It is 2.40am. You have done everything you were told to do. No screens, no coffee after midday, room dark, room cool. You have been lying still for three hours doing an excellent impression of a person who is asleep. And somewhere behind your eyes the day is still running: the thing you said in the meeting, the fridge cycling on in the next room, the seam of the sheet against your ankle, the specific pressure of knowing the alarm goes off in four hours. You are not bad at sleeping. Your nervous system has not been given a reason to stand down yet.
Autistic insomnia in adults is chronic difficulty falling asleep, staying asleep, or waking rested, and it is driven by neurology rather than habit. Three things stack. Your melatonin system may make less of the hormone and release it later, so the sleep signal arrives hours after the clock says it should. Your sensory system does not switch off after dark, so every texture, sound and light source keeps registering. And a day of masking leaves your nervous system still in performance mode at bedtime. Standard sleep hygiene advice was not built for any of these, which is why you can follow all of it and still lie awake.
What the research shows
- Pooling eight datasets covering 194 autistic adults and 277 non-autistic controls, autistic adults were significantly worse off on 10 of 17 objectively measured sleep outcomes, including how long it took to fall asleep and how much of the night was spent awake after first getting to sleep. Morgan et al. (2020)1
- In a study of 631 autistic adults aged 18 to 65, 20% met criteria for moderate or severe insomnia and a further 34% for sub-threshold insomnia. Insomnia severity tracked with sensory hyper-reactivity, and visual sensitivity showed the strongest link of all. Hohn et al. (2019)2
- Comparing 250 autistic people with 255 controls, researchers found a highly significant reduction in the activity of ASMT, the enzyme that carries out the final step of melatonin synthesis, alongside lower melatonin levels. Melke et al. (2008)3
- Across 730 autistic adults aged 18 to 78, sleep quality was poor at every age. Age itself contributed almost nothing to the models, and depression symptoms did not predict sleep quality at all. Anxiety, perceived stress and physical health did. Charlton et al. (2023)4
Why sleep is so hard when you are autistic
The short version: falling asleep is not a decision, it is a handover. Your body has to pass control from the system that keeps you alert and scanning to the system that lets you switch off. In an autistic nervous system, that handover is slower, later, and much easier to interrupt.
Three mechanisms do most of the work, and they compound rather than take turns. Your melatonin timing may be shifted late or blunted, so the chemical instruction to sleep arrives on a different schedule from the one your life is built around. Your sensory thresholds stay low after dark, so the room keeps sending you information long after a non-autistic person would have stopped noticing it. And the regulatory effort you spent all day on masking, on managing input, on getting through, has to be paid back before your body will let go.
None of this is a habit problem. It is worth saying plainly, because most of us have spent years being handed advice built on the assumption that it is: that if you were just more disciplined about your bedtime, more consistent, less on your phone, this would resolve. You have probably tried a version of that. It did not work, and that was never a character result.
What autistic insomnia actually is
Insomnia is not just “not sleeping”. Clinically it means trouble getting to sleep, trouble staying asleep, or waking too early, happening often enough and for long enough that it starts eating into your days. That last part matters, because it is where most of us actually live: the sleep itself is a problem, but the flattened, foggy, brittle day afterwards is the thing that costs you.
The numbers are stark. In Hohn and colleagues’ sample of 631 autistic adults, one in five met the threshold for moderate or severe insomnia, and another third were sitting just underneath it.2 That is more than half of a large autistic adult sample carrying at least some clinically recognisable insomnia. If you have assumed you were unusual in this, you are not. You are in the majority of the room.
There is also a shape to it that is worth naming. Autistic insomnia tends to be chronic rather than episodic. It is not usually the classic pattern of a stressful fortnight followed by recovery. It is more often a permanent feature of your relationship with the night, present since childhood in many cases, going quiet in low-demand periods and getting much worse when demand climbs.
The melatonin problem, and why your body clock runs late
Melatonin is the hormone that tells your body the day is over. It should rise as evening light drops and fall again as morning comes. The Melke study found something specific about that process in autistic people: significantly reduced activity in ASMT, the enzyme that performs the last step of making melatonin, and lower melatonin levels overall.3
What that means for you, practically, is that the signal may be quieter than it should be and may arrive later than the calendar wants it to. If you have spent your life being told you are a night owl by choice, this is worth sitting with. For a lot of us it is not a preference or a bad habit picked up in our twenties. It is where the internal clock genuinely sits.
