Outside view
Why I never feel sure when I look after him
The grandson stays overnight, and his grandmother lies awake wondering whether she did everything right. It was never like that with her own children. Many grandparents of autistic grandchildren know this tension and take it as a sign that they are doing something wrong.
This article explains where the uncertainty comes from. It shows three typical moments and what lies behind them neurologically. It gives no care instructions and judges no one.
When old knowledge no longer fits
The first question comes from a grandmother who never felt unsure with her own children. The answer explains why that earlier knowledge does not apply here.
The following answers were generated in the Autistic Mirror app in German and are shown here as a faithful English translation, without edits to content.
User question
When my grandson stays overnight with me, I am tense the whole time about whether I am doing everything right. It was never like that with my own children. Why do I feel so unsure when I look after him?
Autistic Mirror (Outside view)
This uncertainty does not come from incompetence but from a changed information situation.
With your own children, an inner model was built up over years: what calms this child, what triggers stress, when hunger is the real reason, when exhaustion. This knowledge came through thousands of repetitions, and it was implicit. You did not have to think, you simply knew. The brain then runs in so-called predictive coding. It constantly compares expectations with what actually happens. The more reliable the inner model, the more quietly this comparison system runs in the background.
With an autistic grandchild, the old model often does not apply. The signals you learned to read as a parent show up differently or not at all. A child who is quiet could be relaxed or just before a breakdown. A child who refuses to eat may simply not feel like it or may be reacting to a texture that is genuinely unbearable for the child. A child who needs what it always needs at bedtime is not deviating from it out of defiance. The prediction model that worked well for decades does not fit this child, and the brain registers this gap as uncertainty. That is intolerance of uncertainty, more precisely: a system that currently has to recalibrate more than usual because the environment contains more open spots than previous knowledge covers.
So the tension is not a sign that something is going wrong. It is the honest signal of a model that is not finished yet. And this model gets better over time, but through concrete observation of this child, not through general principles. What calms exactly this grandson, what helps with transitions, what he eats and what not, what the bedtime ritual is in the form the child knows it. This information is best gathered once, systematically, from the parents before the overnight stay begins, not as a check, but as a handover of the inner model the parents already have.
When the same situation ends differently
The second question is about reactions that cannot be predicted. The answer describes why the visible trigger is rarely the whole story.
User question
My grandson sometimes reacts very calmly and sometimes very intensely to things that look the same to me. Why can I predict his reactions so poorly?
Autistic Mirror (Outside view)
His nervous system evaluates every situation not only on the basis of what is happening right now, but on the basis of a running budget. That is predictive coding. The brain constantly compares what it expects with what actually arrives and regulates the reaction accordingly. In autistic people, this comparison runs with much less dampening of incoming stimuli. Every deviation from expectation, every unexpected sound, every touch, every changed mood in the room produces a stronger signal than in a neurotypical child. That costs energy.
What this means for predictability: the same trigger, for example a radio that is too loud, can produce no reaction at all in one moment and trigger a meltdown an hour later, because the available regulation budget has emptied in the meantime. You only ever see the trigger, not the balance before it.
A school day uses up considerably more of it than a free Saturday. Fluorescent light, the noise of the class, social routines he has to steer consciously, not fidgeting, not being too loud, saying the right things: none of this runs automatically, it costs him active attention. Masking is the name for this process, the active adjustment of one's own behaviour to neurotypical expectations. What he spends of it at school is no longer available for the afternoon.
This means: the moment in which you see him is often not the moment in which the actual load arose. What looks like a disproportionate reaction to a small thing is often the last layer on a full stack. That does not make his reactions illogical, they simply follow an inner bookkeeping that you cannot see from the outside.
His behaviour becomes more predictable through times of day and contexts than through single triggers: after school the budget is smaller than in the morning, after unexpected changes in routine smaller than after a predictable day, in a familiar environment larger than in a new one.
When the follow-up feels like a test
The third question is about the handover. The answer explains why detailed questions from the parents do not mean the same thing as mistrust.
User question
When I bring him back, my daughter always asks exactly how the day went, and I feel like I am being tested. What is happening for her and for me?
Autistic Mirror (Outside view)
Your daughter asks about the concrete course of the day because she is updating a prediction map of her world with it, not because she is judging you. That is predictive coding. Autistic brains build detailed inner models of what has happened and what will happen. The exact question about how the day went is not control but data collection: she needs the information to close uncertainties. When you were not with her, her prediction model ran with gaps. The question closes these gaps. The more complete the answer, the lower the open load of uncertainty.
What you experience as a "feeling of being tested" arises through the Double Empathy Problem. Two different communication styles collide: at the handover you might expect a more casual exchange that signals "all good, we are together again". Your daughter asks directly and in detail because that fulfils exactly this function for her. Both sides are looking for security, but through different channels. Her direct, content-focused question sounds like accountability to you. Your expectation of a warmer, more casual welcome might be incomprehensible to her, because phatic communication, meaning conversations whose main task is signalling relationship rather than exchanging information, has a different neurological weight.
This means: the exact questions are her way towards you, not mistrust. Complete, calm answers fill the prediction model and end the questions faster than evasive answers.
Context
The answers rest on well-described mechanisms. Predictive coding describes how the brain compares expectations with impressions; in autistic perception this comparison is weighted differently. Intolerance of uncertainty describes how stressful open situations can feel. Masking means hiding autistic reactions at high inner cost. The Double Empathy Problem describes gaps in understanding between people who process differently.
Three points should be named openly. The first answer applies the concept of intolerance of uncertainty to the grandmother herself; research has mainly studied it in autistic people, so here it serves as a description, not a finding. The third answer assumes that the daughter herself is autistic; the question leaves that open. And the picture of an emptied budget in the second answer is a model, not a measured finding for this child. The last paragraph of the first answer goes beyond pure explanation. This article does not replace a diagnosis or advice.
Autistic Mirror explains such mechanisms for your own specific situation, also from the perspective of grandparents, without instructions. Try it for free.
Read on: Well meant, heard as blame and Guide for grandparents.
Sources
- Pellicano & Burr (2012). DOI: 10.1016/j.tics.2012.08.009
- Vasa et al. (2018). DOI: 10.1002/aur.1916
- Milton (2012). DOI: 10.1080/09687599.2012.710008
- Hull, Mandy, Lai, Baron-Cohen, Allison, Smith & Petrides (2019). DOI: 10.1007/s10803-018-3792-6