In my previous piece, I wrote about what happens when a child can no longer attend school. I described the patterns that emerged from my research with painful consistency, distress misread as defiance, parents who raised concerns early and were not listened to, support that arrived too late or not at all, and a system that looked at these children and saw a behaviour problem, when what was actually present was a child in crisis.

Thankfully, that is not where every story ended…

Within the same research, a qualitative study drawing on the experiences of 62 families of autistic children with a PDA profile, all of whom had missed school for six months or more, a second pattern frequently emerged. One that looked entirely different. These were the accounts of what happened when something around the child began to change. Not a new policy, not a revised programme. Something more fundamental: a change in how the adults around the child understood what they were seeing, and what they chose to do about it.

Before I share what worked, it’s worth revisiting the scale of what these families had been through, because the contrast matters.

Eighty-one per cent of the children in the study had become unable to attend school during primary school. Nearly a third had been out of education for over three years. In 68% of cases, it was the parent who first recognised that their child might have a PDA profile, not a teacher, not a SENCO, not a healthcare professional. Parents who raised those concerns early, often repeatedly, reported that they were dismissed, blamed, or simply not believed. One parent described professionals offering “dismissive responses to parental advocacy, despite clear evidence of deterioration.” Another said, “If they’d only just listened to us when we first raised concerns, none of this would have had to happen.” By the time many of these children reached the turning point, the moment where something began to shift, they had already spent months, sometimes years, in a state of profound and sustained distress. Their nervous systems, shaped by repeated experiences of being misunderstood and overwhelmed, had adapted to expect threat. Some children had developed significant mental health conditions. Twenty-five of the 62 families reported a co-occurring mental health diagnosis. Eighteen children were, at the point of the study, unable to engage in any form of education at all.

This is the context within which change eventually happened. It matters because it tells us that what made a difference was not a minor adjustment. It was a fundamental shift in approach, and the families who experienced it knew the difference.

The first thing that changed was not attendance. It was safety.

This came through with striking consistency across the data. When the sense of threat around a child reduced, even slightly, even imperfectly, something shifted. A child who could leave the house again. A child who could tolerate a small request. A child who could begin, tentatively, to re-engage. As one parent described it: “I’ve finally found a setting who understands that psychological safety takes months to create.”

For children with a demand-avoidant profile, this language of safety is not metaphorical; it is neurological. Research on the polyvagal system helps us understand that felt safety is fundamental to a child’s capacity for social engagement and emotional regulation. When a child’s nervous system is locked in a state of threat detection, compliance-based approaches, the rewards and punishments that characterise much of current mainstream educational practice, do not reduce that threat response. They compound it.

Relationships were at the heart of how safety developed. Not relationships in the abstract, but specific adults who brought particular qualities to every interaction. Parents described professionals who were calm and unhurried, who were predictable without being rigid. Adults who did not escalate when things became difficult, but instead stepped back, reduced demands, and stayed present. Adults who understood, as one parent put it, that once a trusting relationship has been built, my child doesn’t perceive the person as a threat to his autonomy.

The data was consistent and clear on this point. One parent described the qualities their child needed in an adult: calm, patient, not remotely patronising. Treating her like an equal, using humour. Another described the relief of finding provision built on flexibility, predictability, and a relational approach. A third parent, whose child had spent years in environments that did not work, said something that the wider data consistently reflected: The biggest part of this, she has a connection with the staff. Connection is EVERYTHING.

Time and time again, parents described professionals who had moved away from the idea that behaviour was something to be corrected, and towards an understanding that behaviour was communication, a signal from a child whose needs were not being met. These were professionals trained in neurodivergence, familiar with the specific demand-avoidant profile, and willing to adapt their own approach rather than waiting for the child to comply. One parent described their child’s experience with a specialist provider: “They have a complete understanding of the demand avoidant profile and have worked so gently to support him, offering just enough encouragement to push him forwards, whilst never pushing him too hard”. Another noted: “Staff experienced in neurodivergence has so far enabled her to mask less and begin to re-engage with education.” The contrast with earlier experiences was, for many families, stark and painful to sit with. These same children who had been described as defiant, non-compliant, refusing, began to engage, to trust, to learn, when the adults around them were equipped to understand them.

Central to all of this was autonomy. It is worth being precise about what that means here, because the nuance matters.

Parents were clear that superficial or token autonomy made no difference. A choice between two pre-approved activities. A decision about whether to sit in the classroom or the corridor. These gestures did not shift anything. What their children needed was genuine autonomy, the real ability to say no, to direct their own learning, to decide how, when, and whether they engaged. One parent wrote: “genuine autonomy, false autonomy, such as a choice between activities, are not enough.” Another described their child’s need for complete autonomy over how and when she engages with learning. One parent was direct about the implications: “She needs total autonomy, which obviously can’t happen in school.”

That last line contains a great deal. For children with a demand-avoidant profile, the need for autonomy is not a preference or a parenting choice; it is a well-documented need, directly linked to their nervous system response to perceived threat. The research was unambiguous: when autonomy was respected, their sense of threat reduced, and children could begin to engage. When it was restricted, distress escalated rapidly. The systems these children are asked to navigate, built on structure, routine, and compliance, are, by design, in direct tension with what these children need to be safe. This is frequently where placements break down. Not because the child cannot learn, but because the environment cannot adapt to meet them.

What the research showed, clearly and repeatedly, was that when children were given safety, genuine relationships, and real autonomy, learning did not disappear; it merely found a different shape.

Many children began to re-engage with education outside of formal school settings, through alternative provision and flexible, interest-led approaches. One parent described the approach that worked: “Following the child’s lead. She decides what provision she wants to do more or less of. We follow her, in collaboration with he”. Another wrote about their child’s capacity within this kind of learning: “EXCEPTIONAL ability to learn in this way”. A further parent, reflecting on a sustained period of low demand, reported: “with low to zero demand for over two and a half years, my child’s mental health has recovered 100%.”

These outcomes were not accidental. They were the result of deliberate, informed choices, made most often by professionals and settings who listened and understood not just autism broadly, but the specific needs of a demand-avoidant profile. The study found that the focus of effective provision was consistently on connection rather than compliance, delivered by adults who understood the child’s individual neurodevelopmental profile and were willing to put the relationship before the curriculum.

That is not how most of these children’s journeys had begun. In 68% of cases, parents recognised the profile before any professional did. Most of those parents reported raising concerns that were dismissed. Many watched their children deteriorate, some to the point of severe mental health crisis, before any of these approaches that work were implemented.

If 81% of children are unable to attend school during primary years, and they are labelled as non-compliant. If parents are consistently the first to identify the profile, often years before professionals act, then the argument for earlier, more responsive support is not simply compassionate. It is urgent and evidenced. For many of the families in this research, the crisis was not inevitable. It was clearly the product of a system that did not listen, did not adapt, and did not equip its professionals with the understanding they needed.

When professionals were trained, when settings were flexible, and when relationships came before expectations, children engaged. Some, who had not accessed any form of education for years, began learning again. Not because they were made to. Because the conditions finally existed in which they could.

Behind every percentage in this research is a child. A child who was told, in one way or another, that the problem was them. This research documented something different: when the approach changes, the child’s experience changes with it. They were never unreachable. They were never unwilling. They were simply waiting for someone to stop trying to change them and start changing the conditions around them instead.