The trauma roots of addiction: why willpower, brain-disease, and bad-company models all miss the mechanism

Addiction as self-regulation.
What it actually regulates.

The pharmacology of addictive substances is well-described. Alcohol depresses central nervous system activity. Opioids occupy mu receptors and produce analgesia and euphoria. Stimulants flood the dopaminergic reward circuit. Gambling triggers the same reward circuitry through behavioral rather than chemical means. Food modulates serotonin, dopamine, and opioid systems through ingestion. All of these are mechanisms of effect. None of them are mechanisms of cause. The question of why a specific person becomes dependent on a specific substance or behavior is upstream of the pharmacology entirely.

The clearest answer to that question is the self-medication hypothesis, formulated by Edward Khantzian in clinical work with substance-dependent patients and articulated most fully in his 1997 reformulation (Khantzian, 1997). The hypothesis is that people use substances to manage internal states they cannot otherwise tolerate. Affect dysregulation, dissociation, hypervigilance, intrusive memories, emotional numbness, none of these are problems addiction creates. They are problems addiction is solving, often without the user having explicit language for what is being solved.

Substance choice tracks the state being regulated. Alcohol and opioids tend to attract people whose default arousal is too high to tolerate: the hyperaroused trauma survivor reaching for the depressant function. Stimulants tend to attract people whose default state is dissociative-flat or depressed: the system that has gone offline reaching for activation. Food tends to function as emotional regulation in clients whose primary symptom is affective flooding. Gambling tends to function as activation against numbness; the addict described as “only feeling alive at the table” is reporting a clinical observation, not a metaphor. Emotion dysregulation has been identified as the central pathway connecting childhood adversity to substance use in formal mediation analyses (Schick et al., 2020). The third step is not the issue. The middle step is.

The pattern is often visible from the first encounter. A client who reports, “I drank for the first time at fourteen and it was the first time I felt normal”, is not describing the discovery of a pleasant intoxicant. They are describing the discovery of a regulatory tool that did, in fact, regulate. The work begins by recognizing what was being regulated.

Trauma as structural cause, not a 'factor'.
The ACE data and what it changes.

The Adverse Childhood Experiences Study, published in 1998, examined the relationship between ten categories of childhood adversity and adult health outcomes in over 17,000 adults (Felitti et al., 1998). The substance-use findings were among the strongest in the dataset. An ACE score of four or higher was associated with a four-fold increase in alcohol abuse, a four-fold increase in self-reported illicit drug use, and a twelve-fold increase in injection drug use, compared to controls with ACE scores of zero. The relationship was dose-response: more adversity, more risk, with no threshold below which the effect disappeared.

The study has been replicated and extended in dozens of populations since, with consistent findings across countries, age cohorts, and substance categories. The ACE score remains one of the strongest single predictors of adult addiction risk that population health research has produced. This is not the language of “trauma as one risk factor among others”. This is the language of trauma as the upstream structural condition that organizes the risk landscape.

The mechanism that links childhood adversity to adult addiction is increasingly well-understood. Chronic stress, particularly stress that occurs during developmental windows of HPA-axis calibration, produces durable dysregulation of cortisol response and dopamine system function (Sinha, 2008). The result is a nervous system that produces too little of its own regulatory chemistry, particularly the chemistry that signals safety, satiation, and reward. The substance then becomes neurobiologically efficient at producing exactly the regulatory effect the dysregulated system cannot produce endogenously. The substance is not creating dependence by acting on an otherwise healthy reward circuit. It is filling a gap that the prior stress dysregulation opened. This is why addiction risk concentrates so heavily in trauma-exposed populations, and why the same substance produces dependence in some users and not others.

Even mainstream addiction neuroscience now incorporates this frame. Volkow and Boyle's 2018 review for the American Journal of Psychiatry positions chronic stress and trauma exposure as upstream factors that modulate the brain's response to substances, not as separate comorbid conditions (Volkow & Boyle, 2018). The shift from “comorbidity” framing to “structural cause” framing is more than terminological. If trauma is a co-occurring condition, addiction treatment can proceed in parallel with trauma treatment, or sequentially after it. If trauma is the structural condition that produced the addiction, addiction treatment that ignores trauma is reproducing the original gap and predicting its own failure.

