanation for one of domestic violence’s most troubling puzzles: why victims often remain deeply attached to their abusers. Published in Violence and Victims, this research has become foundational for understanding abuse dynamics, cited extensively in clinical literature, legal proceedings, and survivor resources.

The two necessary conditions for traumatic bonding: Dutton and Painter identified two specific factors that, when combined, create powerful emotional bonds between abuse victims and their perpetrators. First, there must be a significant power imbalance—one partner dominates while the other is subordinate, creating dependency. Second, the abuse must be intermittent rather than constant, punctuated by periods of kindness, affection, or normalcy. Neither condition alone produces traumatic bonding; it is specifically their combination that hijacks normal attachment mechanisms.

The paradox of intermittent reinforcement: Perhaps the study’s most counterintuitive finding is that constant abuse does not create strong attachment bonds—intermittent abuse does. When abuse alternates with warmth, the victim’s dopamine system responds to unpredictable reward with heightened activation, similar to gambling’s pull. The “good periods” become desperately anticipated, and their occasional occurrence reinforces hope despite mounting evidence of harm. This intermittent reinforcement pattern creates attachment stronger than consistent positive treatment would produce.

The neurobiological mechanism: While Dutton and Painter’s study was primarily psychological, subsequent research has illuminated the neurobiological underpinnings they described. The cycle of fear and relief, cruelty and kindness, creates neurochemical patterns resembling addiction. Stress hormones during abuse are followed by dopamine and oxytocin during reconciliation, creating a biochemical rollercoaster that progressively rewires the brain’s reward system and attachment circuits. The victim becomes neurologically dependent on the abuser, experiencing genuine withdrawal symptoms during separation.

Relationship duration intensifies the bond: The study found that longer abusive relationships produced stronger traumatic bonds, independent of abuse severity. Each cycle of abuse and reconciliation reinforced the attachment pattern. Women who had been in relationships for longer periods reported stronger attachment to their abusers, even when they had successfully left. This finding explains why victims who return after leaving often report each subsequent departure becoming more difficult—the bond strengthens with each cycle.

How This Research Is Used in the Book

Dutton and Painter’s traumatic bonding research appears throughout Narcissus and the Child as the primary framework for understanding why victims remain attached to narcissistic abusers. The research is cited in five chapters, explaining attachment patterns in romantic relationships, parent-child dynamics, and the neurological mechanisms underlying abuse’s addictive pull.

In Chapter 20: Field Guide to Narcissistic Situations, Dutton and Painter’s research provides the scientific foundation for understanding why survivors struggle to leave:

“Psychologists Donald Dutton and Susan Painter identified two necessary conditions for traumatic bonding: power imbalance and intermittent reinforcement. The victim must be in a lower-power position, and the abuse must be punctuated by periods of kindness, affection, or normalcy. Constant abuse does not create the same bond; it is the alternation that hooks the brain.”

The book uses this insight to address survivors’ most painful question—“Why did I stay?”—with neurobiological explanation rather than blame. The same chapter details how dopamine, cortisol, and oxytocin create addiction-like attachment through the unpredictable cycles narcissists naturally produce.

In Chapter 3: The Anxious Sibling—Borderline Personality, the cycling between idealisation and devaluation is explicitly linked to trauma bonding:

“This cycling between idealisation and devaluation creates trauma bonding through intermittent reinforcement of reward and punishment—a mechanism similar to Stockholm Syndrome.”

Chapter 11: Neurological Contagion uses Dutton and Painter’s work to explain why unpredictable abuse is more neurologically devastating than predictable stress:

“If chronic stress alone were the issue, victims might adapt. Humans are highly resilient when they can predict and prepare for challenges. But unpredictability characterises narcissistic abuse—the intermittent reinforcement pattern that makes it neurologically devastating.”

