If you're a therapist who has just identified a PDA profile in your client and you've typed "how to treat PDA" into a search engine, you've already discovered the problem: there's almost nothing out there telling you what to actually do in session on Monday morning.
That's not because PDA isn't treatable. It's because the PDA field has been stuck at the identification stage for years. Clinicians can now recognize the profile — the pervasive, anxiety-driven avoidance of everyday demands, the sophisticated social strategies, the need for control that goes far beyond typical oppositional behavior. What they can't find is a structured clinical framework that tells them what to do once they've recognized it.
This post is a starting point. Not a complete clinical guide — that's what the RELATE manual is for — but an honest map of what works, what doesn't, and why most of what you were trained to do will make PDA worse before it gets better.
Most therapists encountering PDA for the first time try the tools they already have. This is reasonable and almost always counterproductive. Here's why.
If you're using CBT for the anxiety component, you're treating anxiety organized around feared outcomes — spiders, social rejection, contamination. PDA anxiety isn't organized around outcomes. It's organized around autonomy. The person isn't afraid of what will happen. They're responding to the experience of someone else controlling what happens. Cognitive restructuring doesn't address this because there's no cognitive distortion to restructure. The nervous system is accurately detecting a threat — the threat is just different from what your CBT training prepared you for.
If you're using behavioral approaches — token economies, reinforcement schedules, contingency management — you're adding demands on top of a nervous system that is already overwhelmed by demands. Every new contingency ("If you do X, you earn Y") is processed as another thing someone else is controlling. The behavioral system itself becomes a demand, and the PDA response intensifies. This is why behavioral programs often report that PDA clients "get worse before they get better." They don't get better. They get more overwhelmed, then they shut down or escape the program.
If you're using traditional rapport-building with an implicit clinical agenda — "I'll build this relationship so I can eventually challenge them" — the PDA nervous system will detect the agenda. This is not paranoia. PDA individuals have exquisitely calibrated social detection systems. They can feel the demand embedded in your warmth. The relationship stalls, and the therapist concludes the client is "resistant" or "not ready." The client was ready. The approach wasn't.
PDA operates through a specific nervous system process that, once you understand it, changes everything about how you approach treatment.
Demands — including pleasant ones, self-chosen ones, and ones the person genuinely wants to comply with — are processed by the nervous system as threats to autonomy. This triggers a fight-flight-freeze response that looks like defiance (fight), avoidance or escape (flight), or shutdown (freeze). The critical clinical insight is that this response is happening below conscious control. It is not a choice, a manipulation, or a behavioral pattern that can be extinguished through contingency management.
Two mechanisms drive most of what you're seeing. First, autonomy threat: the nervous system detects that someone else is determining what happens, and responds with a survival-level resistance. Second, intolerance of uncertainty: not knowing what's coming, how long something will take, or what "done" looks like activates the same threat cascade. In most PDA presentations, both mechanisms are running simultaneously.
The clinical question isn't "how do I get this person to comply?" It's "how do I reduce the nervous system threat enough that the person's own capacity comes back online?" This reframe changes everything — from your treatment plan to your session structure to the words you use.
Effective PDA treatment is built on a set of principles that may feel counterintuitive if you're trained in directive therapeutic approaches.
Start with yourself, not the client. Your own regulation state is a clinical variable. If you're anxious about the session going well, frustrated that the client isn't engaging, or subtly pushing an agenda, the PDA nervous system reads all of it. Before you can create safety for the client, you need to be genuinely regulated yourself. This isn't a wellness platitude — it's a prerequisite for the intervention to work.
Reduce demands systematically, not intuitively. Most therapists think they've lowered demands when they've lowered obvious ones — "I won't make them do worksheets." But demands include the implicit expectation to make eye contact, sit in a specific chair, talk about feelings, stay for the full session, show progress, and demonstrate engagement. A genuine demand reduction means auditing every expectation you're carrying into the room — including the ones you haven't consciously identified — and releasing the ones that aren't essential.
Prioritize the relationship over the treatment plan. In PDA treatment, the therapeutic alliance isn't a prerequisite for the real work. It is the real work. The client's nervous system needs to experience a relationship where demands are genuinely low, autonomy is genuinely respected, and ruptures are genuinely repaired. This relational safety is what eventually widens the window of tolerance and makes other interventions accessible.
Follow the client's nervous system timeline, not yours. Progress in PDA treatment doesn't look like linear improvement on standardized measures. It looks like the window of tolerance getting slightly wider, recovery from dysregulation getting slightly faster, and the client tolerating slightly more uncertainty without activating. These changes happen on the nervous system's schedule, which is almost always slower than the therapist's preference or the institution's expectations.
A PDA-informed therapy session looks different from what most training programs prepare you for. You might spend sessions following the client's lead completely, with no agenda. You might sit in silence. You might talk about their special interest for 40 minutes because that's where their nervous system is regulated, and regulated is where you need them to be before anything else is possible.
When the client is dysregulated, you're not intervening. You're reducing — less talking, more space, fewer questions, explicit release of any expectations for the session. "We don't have to do anything today" is sometimes the most therapeutic thing you can say, and you have to mean it.
When the client is regulated, you're not rushing to capitalize on it. You're testing — the smallest possible next step, with genuine willingness to withdraw if it increases threat. "I had a thought about something we could try — but only if it sounds interesting to you, and we can drop it immediately" is the kind of language that respects the mechanism.
The principles above are necessary but not sufficient. Knowing that you should reduce demands and follow the nervous system is helpful. Having a structured framework that tells you how to assess, implement, troubleshoot, and track outcomes — with named protocols and specific clinical tools — is what turns good instincts into reliable clinical practice.
That's what we built RELATE to be: a framework where every strategy traces back to the nervous system mechanism it's addressing, with assessment instruments that capture the specific PDA profile of each individual, an in-the-moment triage protocol for crisis, and implementation phases that sequence the work appropriately.
If you're a therapist trying to figure out how to treat PDA, you're not alone, and the fact that your training didn't prepare you for this isn't a personal failing. It's a gap in the field. We built RELATE to close it.
The RELATE Foundational Training Manual gives you named protocols, assessment instruments, and implementation phases — everything your training program didn't cover.