Intolerance of uncertainty (IU) refers to the degree to which someone finds it difficult to tolerate situations where outcomes are unknown or unpredictable.1 It isn’t a diagnosis, but a trait — one that varies in intensity across individuals and shows up across many different presentations, including anxiety disorders, OCD, autism, and ADHD.1 Researchers have increasingly understood it as transdiagnostic: an underlying characteristic that shapes how a nervous system processes not-knowing, regardless of what diagnostic label sits above it.
If you need to know the plan before you can settle into an experience, if you find yourself gathering information in advance so you can actually be present rather than managing as you go, if a change of plans lands harder than you’d expect it to, then intolerance of uncertainty may describe something real about how your nervous system works.
Table of Contents[Hide][Show]
What is intolerance of uncertainty?
For autistic and ADHD adults, intolerance of uncertainty tends to run at a higher baseline than in the general population. A large systematic review and meta-analysis found the correlation between IU and anxiety in autistic people to be substantial — a large effect, comparable in size to the same relationship in neurotypical populations.2 The same mechanism, at greater intensity. Worth noting is where this intensity tends to show up: in autistic people, IU appears to manifest primarily as immediate emotional dysregulation — stress, an inability to settle, being destabilised by unforeseen events — rather than as catastrophic worry about future outcomes.3 The distress is present-tense and physical first.
Neurological background for IU
There may also be a neurological dimension to this. The mTOR pathway, which is overactive in autistic brains, produces denser local neural wiring and a higher baseline of brain activity. A nervous system that is already processing more — more sensory data, more internal signals, more competing connections — has less spare capacity for the additional cognitive load that uncertainty brings. When the system is already running near capacity, “I don’t know what’s going to happen” is one more demand on a brain that is already doing a lot.
Fundamentally, IU means a nervous system that needs enough predictability to function. A brain that processes uncertainty as high-urgency information needs more information to reach a settled state than a brain that doesn’t. This isn’t a flaw, and it isn’t a choice. It’s a description of how the system works.
What Intolerance of Change Feels Like From the Inside
When something remains unresolved — a plan that hasn’t been confirmed, an outcome that isn’t yet clear — there’s a quality to the experience that’s hard to describe to someone who doesn’t share it. Untethered might be close. Without solid information to anchor to, the nervous system begins generating contingencies instead: a cascade of open loops, all the unanswered questions running at once. Where are we going? What time? What will the food be like? Is it noisy? How are we getting home? For a nervous system with high IU, these questions occupy the foreground until they’re resolved.
The information-gathering serves a purpose. It’s preparatory work — solving the logistical and sensory unknowns in advance, so that when the event arrives, bandwidth is free to do something other than manage it in real time. The restaurant address, the sensory profile of the venue, the rough shape of the evening — once those are accounted for, there’s room to actually engage with the experience. Without that groundwork, you spend the whole time doing accommodation work on the fly, and the accommodation work crowds out the event itself. Research into how sensory uncertainty and IU interact in autistic adults confirms this: sensory unpredictability amplifies the experience of uncertainty significantly, and managing both simultaneously is a genuinely high-cost combination.4
A grounded theory study that centred autistic adults’ own accounts of how they experience uncertainty found something that most people with high IU will recognise immediately: uncertainty isn’t a single thing.5
- There are knowns — the stable constants you can rely on.
- There are known unknowns — situations whose outcome is uncertain but can be planned around, because the parameters are understood.
- And there are unknown unknowns — the kind of uncertainty that arrives without warning and offers no anchor to replace what it has dislodged.
That last category is where the real distress lives.
The same research found that uncertainty-related anxiety is most acute when control is absent, not merely when uncertainty is present.5 This is a meaningful distinction. A surprise can be entirely tolerable — even enjoyable — when you trust the person creating it and have a sense of the possible range. Someone who tends to bring small things home, or order things online, who has never surprised you with anything requiring you to dismantle your plans — their surprises land within a known range. The unknown is bounded. And the bounded unknown is a manageable version of not-knowing in a way the unbounded unknown is not. The parameters themselves are the accommodation.
