Earlier this month, the typical media outlets were abuzz (“Childhood nightmares may point to looming health issues“) with the results of a newly published study linking early childhood nightmares and night terrors with future psychotic experiences. Expressing little in the way of skepticism, most reports simply regurgitated the University of Warwick press release. The research, published in the quite legitimate journal Sleep, is interesting but I’m not sure it tell us anything that we don’t already know. And it certainly doesn’t support any causal relationship between sleep disorders of any variety and “delusions, hallucinations, and thought interference”. But before we delve into the specifics of the paper, I believe a quick review of sleep, and sleep problems, in children is in order.

What is sleep?

To the outside observer, sleep appears as an altered level of consciousness where response to our environment and voluntary movements are noticeably decreased. But, with a certain degree of variability, the line between sleep and wakefulness is pretty thin. This distinguishes it from the increasing stimulation required to reverse other states of altered consciousness such as lethargy, obtundation, stupor and ultimately coma, which is not acutely reversible. I don’t plan on getting too technical, but there is obviously much more to sleep than that. Physiologically our metabolic demands drop a bit, and we enter a generalized anabolic or “growth” state during which a number of beneficial processes take place, predominantly, we think, involving the brain.

Sleep is a vital aspect of human life that has appears to have both physiological and psychological purpose, and is essentially universal in the animal kingdom. All you need to do is observe a cat for more than five minutes to see that we aren’t the only animal species that both needs and seemingly enjoys sleep. In fact, if you could talk to a nematode, it would likely go on for hours about how much it enjoys sleeping in on Sundays. Humans spend roughly a third of their lives asleep, but the percentage of each day devoted to sleep is significantly higher during infancy and early childhood.

We don’t know why the need to sleep became part of the blueprint for life so early on in our evolutionary history, and researchers certainly haven’t worked out all of the nuances of why humans and other animal species continue to be so dependent on it throughout the lifespan. It is likely that its purpose has broadened over time as species branched out into new environments. There are a number of leading hypotheses, however. And barring some amazing technological or medical advance, we appear to be stuck with sleep.

Why do we sleep?

Again, there are many proposed purposes for sleep. Much of the evidence relies on the observation of what happens to our bodies when we don’t get any. Poor sleep affects a number of physiologic processes, and there are probably consequences that we aren’t even aware of yet. Ultimately, a complete lack of sleep, although rare, is fatal.

Many of the hypotheses for the purpose of sleep involve its potential restorative powers, with proponents invoking evidence of increased clearance of waste products from the brain during sleep, as well as improved wound healing and immune system function. How do you truly boost your immune system? Sleep is probably a better means than any supplement or bogus healing modality and considerably cheaper. Sleep also may play a role in clearing out the cobwebs so to speak, by weeding out weak neuronal connections, allowing memory and learning to function optimally and supporting cognition. There are numerous studies detailing the effects of poor sleep on these processes.

Another school of thought involves the role of sleep in development. This stems from observations that babies sleep more than older kids and adults, and in particular active or REM sleep occurs for longer periods of the day in the very young in the vast majority of animal species. We also know that interference with REM sleep in the young can lead to a variety of negative development outcomes later in life. Perhaps this explains the origins of both muscle inhibition and dreams, as activating the brain without the risk of injury may be a pivotal aspect of proper development. Maybe REM sleep persists into adulthood as part of general brain maintenance. Who knows? But this doesn’t hold water in all species. Marine mammals don’t develop REM sleep until adulthood.

The weakest of the potential purposes for sleep appear to involve energy conservation and life preservation. Metabolic processes only slow down by about 5-10% during sleep, and animals that hibernate, a time of more significant drop in energy usage, still need to sleep afterwards. And decreasing consciousness with predators around seems like a risky means of taking a rest if that was all it was about. Avoiding predators by sleeping during their peak activity doesn’t explain why apex predators sleep or why sleep is still required the day after a sleepless night.

Clearly there is also a behavioral component to our desire to sleep. We get sleepy when we need sleep, although this doesn’t always translate to actually being able to fall asleep. At times our bodies face an almost overwhelming demand. It makes sense that this drive to sleep occurs because of some vital physiological importance, but sometimes we feel the urge to sleep for purely psychological reasons, even when we have consistently been obtaining a reasonable amount. There can be great pleasure in taking a nap on a day off. Sleepiness seems to occur with certain environmental cues as well, often accompanying boredom and sadness.

When asked about the reasons we sleep, William Dement, a sleep expert who was part of the early research on REM sleep, famously said “As far as I know the only reason we need to sleep that is really, really solid is because we get sleepy.”