Chapter 13: The Great Accelerant extends Dutton and Painter’s insight to explain how social media exploits identical mechanisms:

“Intermittent reinforcement—the unpredictable alternation of reward and punishment—creates trauma bonds in abusive relationships by exploiting dopaminergic reward prediction error. Social media platforms employ the identical mechanism: variable reward schedules explicitly designed to maximise engagement by exploiting the same neural circuitry.”

Chapter 16: The Gaslit Self applies the research to explain why victims struggle to leave gaslighting relationships, with the intermittent reinforcement of “good days” making victims question whether abuse is real.

Why This Matters for Survivors

If you have experienced narcissistic abuse, Dutton and Painter’s research offers something essential: scientific explanation for experiences that likely left you doubting your own judgment and strength.

Your attachment makes neurobiological sense. The pull you feel toward your abuser—the one who hurt you—isn’t weakness, stupidity, or masochism. Dutton and Painter demonstrated that specific abuse patterns create specific attachment responses. When you experienced the narcissist’s cycle of love bombing and devaluation, your brain responded exactly as human brains are designed to respond to intermittent reinforcement. The same mechanisms that help us learn from unpredictable environments can be exploited to create pathological attachment. Your bond to your abuser reflects neurobiology, not character.

The “good times” weren’t an aberration—they were the mechanism. You may have stayed because of the good periods, hoping they would return, believing the “real” relationship existed in those moments. Dutton and Painter’s research reveals that those good periods weren’t evidence that the relationship could work—they were precisely what made the abuse so binding. Without the intermittent kindness, you would have left sooner. The hoovering, the apologies, the glimpses of the person you fell in love with—these created the intermittent reinforcement pattern that trapped you. This isn’t your failure to see clearly; it’s how trauma bonding works.

Your difficulty leaving proves the severity, not your inadequacy. People who haven’t experienced trauma bonding often wonder why victims don’t “just leave.” Dutton and Painter’s research shows that the difficulty of leaving reflects the abuse’s severity and duration, not the victim’s weakness. Longer relationships and more severe abuse cycles create stronger bonds. If you found leaving nearly impossible, it’s because the abuse was effective at creating the conditions for traumatic bonding. Your struggle to leave is evidence of what was done to you, not a character flaw.

Understanding the mechanism supports healing. Knowing that your attachment has identifiable causes and neurobiological substrates changes the recovery process. You’re not trying to overcome some mysterious personal failing—you’re addressing a specific pattern that can be understood, named, and gradually unwound. The bond will weaken with time and no contact, as your neurochemistry recalibrates to the absence of the abuse cycle. Recovery is possible precisely because the bond was created by specific conditions that no longer apply. Many survivors ultimately develop Complex PTSD from prolonged trauma bonding, which requires specialised treatment approaches.

Clinical Implications

For psychiatrists, psychologists, and trauma-informed healthcare providers, Dutton and Painter’s research has direct implications for assessment, treatment planning, and therapeutic approach with abuse survivors.

Recognise trauma bonding as distinct from healthy attachment. Clinicians must understand that the attachment abuse survivors describe is qualitatively different from secure or even anxious attachment. Trauma bonding is created by specific conditions (power imbalance plus intermittent abuse) and maintained by neurobiological mechanisms resembling addiction. Assessing for these conditions helps identify trauma-bonded patients who may require different treatment approaches than those with standard relationship difficulties.

Avoid blame while maintaining therapeutic honesty. Dutton and Painter’s research supports validating survivors’ experiences without collusion. The attachment is real and has neurobiological basis—survivors aren’t imagining their pull toward abusers. But validation shouldn’t become enabling of return to abuse. Clinicians can acknowledge the bond’s power while helping patients understand its origin and develop strategies for managing urges to return. The goal is informed agency, not judgment-free return to danger.