The Recalibration Cost of Changing Plans
There’s another dimension to this that sits slightly apart from the question of unknowns, and it’s one that the clinical literature captures poorly compared to the lived experience.
When you’ve already assembled yourself around one expectation — calibrated your nervous system to a specific context, built a mental model of how the evening was going to go — a change of plans doesn’t mean updating a single variable. It means dismantling the entire context you built and constructing a new one: a different route, a different sensory profile, a different texture and timing and an internal map of how the event will feel. None of that transfers. One element changing in a large mental network requires checking if every other connected variable in the framework still holds. A simple change is never a simple change.
Example: Taco night is cancelled!
If you were going for tacos and the plan changes to pizza — somewhere you genuinely like — the difficulty isn’t about pizza. Pizza is fine, you love pizza! The difficulty is that “tacos” was never just a food. It was a whole assembled context: a specific route, a particular mouthfeel, a smell, an evening mapped in your mind in a particular way. All of that has to come down, and something new has to go up in its place. Being told it’s no big deal, that you’ll enjoy both, that it doesn’t really matter — well-intentioned as it is — misses the cost entirely. The cost isn’t in the destination. It’s in the demolition and rebuild.
Neuroscience offers a useful frame here.
There’s a concept called predictive processing — the brain constantly generates predictions about what’s going to happen next, and monitors for when reality doesn’t match those predictions. When a mismatch is detected, the brain generates what’s called a prediction error — basically a signal saying something is different from what I expected. So far, so good; all brains do this.
However, in autistic brains, those mismatch signals are weighted as more important, more urgent than they would be in a neurotypical brain. So the same deviation from expectation triggers more processing and more of a response.6 A deviation from expectations doesn’t just register as huh, that’s different — it registers as this matters! deal with it now!. And it registers that way even for small deviations. That’s what makes plan changes have such a seemingly disproportionate reaction. The brain is structurally set up to treat departures from its mental model as high-priority information.
More recently, neuroimaging research found that autistic people are already doing significantly more cognitive work during the preparation phase before a cognitively demanding task, before anything uncertain has actually arrived yet. And the brain is working harder in anticipation, not just the event itself.7
The cognitive work of bracing for uncertainty begins earlier and costs more from the outset. When a plan changes, that preparation has already been made, and now needs to be remade.
If the brain builds a very detailed, specific internal model of how something is going to go, then when that model turns out to be wrong, there’s a lot to dismantle. The more detailed the blueprint, the more work it takes to tear it down. A vague, loosely-held expectation being “wrong” costs less to update than a fully-constructed mental picture of how the evening was going to feel, smell, and run.
This is also why the nature of the transition matters as much as the transition itself. Recalibrating to a new context can only happen if there is a new context to adjust to.
If we agreed to leave at 2 pm, it is now 4 pm, and we still have not left, this means that we are still in uncharted waters.
What is the new timeline? Nobody knows. When will it happen? Soon. How soon? Just be patient, and do something else.
Except… not knowing the boundaries of the new “free” time that appeared means nothing can be started, in case it will be interrupted in five minutes. So we enter limbo, waiting mode, and stay in the discomfort of the liminal space.
A clear new timeline — we’re leaving at six — would convert an unknown unknown into a known unknown. It gives the nervous system something concrete to recalibrate around, and means the intervening time can be used rather than held in suspension. Instead of hearing “just five more minutes” repeated for three hours, something tangible and concrete is now there. Clarity is kindness.
Moving the goalposts with changing plans
There is an additional layer that compounds intolerance of uncertainty further that is mostly invisible from the outside. When a mental model includes an endpoint — “this difficult thing ends in seven days” — the tolerance built around it is conditional on that deadline. Sacrifices are made, sensory or social discomforts are endured, resources are rationed, all calibrated to a known finish line. The endpoint is not just another data point to the plan, but it is load-bearing information that everything rests upon.
When a change pushes that deadline back, it does more than simply add more time, because it invalidates the rationing strategy entirely, rendering it useless. Things that were being endured because they were nearly over are now not nearly over. Resources allocated to last a specific duration are suddenly insufficient for the new one. And the person announcing the change — who has no visibility into how many concessions were already built into the original agreement — experiences the response as disproportionate to what, from the outside, looks like a simple delay.