Why 'just stop' is not the first thing to do.
The modal error of treatment.

The dominant treatment models in addictions remain abstinence-first. The Alcoholics Anonymous tradition, drug-court mandates, most residential treatment programs, and the majority of outpatient addictions counseling all proceed on the assumption that removing the substance is the first task, and that recovery proceeds from abstinence. For a subset of addictions, this works. The peripheral addiction, the one that has not yet acquired a deep self-regulatory function, the situational pattern around a specific environment, often resolves with abstinence and structure.

For the clinically presenting majority, the arithmetic is different. The substance is not the problem. The substance is the solution to a problem. Remove the solution without addressing the problem, and the underlying dysregulation reinstates at full intensity, with no regulatory tool available to manage it. The system has to find a way. The same neurochemistry that previously responded to alcohol begins demanding alcohol again, or rerouting through another substance, or expressing itself as a new compulsive behavior. The published relapse rates are stable across decades: forty to sixty percent in the first year following abstinence-only treatment, higher in the five-year horizon. The industry calls this “the chronic relapsing nature of addiction”. A more accurate description is that it is the predictable consequence of removing the only regulatory tool the system had access to, without replacing the function the tool was performing.

The cross-addiction phenomenon makes the regulatory frame visible directly. A client stops drinking, begins working compulsively, develops an eating disorder eighteen months later, transitions to gambling, finds religion in a way that resembles the structure of the prior compulsions. The substance changes. The function does not.

This is not an argument against abstinence as a clinical goal. It is an argument against abstinence as the first move. Abstinence works when it follows the establishment of alternative regulatory capacity. Imposed in advance of that capacity, it produces predictable failure, and the failure is then attributed to the client's “lack of motivation” rather than to the structural problem with the treatment design.

What changes when treatment addresses the mechanism, not the symptom.
Sequencing trauma-informed addiction work.

Trauma-informed addiction treatment begins from a different premise: the substance is evidence of a dysregulation, not the disease itself. The work is to address the dysregulation, and the substance use typically follows. Sequencing is the central design question, and it is the opposite of the abstinence-first sequence.

The first phase is stabilization. This is not abstinence as such, although for some clients it includes substance management. It is the establishment of basic regulatory capacity through means other than the substance: somatic anchoring, predictable sleep architecture, relational scaffolding, structured affect-tolerance work. The Najavits Seeking Safety protocol (Najavits, 2002) and its derivatives codify this approach with documented outcomes in trauma-comorbid populations: across more than fifteen controlled trials, integrated trauma-and-addiction treatment has produced greater reductions in both PTSD symptoms and substance use than parallel or sequential models that treat the two conditions separately. The effect sizes are not enormous in any single study, but the direction is consistent, and the population in which the effect is strongest is precisely the population most poorly served by abstinence-first care. Stabilization is not glamorous, and the client often does not feel changed during it. The change is structural: the system is being given alternative tools.

The second phase is capacity-building, in which alternative regulatory strategies are practiced until they begin to compete, functionally, with the substance. This is often where substance use begins to reduce without explicit cessation pressure, because the underlying dysregulation is being addressed through other means, and the substance is no longer the only available regulation.

The third phase is trauma reprocessing, which uses EMDR, IFS, structured exposure protocols, or other modalities to reduce the upstream load that is driving the dysregulation in the first place. This is where the gain dial begins to come down, and where the substance often loses both its function and its grip.

In a small clinical-phenomenological study (Laugman, 2026), clients presenting with comorbid addiction and PTSD reported reduction in substance use that tracked alongside reduction in trauma symptoms, not preceding them. The reduction was rarely framed by the clients as “I quit”. It was more often framed as “I noticed I didn't need it the same way”. The function had moved.