Adult ADHD and trauma: when the treatment misses the architecture

The first reason.
The two architectures look identical from the outside.

The presentations overlap almost completely at the surface. Difficulty sustaining attention. Executive function that drops out at unpredictable intervals. Restlessness that does not yield to exhaustion. A nervous system that needs more stimulation than the room is offering, and that finds workarounds for the deficit through behaviour. Anyone working from a symptom checklist will have a hard time telling the two apart, because the symptom checklist was built to catch the surface pattern, not the architecture producing it.

The architecture matters because it determines which intervention reaches the right layer.

In a primary ADHD presentation, the architecture is dopaminergic. The reward and attention systems calibrate at a setpoint that produces under-arousal in routine contexts. Stimulant medication moves the setpoint. Function improves at the layer where the deficit lives.

In a trauma-driven ADHD-like presentation, the surface looks the same. The architecture underneath is different. The attention dysregulation is downstream of a hypervigilance pattern that has been continuously running since childhood. The restlessness is downstream of a freeze-fight loop that never closed. The executive function dropout is downstream of dissociative gaps that the patient learned to call “distractibility” because that was the available vocabulary.

A patient with this second architecture, on stimulants, often feels significant improvement. The improvement is real. It is also at the wrong layer.

The clinical signal of this misfit is recognisable, with time. The patient does well on a structured task list. They sleep poorly anyway. They concentrate when medicated. They jump at unexpected sounds, freeze in conflict, and find themselves bracing for things they cannot name. The medication is doing its work in the room it was designed for. The other rooms in the house do not warm up.

The second reason.
The microsequence underneath is different.

There is a microsequence that runs in any regulatory loop. Trigger, then tension, then discharge, then relief, then a small new deficit. The shape repeats across substance dependency, behavioural compulsion, and the achievement loop in high-functioning trauma. It also repeats in the trauma-driven ADHD-like presentation, with one variation that matters.

In primary ADHD, the loop sits around cognitive engagement. Trigger is task initiation. Tension is the cost of holding attention. Discharge is novelty seeking. Relief is the brief reward signal from a new stimulus. The deficit is the small post-novelty drop. Stimulants flatten this loop usefully, by raising baseline arousal so the discharge-relief cycle is not the only available regulator.

In the trauma-driven version, the loop sits around physiological activation. Trigger is anything that resembles the archived threat pattern. Tension is the activation that follows. Discharge is whatever the patient has built into their behavioural repertoire to bring the activation down: scrolling, switching tasks, restless movement, micro-rituals. Relief is brief. The deficit is the cumulative physiological cost of running this cycle continuously, year after year, with no exit.

Stimulants reach the cognitive engagement loop. They do not reach the activation loop. A patient on stimulants whose underlying loop is the activation one will report better focus and continued symptoms that look like everything else: poor sleep, body that does not unwind, a sense of running fast without arriving anywhere.

The somatic signature of the activation loop is what often makes the differential possible. The body in primary ADHD is restless in a way that responds to physical movement and to the medication. The body in trauma-driven activation is restless in a way that movement does not discharge and that medication only quiets at the cortex. The patient may not be able to articulate the difference, but they can usually tell, when asked carefully, that the medication helped them concentrate while something else, lower in the body, kept doing what it had been doing for years.

The third reason.
The prediction circuit was not addressed.

The deeper layer is the one already described in the related mechanism literature on this site. The prediction circuit, sitting under the cortical surface, generates the activation that the patient experiences as “I cannot focus” or “I cannot stop moving.” It generates that activation against an archive of patterns that previously preceded threat, and it does this in milliseconds, well before any cognitive content registers (the architecture of this layer is described here in detail).

Stimulants do not introduce a prediction-error signal to this layer. They do not contradict the archived pattern. They do not provide the kind of physiological evidence that would cause the prediction template to update. They raise baseline arousal in the cortical regions that handle attention, which has the side effect of making the underlying activation slightly more tolerable. The patient is, in a precise way, better able to function while the architecture that produces the dysregulation continues to operate, untouched.

This is why the trauma-driven version often becomes more visible after a year or two on medication, not less. The cortical surface is calmer. The signal that lives below the cortex starts to come through more clearly. The patient who arrived saying “I think I have ADHD” returns saying “the medication is working, but something else is happening that was always there underneath.”

What is happening, in those cases, is that the patient is finally able to perceive the activation pattern that the cognitive noise had been masking. The pattern was not produced by the medication. The medication just made the room quiet enough to hear it.

What changes when the audit
names both.

The audit, in this presentation, is not “is it ADHD or is it trauma.” For a substantial number of adults, it is both, and the question of which mechanism is doing what work has to be answered separately. The DSM categories were built to be discrete. The architectures they describe are not. The patient's nervous system runs both loops, and a treatment plan that accounts for only one will produce partial results that look like progress and stall.

Two things, sequenced, do most of the work.

The first is keeping what the medication does well. If stimulants are providing genuine improvement at the cortical layer, they are doing useful work, and there is rarely a reason to remove them prematurely. The patient is functioning better. Functioning better is not the problem.

The second is naming the architecture that the medication does not reach, and treating it on its own terms. This is the work that addresses the prediction circuit and the activation loop, the layer where stimulants were not designed to operate. In a small clinical-phenomenological study (Laugman, 2026), when the trauma component was treated structurally on its own layer, the regulatory profile shifted in a way that medication alone had not produced. Sleep stabilised. The body began to unwind. The “running fast without arriving” sensation reduced. The improvement was not in attention metrics, where the stimulants had already done their work. It was in the layer underneath that those metrics had been hiding.

The work, when it is the right work, names which system is running where, and stops asking either of them to do the other one's job. The medication continues if it is helping. The other architecture gets the conversation it never had. The room that was always cold has, for the first time, a thermostat of its own.