Sleep problems in kids

Sleep is incredibly important to growth and development in children. Unfortunately sleep problems are very common, occurring across the age spectrum. They range from mild and occasional inconveniences accepted by most parents as normal aspects of childhood development to full blown disorders with a significant negative impact on the quality of life for patients and their loved ones. In general, anywhere from a quarter to half of children have difficulty with sleep at some point.

The structure and patterns of sleep are in flux throughout childhood. How many hours each day that we sleep, the specific percentages of REM and non-REM sleep, the duration of sleep cycles and how sleep is distributed over a day all change over time. So do the behaviors associated with sleep. Sleep problems manifest in a variety of ways during childhood as well.

An infant deprived of the recommended average of 15 hours of daily sleep over the first year of life (newborns often need 16-20 hours) may be irritable, feed poorly, and have delays in the acquisition of developmental milestones. And adding parental lack of sleep to the mix can be disastrous as it decreases their ability to handle stress appropriately. Lack of sleep in older children can result in learning difficulties and continued developmental challenges. Behavior, memory and attention problems can affect school performance. Even issues with weight gain and obesity have been linked to lack of sleep. And unfortunately, difficulties with sleep in childhood have often been found to be predictive of a wide range of future problems such as depression and anxiety, drug use, obesity and sleep disorders.

As in the adult population, there are a variety of sleep disorders found in children. The typical screening tool used by pediatricians is the BEARS, which consists of trigger questions for different age groups that can help reveal problems in one or more of five sleep related categories: Bedtime problems, Excessive daytime sleepiness, Awakenings during the night, Regularity and duration of sleep and Snoring. Probably the most common sleep disorder in kids is behavioral insomnia. Though it is extremely common for young children to wake up during the night, it can be a big problem if they have not learned how to fall back asleep without parental assistance and it occurs repeatedly each night. Behavioral insomnia also occurs, typically in older kids, because of a lack or inconsistency of limit-setting. These children simply refuse to go to bed. Again, it is common for there to be some resistance to calling it a night during childhood but at a certain point it becomes excessive.

Children can also suffer from disorders of the circadian rhythm, typically presenting in adolescence or young adulthood. The most common circadian rhythm problem is delayed sleep phase syndrome, where there is an inability to fall asleep and wake up during conventionally desired time periods. Sleep-related breathing disorders, in particular obstructive sleep apnea, is an increasingly common problem in children. This occurs when there is intermittent blocking of the upper airway, either partial or complete, that impedes airflow during sleep. It can lead to drops in the amount of oxygen in the blood and is associated with heart disease and hypertension in addition to behavior and learning difficulties.

The final category of sleep disorders commonly seen in children, and the one which the study on nightmares and future psychosis looked at, are the parasomnias. These are, at least to me, the most fascinating of the sleep disorders and certainly the group that gets the most coverage in the media. Parasomnias are discrete behavioral episodes that happen at any point during the sleep process, causing an interruption but generally not impacting the overall quality of sleep like insomnia or circadian rhythm disorders do. They do often lead to a great deal of parental distress however.

The parasomnias, which include nightmares, night terrors, sleepwalking, sleep talking, confusional arousals, bedwetting, teeth grinding, excessive or unusual movements and sleep paralysis are very common. There isn’t a lot of long term data, but one large cross-sectional study revealed that nearly 90% of children had at least one between the ages of 2.5 to 6 years of age. My 5-year-old daughter for example both talks in her sleep and often has confusional arousals during the night. It’s adorable. There appears to be a huge decrease in incidence once children reach school age but some parasomnias do persist or occur later in childhood and even in adulthood. Some only occur in adults. Studies on kids have primarily consisted of parental recall however, and mild events may not be recognized by parents. So we have to take the numbers in older kids with a grain of salt.

Night terrors

I could write an entire post just on these parasomnias, but for this post I’ll focus on nightmares and night terrors as they were focused on in the study. Night terrors occur during stage 3 and 4 non-REM sleep and are classified as a disorder of partial arousal along with confusional arousals and sleepwalking. They tend to happen as children are transitioning into lighter sleep or even into wakefulness.

There appears to be a genetic predisposition to having this type of parasomnia. One twin study that looked 323 pairs was very supportive, showing 6 times the incidence in identical versus fraternal twins, and another study revealed more than twice the likelihood of night terrors in the child of a sleepwalking parent. There are common triggers for these arousals as well, including obstructive sleep apnea, restless leg syndrome and reflux. Removing the trigger can stop the parasomnia. Anxiety, poor sleep and fever have also been shown to be triggers. The partial arousal parasomnias all typically occur during early childhood and gradually go away with time as the amount of slow wave sleep decreases, with very few teens and adults having them. EEG findings reveal unstable deep sleep.

