One of the more counterintuitive aspects of harmful relationships is that the people inside them are often intensely attached — not despite what’s happening, but in ways that seem bound up with it. The hurt doesn’t produce distance. It can produce a pull. The prospect of permanent separation may feel more like grief than relief.
“Trauma bonding” is the term most widely used to describe this experience. It’s worth being careful about the term, because it carries specific meaning in some contexts and is used more loosely in others.
Where the concept comes from
Patrick Carnes introduced the term “betrayal bond” (later often called trauma bonding) in his 1997 book The Betrayal Bond, drawing on clinical work with people in exploitative or abusive relationships. Carnes described how people in these situations could develop strong attachment to the person harming them — a pattern he observed particularly in situations involving cycles of threat and intermittent kindness or relief.
Judith Herman’s Trauma and Recovery (1992) is frequently cited alongside this concept. Herman documented complex psychological responses in people held in captivity or sustained abuse situations, including what she called complex PTSD. Her work does not use the term “trauma bond” itself, but her clinical descriptions of attachment in captivity contexts are often referenced in discussions of this phenomenon.
It’s worth being clear: “trauma bond” is not a DSM diagnosis. It does not appear in the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders as a formal category. It functions as a descriptive concept in survivor education and some clinical contexts, not as an established diagnostic entity.
What the concept describes
The practical observation behind the term is real and clinically recognized in some form: people in sustained abusive or exploitative relationships can develop strong attachments that persist despite harm, and may feel significant distress at the prospect of leaving — including pull toward the person who has hurt them.
A range of mechanisms may contribute to this. The intermittent reinforcement literature from behavioral psychology offers one possible explanation (discussed in a separate article on this site). Social isolation from protective relationships is another factor well-documented in the coercive control literature. Fear of what leaving will bring is a practical consideration that gets underestimated by outside observers.
None of this has a single clean mechanistic explanation that has been experimentally established in humans in real-world relationships.
What the neurochemistry claim gets wrong
You’ll frequently encounter the claim that trauma bonding creates a “neurochemical addiction” — that stress hormones, oxytocin, and dopamine produce something neurologically identical to drug dependency. This framing appears in popular self-help content but significantly overstates what the research supports.
There is relevant neuroscience on bonding, reward, and stress — but the specific claim that attachment in abusive relationships is neurochemically identical to drug addiction has not been established by the current evidence. The experience may feel that way. The mechanism has not been demonstrated at that level of specificity.
Framing it as neurological addiction can also discourage people from seeking change by suggesting they’re powerless until something resets their brain chemistry. That overstates both the constraint and what would be required to leave.
What tends to support recovery
Research on recovery from coercive and abusive relationships — including Herman’s work on complex trauma — points toward several consistent factors: physical distance from the harmful relationship when this is safe and possible, reconnection with people and support networks outside it, and professional support from therapists experienced in trauma and coercive dynamics.
None of these work on a fixed timeline. Recovery from sustained relational harm is typically not a clean, linear process.
If you’re finding the pull to return to a harmful relationship overwhelming, support from a trained therapist or domestic violence advocate is appropriate — not a sign that something is more seriously wrong.
- National Domestic Violence Hotline (US): 1-800-799-7233 | https://www.thehotline.org/resources/trauma-bonds-what-are-they-and-how-can-we-overcome-them/
- Outside the US, contact a local domestic violence or family violence organization for confidential support.
Sources and Further Reading
- Patrick Carnes. The Betrayal Bond: Breaking Free of Exploitive Relationships. Health Communications Inc., 1997. Practitioner book introducing the betrayal bond/trauma bonding concept for a general audience; not peer-reviewed research.
- Judith Herman. Trauma and Recovery. Basic Books, 1992. Foundational clinical work on complex PTSD and responses to sustained captivity and abuse; does not use the term “trauma bond” but addresses related clinical observations.
- American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5). American Psychiatric Publishing, 2013. Establishes that “trauma bond” is not a formal DSM diagnosis.
- National Domestic Violence Hotline. “Trauma Bonds: What Are They and How Can We Overcome Them?” https://www.thehotline.org/resources/trauma-bonds-what-are-they-and-how-can-we-overcome-them/
This article is for educational purposes and does not provide medical, mental-health or legal advice. Only a qualified professional can diagnose a mental-health condition.