The chemistry of downshifting
Melatonin sets the timing, but it is not the only part of the handover. Getting from alert to asleep also depends on the brain being able to quieten its own activity, and that is largely the job of GABA, the main inhibitory neurotransmitter, with serotonin involved both in arousal and as the raw material melatonin is built from. Differences in GABA function in autistic brains are one of the more consistent findings in the wider literature, and the practical consequence is the one you already know: the volume does not come down on command. What gets called overthinking at bedtime is often this, a brain that has not received a strong enough instruction to stop processing.
Delayed sleep phase, in plain terms
Delayed sleep phase is when your whole sleep window is shifted later than the world around you. You are genuinely not tired at 10pm. You are properly tired at 2am. Left alone, you would wake naturally at 10am feeling reasonably human. The sleep itself is not necessarily broken, it is just happening on a different timetable.
The damage comes from the collision. Work starts at nine, so you go to bed at eleven and lie there, then get up at seven having had four hours. Do that five days a week and you accumulate a sleep debt that has nothing to do with sleep quality and everything to do with schedule mismatch. Then you catch up on the weekend, which shifts your clock later again, and Monday is worse. It is not a discipline failure. It is two clocks that do not agree, and only one of them is negotiable.
Why the sensory piece feeds the clock piece
Here is the part that usually gets missed. Your circadian rhythm is not set by an internal timer alone. It is set by light and by the rhythm of your daily life, and both of those are things sensory difference changes.
Hohn and colleagues put this forward directly: if bright light is aversive to you, you avoid it, and avoiding it reduces the light signal your body clock needs to stay anchored to the actual day.2 The same applies to social rhythm. A day with fewer scheduled human anchors in it gives the clock less to synchronise to. So the sunglasses indoors, the drawn curtains, the cancelled plans, all of which are sensible responses to sensory overload, may also be quietly loosening the clock that decides when you get tired. That is not a reason to stop protecting yourself. It is a reason to stop blaming yourself for the 2am wakefulness that follows.
Why the sensory world gets louder after dark
During the day, sensory input competes. There is a lot of it, and some of it drowns the rest out. At night that competition disappears and whatever is left gets the whole stage.
Touch
A fabric that was mildly annoying at 3pm can be genuinely intolerable at midnight. The seam. The tag. The particular scratch of a sheet that has been washed with the wrong thing. The way the duvet is heavy on one shoulder and not the other. You are not being precious about it. With nothing else pulling at your attention, tactile input that your brain was suppressing all day becomes the loudest thing in the room, and it can hold you at the edge of sleep for hours or pull you back out of it at 4am.
Sound
Most people habituate to steady household noise. The fridge starts, and within a minute their brain has filed it as irrelevant and stopped reporting it. That filing step is less reliable for a lot of us. The fridge starts, and it keeps being a thing that is happening. So does the neighbour’s television through the wall, the car three streets away, the tick of something cooling. These sounds do not need to be loud to keep you awake. They only need to be unpredictable, because an unpredictable sound is one your nervous system cannot safely stop monitoring.
Light
The standby LED on the television. The strip of streetlight where the curtain does not quite meet the wall. The phone charger. For a visually sensitive nervous system these are not background, they are things in the room, and your eyes keep going back to them. Given what the research says about visual hyper-reactivity being the sensory channel most strongly tied to insomnia severity, this is not a small detail.2
Temperature
Your comfortable range may be narrow, and you may cross out of it faster than the people you share a home with. Too warm and you cannot settle. A draught on one arm and you are awake. The heat coming off another body in the bed shifts the whole equation. Waking at a temperature change that someone else would sleep straight through is a normal consequence of interoceptive and tactile sensitivity, not evidence that you are a light sleeper by personality.
The inside of your body counts too
Interoception is the sense that reports on your internal state, and for many of us it runs either too quiet or too loud. Too quiet and you get into bed without registering that you are actually hungry, or that you needed the bathroom, and then lie there vaguely wrong. Too loud and you can hear your own heartbeat, feel your pulse in your jaw, notice your digestion. Neither is imagination, and neither responds to being told to relax.
The masking cost that arrives at bedtime
Sleep research almost never mentions this one, and it may be the biggest single factor in your week.
A day of masking is a day of continuous voluntary regulation. Managing your face. Timing your responses. Suppressing the stim that would have helped. Tracking the room. Working out what was meant rather than what was said. None of that is free, and all of it is drawn from the same pool of regulatory capacity that would otherwise be available to bring you down into sleep. By the time you get home, the pool is empty. By the time you get into bed, your body is exhausted and your nervous system is still standing at the door in uniform.
