It is 3 a.m. and my son is awake. Not drowsy awake, not almost back to sleep awake. Awake awake. His sister is awake now too, because that is how sound works in a house, and I am awake because that is how parenting works in this house. This has been our normal for so long that one of my websites carries the line it was born at 3 a.m., up with my son. Parents online have a name for this life... nocturnal autism. Nobody official calls it that. There is no diagnosis code for it. But if you know, you know.

It has a name in parent land because it is that common

Sleep problems are one of the most studied parts of autism, and the numbers are not small. Reviews of the research put sleep disturbance in autistic children and teens somewhere between 50 and 80 percent, compared with roughly 11 to 37 percent of other kids. Depending on the study, around 40 percent struggle to fall asleep and about 35 percent struggle to stay asleep. Measured with sleep studies and activity trackers rather than parent reports, autistic kids show less total sleep, longer to fall asleep, and lower sleep efficiency than their peers. One long-running study following children from toddlerhood found autistic kids sleeping roughly 17 to 43 minutes less a night, every night, for years.

So if your house runs on a different clock than the rest of the street... that is not a discipline problem. That is the majority experience, documented.

What a nocturnal night actually looks like

The patterns researchers describe will sound familiar to anyone living them. Trouble falling asleep in the first place. Waking again and again through the night. Waking for the day at an hour most people would call the middle of the night. And the one parents call split nights... asleep fine for a few hours, then wide awake and cheerful for one to four hours in the small hours, then maybe back down before the alarm. Split nights is a parent term rather than a clinical one, and researchers mostly record it as prolonged night waking, but spend five minutes in any parent forum and you will meet it by name.

My favourite example is a thread on the National Autistic Society community titled, in all caps energy, "Sleep or LACK OFF..!!" A parent of four boys, three on the spectrum, describes bedtime at 6 or 7 p.m. and all of them awake without fail between 3 and 5 a.m... even when bedtime got pushed to midnight to try to break the pattern. The post mentions locked doors, one son described as a Harry Houdini, toothpaste eaten like ice cream, and ends by asking whether coffee will simply be their best friend forever. That is not a family doing bedtime wrong. That is a body clock running on its own schedule.

Why their brains do not clock out

The research points to a pile of overlapping causes rather than one switch. In some autistic people, the body produces less melatonin, the hormone that signals night. One widely cited study found reduced activity in ASMT, a gene involved in making melatonin, alongside lower melatonin levels in autistic participants. Genes that run the body clock itself have also been implicated. Layered on top of that is a nervous system that stays on high alert... sensory hyper-responsiveness keeps the brain scanning the room instead of powering down.

Anxiety deserves its own sentence, because in one study of predictors it came out as the strongest single predictor of sleep disturbance, with sensory issues and gut symptoms also in the mix. And the gut matters more than people expect. Reflux, constipation, and pain that a child cannot describe can run a whole night. Add in co-occurring conditions like ADHD, epilepsy, sleep apnea, and restless legs linked to low iron, and you start to see why there is no single fix. The AAP-endorsed practice pathway for sleep in autism basically says: screen for the medical stuff first, because some of it is treatable.

What the evidence says actually helps

The first-line recommendation in the American Academy of Pediatrics and Autism Treatment Network pathway is parent-based behavioural sleep support before medication... routines, sleep schedules, and coaching parents through a plan. It works, modestly: a meta-analysis of the randomised trials found about 24 extra minutes of sleep a night on average. Twenty-four minutes does not sound like much until you have lived without it.

Melatonin is the best-studied supplement here, and the evidence has a very specific shape. Across trials, melatonin helps children fall asleep faster and sleep longer overall... roughly 48 extra minutes and half an hour less lying awake in one meta-analysis of neurodevelopmental trials. What it did not do in those trials is reduce how often kids woke in the night. Parents in the forums say the exact same thing in plain language: melatonin gets them to sleep, it does not keep them asleep. The regular kind can also wear off partway through the night, and in one large trial kids on immediate-release melatonin fell asleep faster but woke about half an hour earlier in the morning. A prolonged-release version designed for autistic children did better on staying asleep in its trial, and a two-year follow-up found the benefits held, with no sign of delayed puberty or stunted growth. That follow-up matters, because the long-term safety question is the one that keeps parents up at night... well. One of the things.

And the weighted blanket, because someone has already suggested one to you. In the best trial we have, 73 autistic children were randomised to a weighted blanket or an identical-looking control blanket. The weighted blanket produced no measurable improvement in total sleep, falling asleep, or night waking. The children and parents in the study preferred it anyway. Both of those things are true, and honestly that sums up a lot of autism parenting gear... if your kid loves the blanket, the blanket is doing something. Just know the sleep numbers did not move.

Beyond melatonin, the cupboard gets thin. Medications like clonidine get used, mostly on small studies and clinical experience rather than big trials, and that is firmly talk-to-your-doctor territory. Magnesium gets recommended constantly in parent groups; the trial evidence for it is weak to nonexistent. The gap between what is sold to tired parents and what is proven is... wide.

The safety part nobody warns you about

Night waking stops being only about sleep when a child can get out. In a well-known Pediatrics study, 49 percent of autistic children had tried to wander or run off after age four, and just over half of those children went missing long enough to cause real concern. The National Autism Association reports that 40 percent of parents say fear of wandering disrupts their own sleep, which is its own bleak statistic... you cannot sleep deeply beside a door you do not trust. Door alarms, locks up high, a bell on the bedroom door. It reads as extreme until you have met your child in the kitchen at 4 a.m. standing on the counter.

And then there is us

The research on parents is smaller but it is not subtle. In a 2023 survey of 409 parents of autistic children, 86.6 percent of parents were poor sleepers themselves, and parents whose children had sleep problems reported significantly more stress, anxiety, and depression, and lower quality of life, than parents whose autistic children slept. Read that again. Your exhaustion is not a side note. It is in the literature.

The forums fill in what the surveys cannot. Parents describe taking shifts like it is a newborn rotation that never ended. A double bed bought for the child so a parent can crawl in instead of commuting down the hall all night. One grandmother's thread describes the family waking with their boy most mornings until he was twelve... and then, slowly, it eased. By fifteen it was rare. That thread is the one I think about. Not because improvement is guaranteed. The long-term data is honestly mixed, and autistic adults report high rates of insomnia too. But because for that family, the nights did change.

Belonging to the 3 a.m. club

Nobody applies to join this club. You join it the way we did... one ordinary night that turned into a thousand ordinary nights. The science says your child's brain is wired with a different clock and a louder alarm system. The parents say melatonin might buy you the front end of the night, a locked door buys you peace of mind, and the rest is endurance and each other.

If you are reading this at 3 a.m. with a small person wide awake beside you... you are not failing at bedtime. You are just on shift. The rest of the club is awake too. 🌙