Prepare patients for separation as withdrawal. When survivors are ready to leave abusive relationships, clinicians should prepare them for withdrawal-like experiences: intense cravings for contact, physical discomfort, ruminative thinking about the abuser, and powerful urges to return. These symptoms are predictable based on Dutton and Painter’s model and reflect neurobiological recalibration, not genuine love or evidence that the relationship was good. Patients need coping strategies similar to those used in addiction treatment: identifying triggers, developing alternative responses, building support systems, and expecting non-linear progress. No-contact or minimal contact accelerates the bond-breaking process significantly.

Address the abuse cycle’s effect on attachment patterns. Survivors of trauma-bonded relationships often develop disorganised attachment patterns that affect subsequent relationships. They may be hypervigilant for abuse signs, alternatively drawn to and repelled by intimacy, or struggle to recognise healthy relationship dynamics. Treatment should address not only the index relationship but the attachment templates it created. Healthy relationships may initially feel “boring” or “wrong” to someone whose attachment system was calibrated to trauma bonding’s intensity.

Consider treatment intensity matching abuse duration. Dutton and Painter found that longer relationships created stronger bonds. This suggests that recovery duration may need to match abuse duration—brief interventions are unlikely to address bonds formed over years of cycling abuse. Clinicians should set realistic expectations and advocate for treatment intensity that matches the clinical picture. Trauma bonding created by a multi-year abusive marriage will likely require longer treatment than a brief abusive relationship.

Broader Implications

Dutton and Painter’s traumatic bonding theory extends beyond individual therapy to illuminate patterns across families, institutions, and society.

The Intergenerational Transmission of Abuse Vulnerability

Children who witness parental trauma bonding learn that intermittent abuse and intense attachment naturally co-occur. Their internal model of relationships includes the expectation that love comes with pain, that intensity indicates depth, and that unpredictability is normal. This template makes them vulnerable to trauma bonding in their own adult relationships—not through genetic transmission but through learned expectations about what relationships feel like. Intergenerational trauma includes the transmission of relationship templates that predispose children to abuse vulnerability. These patterns often manifest as anxious attachment or disorganised attachment styles that persist into adulthood.

Relationship Patterns in Adulthood

Dutton and Painter’s framework explains why abuse survivors often enter subsequent abusive relationships. The trauma-bonded individual has a neurochemistry calibrated to the abuse cycle—stable, consistent relationships may feel flat or wrong. They may unconsciously seek partners who provide the intermittent reinforcement their system was trained on. Recovery requires not just leaving one abuser but addressing the attachment patterns that make trauma bonding likely. Without this deeper work, survivors remain vulnerable to repeating the pattern.

Workplace and Organisational Dynamics

The same mechanisms that create trauma bonding in intimate relationships can operate in workplace contexts. An abusive boss who alternates between praise and humiliation, warmth and coldness, creates conditions for organisational trauma bonding. Employees may become intensely loyal to abusive supervisors, working excessive hours to earn approval while suffering significant harm. Understanding trauma bonding helps explain why employees stay in toxic work environments and why leaving can feel as difficult as ending an abusive romantic relationship.

Legal and Policy Considerations

Dutton and Painter’s research has significant implications for how legal systems treat domestic violence cases. Judges and juries who don’t understand trauma bonding may discount victim testimony when survivors express continued attachment to abusers, recant accusations, or return after separation. Expert testimony about traumatic bonding can help courts understand that these behaviours are consistent with abuse, not evidence against it. Policy approaches that require victims to immediately and permanently separate from abusers without addressing the neurobiological bond are likely to fail.

Cult and Coercive Group Dynamics

The traumatic bonding model applies beyond intimate relationships to cult dynamics and coercive organisations. Charismatic leaders who establish power imbalance and employ intermittent reinforcement—alternating between love and punishment, approval and shaming—create trauma bonds with followers through mechanisms of coercive control. This explains why cult members often remain attached to leaders despite objective evidence of harm and why exit is so difficult. Dutton and Painter’s framework informs cult recovery work and prevention education.