This is why the autistic overwhelm of a heavy week so often shows up as insomnia rather than as tiredness. The more you had to hold together during the day, the more wired you are at night. It is a horrible piece of maths: the days you most need to sleep are the days you will sleep worst.
“By the time I get into bed I’m exhausted. But my nervous system doesn’t know that. It’s still managing everything from the day. It takes hours to slow down to something that feels like it could actually sleep.”
— Autistic adult, HeyASD community
The residual alertness has a particular quality that is worth distinguishing from ordinary insomnia. It is not racing thoughts exactly, and it is not anxiety exactly, though it can turn into both if you lie in it long enough. It is closer to being on shift. The scanning has not been switched off, so the part of you that has been watching all day carries on watching an empty room.
If the thing keeping you awake is the accumulated weight of performing a version of yourself all day that does not fit, that is the terrain The Unmasking Years covers: what masking actually costs, and what putting it down slowly looks like in a real life.
When you cannot make yourself go to bed
There is a version of this that has nothing to do with lying awake, and it gets talked about even less. You are tired. You know you are tired. You want to be asleep. And you cannot get yourself to start the sequence that ends in bed.
Going to bed is not one action. It is a chain: stop what you are doing, get up, teeth, contacts out, tablets, lights, door, charger, alarm. That is a lot of executive function to summon at the exact hour of the day when yours is most depleted, and the transition out of what you are currently doing is the hardest part of the whole chain. So you stay in the chair. Another episode. Another twenty minutes of scrolling that you are not enjoying.
Two separate things are usually happening. One is the transition cost, which is a genuine autistic difficulty and not procrastination in the ordinary sense. The other is that the late evening may be the only stretch of the day that belongs to you, with nobody making requests of you, and giving it up to go to sleep feels like handing back the only unmasked hours you got. Both are real. Neither is laziness, and neither responds to being told to have better boundaries with yourself.
What autistic sleep actually looks like
If you have never seen your own pattern written down, it can be oddly steadying to recognise it.
Long sleep latency
Lying there for an hour or more before anything happens. This is one of the most consistently measured differences: the meta-analysis of objective sleep studies found significantly longer sleep onset latency in autistic adults, meaning it is being picked up by equipment in a lab, not just reported on a questionnaire.1 The frustration that builds while you wait makes it longer, which is its own particular cruelty.
Fragmented sleep
Surfacing three or four times a night. Sometimes for a minute, sometimes for ninety. The same research found significantly more time spent awake after first falling asleep, and lower sleep efficiency overall, which is the proportion of your time in bed that you actually spend asleep.1 Eight hours in bed can quite easily be five and a half hours of sleep in four pieces.
A clock that runs late
Genuinely alert at midnight, genuinely dead at 7am, and never quite able to convince anyone that this is not a choice. If your mornings are the worst part of your day, that is a schedule collision rather than a personality trait, and building a gentler morning routine around it helps more than trying to become a different kind of person.
Sleep that does not restore
You were unconscious for eight hours and you woke up feeling like you had not been. Given the fragmentation and the reduced sleep efficiency in the objective data, this is a fair description of what happened rather than an exaggeration.1 The hours were there. The sleep was in pieces.
“I sleep for eight hours and wake up feeling like I didn’t sleep. And then I have a whole day to get through. The exhaustion isn’t laziness, it’s structural.”
— Autistic adult, HeyASD community
If you also have ADHD
A large proportion of us are also ADHD, and the two sets of sleep problems do not politely alternate. The autistic contributions are sensory load, masking cost and melatonin timing. The ADHD contributions are a delayed circadian rhythm of its own, a mind that will not stop generating, and the specific trap of finally getting interested in something at 11pm. Stacked together you get a pattern that is later, more irregular and more resistant to standard advice than either would produce alone. Most of what is written about neurodivergent sleep is really describing this combination.
A note on functioning labels
You may have searched something like “high functioning autism sleep problems”, because that is the phrase you were given. It is worth knowing that the label predicts nothing useful here. Being able to hold a job and speak in meetings does not mean less sensory load at night. Very often it means more, because you are masking harder for longer to maintain the presentation that earned you the label in the first place. The amount of support you visibly need and the amount of sleep you are losing are close to unrelated.
What it costs, and when to take it somewhere
It is easy to file chronic bad sleep under things you have simply learned to live with. The research is less relaxed about it. In a two-year longitudinal study of 598 autistic adults, sleep problems came out as the single most important predictor of later quality of life.5 Not the most important sleep-related predictor. The most important predictor, full stop.