Imagine you signed up for and trained for a 5k run. And while already running and nearing kilometre 4, someone shouts from the sidelines: “It’s actually a 10k race, okay?“, waiting for your encouraging positive response with a smile.
But no, it is not okay. This means that the 20% of energy you have left for the last bit suddenly becomes insufficient, as now the energy needed is 120%. But because the race is underway already, and there is no built-in rest or recovery, there is no mental headroom to negotiate, to stand up for yourself, to push back. It is a very vulnerable position with already depleted energy reserves. Additionally, since the goalpost has just moved, the mental framing of “almost done” immediately switches to “if I stop where I was supposed to stop, then I have failed“. The spot where the previous finish line was (with the confetti and the cheering) is now just another marker you need to run by.
Therefore, the newly raised spectre of failure adds to the coercion already set up, because of course you don’t want to fail, you don’t want to give up, you don’t want to disappoint anyone. All of these are taking you closer to a decision that goes against your well-being. And since your energy has been eroded by this time, it is easier to be coerced. Eventually, if it happens enough, this is a one-way ticket to burnout.
How Need for Clarity Gets Misread
The gap between how intolerance of uncertainty looks from outside and how it feels from within is where a great deal of harm accumulates over a lifetime.
- From the outside: someone who asks too many questions before an event, who needs a confirmed plan before they can settle, who becomes upset over what appears to be a minor change of direction.
- From the inside: a nervous system requesting — directly, without subtext — the information it needs in order to function.
The directness tends to be the biggest problem for these queries. Whether it is a request for the restaurant address, a need for a clear departure time or a question about what the evening will involve: these are purely about clarity.
There is no hidden agenda. No attempt to manage or constrain, to manipulate or micromanage. The nervous system needs information and predictability the way it needs sleep.
One autistic person put it plainly: “I am in fact NOT controlling. I just like to have a reliable expectation.“5 The distance between needing control and needing clarity is exactly the distance between being understood and spending years wondering what’s wrong with you.
Are you “high-maintenance” or do you just have needs?
For women, and for anyone socialised to believe that expressing needs is itself a kind of social failure, being told you’re high-maintenance for asking questions is doing a special kind of cruel damage.
All you are doing is advocating, clearly and directly, for what your nervous system requires. But it is read as requiring “too much” attention and care. It teaches a particular lesson: that your needs are too much. And the needs that have been labelled “too much” don’t disappear just because they are told to stop existing. They go underground, pushed down, never being voiced again.
The underlying need remains unmet, you try to manage it alone (if at all), and the people around you miss information they would need to understand what’s happening. The label, apart from wounding, produces worse outcomes for everyone involved.
For adults who spent years absorbing the message that their needs were “too much”, these effects have a head start in compounding and escalating before they get properly named, given new context, and can be unpacked to start the healing process.
Double empathy and “intolerance of uncertainty”
Research adds something useful to the picture of why this misreading persists.
The social world is, by its nature, harder to navigate under high intolerance of uncertainty than almost any other environment — because other people are genuinely less predictable than systems, routines, or physical spaces.
Autistic adults in qualitative research have described neurotypical people as inherently unpredictable in ways that compound IU specifically in social contexts 5 — and neuroimaging data support this. The cognitive switching cost that shows up in autistic people is significantly larger in social contexts (where other people are involved) than in non-social ones. Even more importantly, it largely disappears in non-social contexts altogether 8. The situations that generate the most visible IU responses are, structurally, the ones with the highest unpredictability load, therefore being more cognitively demanding.9
And there’s an asymmetry operating in the background that the neurodivergent community has articulated clearly: in any social situation, the autistic person is typically engaged in continuous, effortful accommodation of neurotypical norms, unspoken expectations, and shifting social rules — from start to finish, without acknowledgement. Only for one request for clarity to become, somehow, the evidence of “inflexibility”.
The double empathy problem17 offers a frame for understanding why the misreading persists so stubbornly. Milton’s concept describes the communication difficulties between autistic and non-autistic people as bidirectional, as in, it happens both ways. It is not a deficit located in one person, but a disjuncture between two differently disposed social actors who struggle to read each other accurately. The difficulty of reading each other is mutual, but the burden of that difficulty is not.