Digital and Online Abuse Patterns

As Chapter 13 of Narcissus and the Child explores, social media platforms employ intermittent reinforcement patterns similar to those Dutton and Painter identified in abuse relationships. The unpredictable nature of likes, comments, and notifications creates dopaminergic responses analogous to trauma bonding. Online abusers can exploit these mechanisms through cycles of attention and withdrawal, creating powerful attachments despite never meeting victims in person. The traumatic bonding framework helps explain online grooming, catfishing, and digital abuse dynamics.

Limitations and Considerations

Dutton and Painter’s influential work has important limitations that inform how we apply it.

Sample characteristics limit generalisability. The study used a sample of 75 women who had left abusive relationships, recruited through shelters and support services. This sample may not represent all abuse victims—those who never leave, those who don’t access services, male victims, and victims in same-sex relationships may show different patterns. Subsequent research has generally supported the findings across populations, but caution is warranted in individual application.

Correlation does not establish causation. While the study found significant relationships between abuse characteristics and attachment strength, it cannot definitively establish that power imbalance and intermittent abuse cause traumatic bonding. Other factors—personality characteristics, childhood attachment history, relationship investment—may contribute to both abuse patterns and attachment strength. The mechanism is theoretically compelling but empirically correlational.

Individual variation is substantial. Not all abuse victims develop trauma bonds, even when power imbalance and intermittent abuse are present. Individual differences in attachment style, neurobiological factors, support availability, and other variables moderate the effect. Clinicians should assess each case individually rather than assuming trauma bonding exists whenever abuse characteristics are present.

The framework may oversimplify complex dynamics. While power imbalance and intermittent reinforcement capture essential elements, real abusive relationships involve many other factors: financial entanglement, children, social networks, cultural context, immigration status, and more. The two-factor model is parsimonious but may miss important nuances in individual cases.

Historical Context

Dutton and Painter’s 1993 study emerged from clinical observations that had puzzled domestic violence researchers and practitioners for decades. Victims’ attachment to their abusers seemed paradoxical—why would someone love someone who hurt them? Earlier frameworks borrowed from hostage research (Stockholm Syndrome) and learned helplessness theory provided partial explanations, but lacked empirical testing in domestic violence contexts.

Dutton had been researching domestic violence dynamics since the 1970s, contributing to understanding of batterer psychology and abuse cycles. Painter brought methodological expertise to the collaboration. Together, they designed a study to test whether the conditions observed to create attachment in hostage situations—power imbalance and intermittent good treatment—also explained attachment in domestic violence.

The study’s publication in Violence and Victims gave it visibility in the domestic violence research community, and its clear conceptual framework made it accessible to clinicians and advocates. The term “traumatic bonding” entered professional vocabulary and gradually spread to survivor communities.

Subsequent neuroimaging and physiological research has largely supported Dutton and Painter’s psychological model, showing that the abuse cycle produces measurable neurobiological changes consistent with the attachment patterns they described. The concept has been incorporated into clinical training, legal expert testimony, and survivor education materials worldwide.

Dutton’s later work expanded understanding of batterer psychology, while the traumatic bonding framework has been applied to child abuse, elder abuse, cult dynamics, and human trafficking. The 1993 study remains foundational—a clear articulation of a mechanism that, once named, helped millions of survivors understand their experiences.

Further Reading

• Dutton, D.G. (1995). The Batterer: A Psychological Profile. Basic Books.
• Dutton, D.G. (2007). The Abusive Personality: Violence and Control in Intimate Relationships (2nd ed.). Guilford Press.
• Herman, J.L. (1992). Trauma and Recovery: The Aftermath of Violence—From Domestic Abuse to Political Terror. Basic Books.
• Walker, L.E. (1979). The Battered Woman. Harper & Row. [Introduced the cycle of violence concept]
• van der Kolk, B.A. (2014). The Body Keeps the Score: Brain, Mind, and Body in the Healing of Trauma. Viking.
• Carnes, P.J. (1997). The Betrayal Bond: Breaking Free of Exploitive Relationships. Health Communications.