That tracks with what a bad stretch feels like from the inside. The sensory threshold drops, so ordinary environments start producing overstimulation that you would normally absorb. The gap between input and meltdown or shutdown narrows. Masking gets more expensive, which costs you more sleep, which is how a rough fortnight turns into the slide towards autistic burnout.
It is probably not just depression
This is worth stating clearly, because you may have been told otherwise. In Charlton and colleagues’ study of 730 autistic adults, depression symptoms did not significantly predict any aspect of sleep quality. Anxiety, perceived stress and physical health did.4 The two do overlap and can feed each other, and it is genuinely worth understanding the relationship between autism and depression. But if a GP has treated your sleep as a symptom that will resolve once your mood does, that is not what the adult data supports.
It does not resolve on its own with time
The same study covered ages 18 to 78 and found that age contributed almost nothing to sleep quality.4 Sleep was poor across the whole adult lifespan. If you have been waiting to grow out of this, that is the honest answer: it does not appear to happen by itself.
Worth a GP appointment
Take it to a doctor if it is affecting your ability to work, drive, regulate or function, and ask about three things specifically. Obstructive sleep apnoea, which is under-diagnosed and produces exactly the fragmented, unrefreshing pattern described above. Delayed sleep phase, which is a recognised circadian rhythm disorder with its own management approaches. And whether anything you are already taking is contributing. If you are offered CBT for insomnia, it does have evidence behind it, but it assumes a nervous system that responds to standard stimulus control, so it works best with a practitioner who will adapt it rather than run the protocol as written.
Understanding the mechanism is the part that makes the rest possible, because it tells you which lever you are actually pulling. Once you know whether you are dealing with a clock that runs late, a room that is too loud for your nervous system, or a masking debt that has to be paid before you can settle, you can look properly at what actually helps autistic adults sleep and choose the things that match your reasons rather than working through a generic list.
Key points
- Autistic insomnia is a neurological pattern, not a habit problem: in one sample of 631 autistic adults, more than half were carrying insomnia at sub-threshold level or above.
- Your melatonin system may make less of the hormone and release it later, which is why being a night owl is often a body clock setting rather than a preference.
- Sensory difference works on the clock as well as the room: avoiding bright light because it hurts also reduces the light signal your circadian rhythm needs to stay anchored to the day.
- The regulatory cost of a day of masking arrives at bedtime as residual alertness, which is why the weeks you most need sleep are the weeks you sleep worst.
- Not being able to start the going-to-bed sequence is a transition and executive function difficulty, not procrastination, and it deserves to be treated as a separate problem from lying awake.
- Sleep problems were the strongest predictor of later quality of life in a two-year study of autistic adults, and they do not resolve on their own with age.
Questions about autism and sleep
Why do autistic adults struggle with sleep?
Three things stack up on you. Your melatonin system may produce less of the hormone and release it later than the schedule you live on, so the biological instruction to sleep arrives hours after you got into bed. Your sensory thresholds stay low after dark, so texture, sound, light and temperature keep registering when there is nothing else competing for your attention. And a day of masking spends the same regulatory capacity that would otherwise bring you down into sleep, so you arrive at bedtime physically exhausted with a nervous system still running. None of these respond to being more disciplined about your bedtime, which is why standard sleep advice so often fails you.
Does autism cause insomnia?
Not as a direct symptom, but the neurology makes insomnia far more likely and far more persistent. Reduced and delayed melatonin makes getting to sleep harder. Sensory hyper-reactivity makes the environment harder to get right and makes you easier to wake. The masking cost keeps arousal high in the evening. Co-occurring anxiety, ADHD and sleep apnoea add more. The result is that insomnia turns up in autistic adults at rates well above the general population and tends to be chronic rather than a rough patch that passes. Treating it properly means addressing the autistic-specific causes, not just applying the standard insomnia playbook harder.
Does autism affect sleep?
Measurably, yes, and it shows up on equipment rather than only on questionnaires. A meta-analysis pooling objective studies of autistic adults found significant differences on 10 of 17 measured outcomes, including how long it took to get to sleep and how much of the night was spent awake afterwards. So when you report that you were awake for hours or that you kept surfacing, you are describing something instruments have independently recorded in autistic adults as a group. It is not a perception problem or an unreliable memory of the night. The difference is real, it is physical, and it has been measured.
What do autistic sleep patterns look like?