It seems like the phenomenon that neurotypicals describe as “autistic people having communication difficulties” is often about neurotypical people refusing to make an effort to understand autistic people.
The perception gap in IU runs exactly along this fault line. The autistic person is being misread as controlling or rigid. The neurotypical person is, from the autistic person’s nervous system, a source of genuine unpredictability. Both experiences are real — but only one person is doing the work of bridging the gap, and they are the one being labelled for it.
What you don’t see: the invisible effects of IU
When circumstances require it, and the system still has the capacity to do it, IU’s effects can be contained. The unconfirmed plan is absorbed internally, while the recalibration happens behind whatever facial expression the social situation requires. The evening continues, and to anyone watching, the person appears — if not entirely at ease — at least “functional”.
This containment (which is part of masking) cannot happen every time, is not a permanent solution, and it does not replace processing. Being able to do it at all depends on available capacity, which is itself a moving variable. And the cost doesn’t disappear; it surfaces later — often significantly later — as what looks to observers like a disproportionate reaction or outburst to something minor. The reaction is causally connected to the unconfirmed plan, the unprocessed transition, and the accommodation work done in real time without preparation. The link is real and direct. But because it’s invisible — because some observers don’t see anything, and some only see the endpoint — the explanation is lost, and the response appears inexplicable.
For late-identified adults, there’s an additional layer. Research on alexithymia — difficulty identifying and describing one’s own emotional states — has found that this difficulty mediates and amplifies the relationship between intolerance of uncertainty and anxiety in autistic adults.12 When you can’t accurately name what’s happening internally, you can’t communicate it to others, and you can’t identify what you need. The overwhelm builds without a label and surfaces without a context. Years of this, without an accurate framework, make the delayed unravelling pattern harder to interrupt and harder to explain to the people around you.
Research consistently shows that autistic adults report a lower quality of life than non-autistic adults across all areas of life. 16 When researchers looked at why this is, intolerance of uncertainty and anxiety together emerged as significant explanatory factors — accounting for a substantial part of that gap not just in psychological wellbeing, but in physical health, social relationships, and material circumstances too.13
How flexibility changes with the stacking effect
One of the most important things to understand about IU in lived experience — and one that the clinical literature consistently overlooks — is that the threshold for what’s manageable is not fixed.
The same objective uncertainty (“Grandma cannot come to visit tomorrow“) lands entirely differently depending on what has already been spent that day.
A plan changing at the end of a day that has involved chronic pain, social exhaustion, masking across multiple contexts, and a string of decisions under cognitive load is a fundamentally different experience from the same plan changing on a rested morning.
The IU itself hasn’t changed. But the capacity available to absorb it has. This is cumulative in the way that fatigue is cumulative: invisible to observers, entirely real in its effects, and often the effects are only visible when the stacked tower of exhaustion topples over.
Research into how autistic people actually manage flexibility in daily life confirms that real-world flexibility difficulties predict anxiety far more reliably than performance on controlled laboratory tasks.9 The lab task arrives in a clean context with nothing else already spent. The daily life version is not this clear-cut. A meta-analysis examining cognitive flexibility across multiple studies found a significant gap between autistic performance in laboratory conditions and in real-world situations — a gap that becomes substantially larger when IU and anxiety are factored in.10 The lab measures a theoretical capacity. The real world measures what that capacity looks like after everything else has already been paid.
For people with both autism and ADHD — an AuDHD profile — the executive function demands of both conditions compound in ways that make cumulative load more severe and the threshold shift more pronounced.11
What this means in practice is that the version of someone’s IU reaction that other people observe is often the depleted version — the exhausted Friday evening version, not the “sun is shining and I slept well!” Tuesday version. They did not see the million instances before where the effect of the uncertainty was processed properly (or masked). The observed response gets attributed to the uncertainty itself, rather than to the accumulated weight of everything that came before it.
