Sleep trouble may be part of the early autism picture that parents should notice

A matched comparison of 492 young children

Two groups of young children were compared in a one-to-one age match. Each autistic child was paired with a typically developing child of the same age. The total sample held 492 children, all under six years old. Half had autism, and half served as controls from local kindergartens. The autistic children came from Guangzhou Women and Children’s Medical Center in China. They were recruited over a period from September 2018 to February 2025. Two independent specialist clinicians confirmed every autism diagnosis using standard clinical tools. Those tools included diagnostic criteria plus two well-tested autism assessments. Children first diagnosed before age two were followed until a firm diagnosis was possible at two or older. Controls were excluded if parents or teachers reported neurodevelopmental disorders. Other genetic, endocrine, or medical conditions also ruled out a control child. Parents reported their children’s sleep through structured interviews and questionnaires. The reports covered bedtime, wake time, night wakings, and how long each waking lasted. Because age was matched, the comparison reduced one obvious source of difference. A toddler and a five-year-old do not sleep in the same way. Pairing by age let the team compare sleep patterns within similar developmental stages. The matching also made the groups easier to balance for other background factors. Analyses adjusted for child age, sex, birth details, and maternal education. They also considered gestational age, delivery mode, birth weight, and mother’s age at childbirth. This design did not prove that autism causes sleep changes. It showed which differences remained after those background factors were accounted for. The sample was large enough to see clear group contrasts in sleep timing and duration. It was also focused on preschoolers, a period with little prior evidence. Most earlier work had looked at school-age children or adolescents. The young age placed the comparison close to when families first seek help. In the autism group, most children were boys, at 84 percent. Mothers in that group were slightly younger and had lower educational attainment on average. These differences did not drive the sleep comparison because the analyses accounted for them. The recruitment period overlapped with pandemic disruptions, which may have shaped family routines. The cross-sectional design could not track how sleep changed over time.

Later wake-ups, shorter nights, and more frequent wakings

In parent reports from 492 preschoolers, children with autism went to bed later on weekdays. They also woke up about twenty-five minutes after the typically developing group. Their weekday sleep lasted roughly sixteen minutes less, even though bedtime and waking both shifted. The shift was not a small rounding error in the numbers; it meant that these children were getting less rest during school days. On weekends, the bedtime gap disappeared, but the later waking remained clear. Autistic children rose about thirty-three minutes after their comparison peers. Their weekend sleep was also shorter by about seventeen minutes. That pattern suggests the problem is not simply a busy weekday schedule; their entire sleep window moved later, placing more of it in the dark hours.

The middle point between falling asleep and waking came later on both weekdays and weekends. A measure of usual sleep timing was about forty-six minutes later on free days, which points to a pattern that leans toward the night. Autistic children were also more likely to wake during the night. Nearly one in five woke on most nights, compared with about seven percent of peers. After adjusting for age and family factors, the adjusted odds were 2.31 times as high. The length of each waking did not differ between the two groups. So the extra restlessness came from more interruptions, not longer ones. Taken together, these children had later nights, shorter sleep, and more waking.

What the sleep pattern adds to early autism care

The sleep pattern found in a matched comparison of 492 young children adds a practical clue to early autism care. It suggests that some sleep problems are not just side effects of diagnosis, but may be part of the condition itself. For families, that changes what clinicians should look for when a child is first evaluated.

A child who falls asleep late, wakes up late, and gets less total sleep may already be showing a pattern linked to autism. The same was also true for children who woke more often during the night. These are not small household complaints, because they can affect attention, learning, memory, and behavior regulation during the preschool years.

The timing matters because early intervention works best while the brain is still highly plastic. If sleep problems appear alongside diagnosis, treating them may support behavioral progress rather than simply adding another task for tired parents. A clinician who asks about bedtime, wake time, and night waking can spot a problem that might otherwise be missed.

This does not mean every child with poor sleep has autism, nor does it prove that sleep causes the condition. The comparison was made at one point in time, so it cannot show whether these patterns appear before diagnosis or how they change as children grow. It also relied on parent reports rather than laboratory sleep measures. Still, the pattern gives early care a concrete target: check sleep early, treat it when needed, and use that information to support development.

A sleep signal that deserves earlier attention

The matched comparison of 492 preschoolers points to a pattern that families may already feel in the bedroom. Children with autism fell asleep later, woke later, slept less, and broke their nights more often than age-matched peers. The timing shift was visible on weekdays and weekends, so it did not disappear when school pressure eased. That persistence matters because sleep is woven into attention, mood, memory, and behavior during a period when the brain is still taking shape.

The broader implication is that bedtime may be an early clinical sign rather than a background complaint. If sleep problems are present around diagnosis, clinicians could ask about them sooner, track them over time, and pair behavioral support with practical help for parents. The opportunity is real: better sleep might make daily routines easier and give children more room to learn social skills. The obstacles are also clear. The comparison relied on parent reports from one Chinese cohort, collected during unsettled pandemic routines. It could not follow children over time or measure sleep with devices. It also did not account for screens, medication, parental habits, or how severe each child’s autism was.

Still, the finding gives families and clinicians a concrete place to start. A later bedtime is easy to notice, but it may be more than a habit; it could be part of the condition that needs care from the beginning. If sleep is one of the first signals, will providers treat it as such before it hardens into years of exhausted nights?