Most commonly: a long stretch of lying awake before anything happens, a sleep window that sits later than conventional schedules want, waking several times in the night, and getting up after what should have been enough hours feeling like you did not sleep. Sleep efficiency, meaning the proportion of your time in bed that is actually sleep, tends to be low. On top of that there is a specific pre-sleep alertness after high-demand days that is distinct from ordinary sleeplessness. Your own pattern may include some of these and not others, and it will usually get noticeably worse when the demands on you increase.
Do autistic people need more sleep?
You may well need more time in bed to get the same amount of usable rest, because fragmented sleep and low sleep efficiency mean the hours convert badly. There is also a recovery argument. Managing sensory load, masking and constant processing costs energy that a non-autistic day does not spend, so there is more to recover from. Many of us find our sleep need swings much harder with recent demand than other people report: unremarkable after a quiet week, enormous after a heavy one. Needing eleven hours after three days of people is not greed. It is the bill arriving.
What are neurodivergent sleep issues?
It is a loose term covering the sleep difficulties that cluster in autistic and ADHD adults, and in the very large group who are both. The autistic contributions are sensory load, melatonin timing and masking cost. The ADHD contributions are a delayed circadian rhythm, difficulty stopping activity in the evening, and a mind that keeps producing. When both are present they compound into something later, more irregular and more resistant to standard advice than either produces alone. If you have read general neurodivergent sleep content and found it nearly right but not quite, it is often because it is describing the combination rather than your specific arrangement.
What is delayed sleep phase in autistic adults?
Your entire sleep window sits later than the world around you. You are genuinely not tired at 10pm and genuinely alert at 1am, and left to your own devices you would wake around ten feeling reasonably functional. The sleep itself is not necessarily poor, it is just scheduled differently. The harm comes from the collision with work and school hours: you go to bed at a time your body does not accept, lose hours at the front of the night, and get up before you have finished. Repeat it five days a week and the debt compounds. It is a timing mismatch, not a discipline failure.
Is autistic insomnia different from ordinary insomnia?
The experience of lying awake is similar. What differs is what is generating it and how it responds. Ordinary insomnia is often triggered by a stressor and eases when the stressor does. Autistic insomnia tends to be lifelong, closely tied to sensory environment and daily demand load, and driven partly by melatonin differences that will not shift with better habits. It also has that specific post-masking alertness that does not feel like anxiety and does not respond to relaxation exercises. Standard treatment can still help, but it usually needs adapting rather than applying straight, which is worth saying to any clinician you see.
Do autistic sleep problems get better with age?
The evidence says not on their own. A study of 730 autistic adults spanning ages 18 to 78 found sleep quality was poor across the whole adult lifespan, and age itself contributed almost nothing to the models. What did predict sleep quality was physical health, anxiety and perceived stress. That is a hard thing to read if you have been quietly waiting to grow out of this, and it is better to know. It also points somewhere useful: the levers that move are stress load, physical health and sensory environment, rather than time passing.
Is it insomnia or is it depression?
They overlap and can drive each other, but do not accept the assumption that your sleep is simply a symptom of low mood. In the largest adult study on this, depression symptoms did not significantly predict any aspect of sleep quality in autistic adults, while anxiety and perceived stress did. If a GP has told you that treating your mood will fix your sleep, that is not what the adult data shows. It is reasonable to ask for the sleep to be assessed as its own problem, alongside anything else being treated, rather than filed underneath it.
Do high functioning autistic adults have sleep problems too?
Yes, and the label predicts nothing useful about it. Holding down a job, driving, speaking fluently in meetings: none of that reduces your sensory load at night or your melatonin timing. Often it means the reverse, because maintaining the presentation that earned you that description takes sustained masking, and masking is one of the main things keeping you awake. The amount of support other people can see you needing and the amount of sleep you are losing are close to unrelated. If anything, being read as coping well makes it less likely anyone asks how you are sleeping.
When should I see a doctor about autistic sleep problems?
When it is affecting your ability to work, drive, regulate your emotions or manage sensory load, which for chronic sleep loss is most of the time. Ask about three things by name. Obstructive sleep apnoea, which is under-diagnosed and produces exactly this fragmented, unrefreshing pattern. Delayed sleep phase, which is a recognised circadian disorder with its own management. And whether any current medication is contributing. If CBT for insomnia is offered, it does have evidence behind it, but ask whether the practitioner has adapted it for autistic adults, because the standard protocol assumes responses that may not be yours.
If you’ve lived this.
Nobody has added anything here yet. You are welcome to be the first.