The clinical questionnaires used to measure IU ask how someone generally responds to uncertainty. They don’t account for sleep, pain, masking hours, social interactions, or existing conditions. They often capture the trait at a stable average, and they miss this dynamic entirely.10
Routines, Sameness, and Self-Accommodation
The logical response to a nervous system that finds prediction errors costly is to engineer predictability wherever it can be engineered.
Eating the same meal, taking the same route, building predictable sequences into the day — these are active management strategies. They work by locking variables in place: removing from the uncertainty budget the things that don’t need to be processed from scratch each time, so that bandwidth is available for the things that can’t be controlled.6
This framing is well supported by the research. Repetitive and sameness-seeking behaviours in autistic people are positively correlated with both IU and anxiety, and the directional relationship matters: these behaviours appear to be responses to uncertainty and anxiety, not sources of them.14 When autistic children and young people are asked to describe their own routines, they’re explicit about the function: sameness is used deliberately to reduce the space in which surprises can happen.15 The routine serves a purpose. Understanding that purpose changes everything about how to approach it.
This is also why approaches that focus on reducing repetitive or sameness-seeking behaviours — without addressing the IU that underpins them — tend to increase distress rather than decrease it.14 Removing a coping mechanism without attending to what it’s managing leaves the underlying need entirely intact, and the person less equipped to meet it.
How misreading IU leads to misdiagnosis
In clinical settings, IU in autistic and ADHD adults is regularly misread as generalised anxiety disorder, OCD, or controlling behaviour within relationships.
The diagnostic distinction that matters most is whether the distress centres on not having information — characteristic of IU — or on catastrophic worry about specific future outcomes, which is more characteristic of anxiety disorders. These can co-occur, but treating IU as anxiety without attending to the information need at its core tends to be ineffective.
For late-identified adults especially, an accumulation of punitive labels over decades — controlling, rigid, high-maintenance, anxious, inflexible — significantly complicates therapeutic alliance. Reframing IU as a nervous system difference rather than a pathological trait is often reparative, because it offers accuracy rather than just shame reduction. Approaches that focus on teaching tolerance of uncertainty, without first attending to what information the person actually needs and why having it matters, can inadvertently reinforce the message that the need itself is the problem.
Both insistence on sameness and cognitive inflexibility are explored in their own entries in this glossary. The current entry introduces them as expressions of the same underlying dynamic: a nervous system that works best when it has enough predictability to function, finding reliable ways to create the conditions it needs.
What actually helps with intolerance of uncertainty
The most consistently useful thing, across all the dimensions of IU described above, is getting information in advance.
Knowing the plan, the location, the likely sensory environment, and the rough shape of the timeline — enough to assemble a working mental model before the event — changes the experience measurably. The preparation work happens in advance, which means the event itself can be enjoyed, in full presence, rather than spent with in-the-moment management and self-regulation.
When advance notice isn’t possible, a clear new timeline helps more than anything else. A concrete, new anchor — we’re leaving at six — converts the unknown unknown into a known unknown. It gives the nervous system something to recalibrate around, and it also means the waiting time can be used rather than suspended in limbo with no defined endpoint. This is the single most practical accommodation available, and it costs almost nothing to offer.
Trust and prior pattern also matter. A surprise from someone whose range you know — whose surprises have consistently stayed within certain parameters — is a bounded unknown. The parameters themselves are the accommodation. They create a space in which the unexpected can arrive without triggering the full alarm.
What doesn’t help:
- Being told the change is “no big deal”.
- Being rushed through the recalibration.
- Being offered an onslaught of new information while the transition is still in process.
- Being reprimanded for any attempt at self-soothing behaviour
These all add load at the moment when the nervous system is already managing as much as it can. The upset that accompanies a sudden plan change isn’t a response to the new destination — it’s a transition in progress. It resolves when it’s allowed to, and it resolves faster when it isn’t being simultaneously suppressed or managed socially on top of everything else.
More broadly, environments that make a material difference to someone navigating high intolerance of uncertainty are those where information is offered proactively, where asking for clarity is understood as a reasonable request rather than a character flaw, and where a nervous system advocating clearly for what it needs is met with that information — rather than a verdict about the person asking.13
That kind of environment doesn’t just reduce anxiety. It changes what quality of life is possible.