Night terrors can be very impressive and scary for parents. The child appears to be sleeping soundly when suddenly they begin screaming. They are red faced and sweat profusely, reflecting the increase in autonomic nervous system activity. Their hearts are racing. They may even jump out of bed as if fleeing from some invisible monster and they are inconsolable or very confused and disoriented if awakened. Eventually, after 10-20 minutes of seeming terror, they fall back asleep and will have no recollection of the event in the morning. It’s no wonder that historically these events may have been attributed to demon possession or evidence of other paranormal activities.

We think that night terrors occur in anywhere from 1% to 6.5% of young children, and in about 2.5% after adolescence. Again, there appears to be a substantial genetic predisposition. They are usually infrequent and benign, perhaps occurring only once or twice a month, so parental reassurance is usually all that is needed. When causing a problem, the first step is making sure that the child is getting adequate sleep. Toddlers need naps during the day for instance but many aren’t given time to do so. Addressing other potential triggers is also an important step in managing significant night terrors. Finally, there are cases of severe refractory events, sometimes putting the child at risk of injury, that require a low dose sedative for a few months. Although there are only anecdotal reports, scheduled awakening prior to the typical time of the event, followed by reassurance and allowing the child to fall back asleep may help.


Nightmares are disturbing dreams, differentiated from night terrors by occurring during REM sleep, that lead to sudden wakefulness and emotional distress. They can involve a variety of intense emotions, such as sadness, anger, or disgust, not just the stereotypical fear and anxiety. Nightmares tend to occur during the early morning, as opposed to late evening with night terrors, and patients usually have good recall of the events of the dream although there ability to describe them varies with age. Also, unlike after night terrors, children with nightmares tend to achieve full wakefulness and have difficulty falling back asleep because of residual distress. There is an absence of symptoms such as flushing and sweating on awakening, but kids may have some mild increased heart rate, and excessive body movements during the nightmare are uncommon because of REM sleep inhibition.

Nightmares occur in a large percentage of children, and often can interfere with sleep and lead to parental awakening. Stress in general, and traumatic events can induce nightmares, so it shouldn’t be surprising that nightmares are more common in kids with PTSD. Similar to night terrors in this regard, poor sleep and anxiety can increase the incidence of nightmares. Unlike night terrors, there are some medications that can also increase the risk. Nightmares tend to peak between 5 and 10 years of age and then significantly decrease in incidence.

Reassurance is also in order with nightmares, and improved sleep hygiene. Simple techniques such as night lights and avoiding any potentially overstimulating television before bed can help, as can relaxation strategies. But with very frequent and severe events, there is potential association with anxiety disorders so a referral to mental health or a developmental-behavioral pediatrician may be necessary.

Nightmares, night terrors and psychotic experiences

The study making the rounds this month is a prospective birth cohort study using data from the Avon Longitudinal Study of Parents and Children (ALSPAC). A longitudinal cohort study, as opposed to a cross-sectional study, follows a group of subjects without a condition and who share common characteristic over time, sometimes even decades, to investigate various exposures or risk factors and their possible link to the development of a condition. The common experience of the cohort used for this study was being born in 1991 or 1992, and in or around Bristol, England. 6,796 children (3,462 girls, 50.9%) who completed a psychotic experiences interview made up the subject pool.

Although prospective cohort studies can be very helpful in telling if a certain risk factor, such as heavy smoking, plays a causal role in the development of lung cancer for example, they aren’t as effective as the randomized controlled trial. But they are a solid epidemiological tool even if there is some increased risk of missing a confounding variable. Of course RCTs aren’t always appropriate. Imagine assigning a study group to smoke a certain amount of cigarettes each day, or having a random assignment to the “no vaccines” arm of a trial. Only IRBs made up of family members and meeting in your basement would approve a study like that.

An advantage of a prospective cohort study is determining potential risk factors for a condition by following subjects over long periods of time. And when information is collected frequently over time, there is less need to rely on potentially biased recall of information. As with any prospective design, subjects can drop out, die, or disappear. And longitudinal cohorts running for decades take time and money, but are often worth it because of the improved reliability over studies that look back retrospectively or involve samples of subjects taken at one moment in time.

The authors compared subsets within the cohort, in this case children whose mothers reported frequent nightmares between age 2.5 and 9 years, to children who did not have frequent nightmare episodes during that period, and then 12-year-old children who self-reported nightmares, night terrors, or sleepwalking during the previous 6 months to those who did not. All of the children were asked about psychotic experiences at age 12 years. The goal of the authors was to examine any link between these parasomnias and psychotic experiences.

According to the authors, potentially confounding variables such as sex, family adversity, emotional or behavioral problems, IQ and potential neurological problems were accounted for, but I think some glaring potential confounders were missed. After analyzing the data, they concluded that “Nightmares and night terrors, but not other sleeping problems, in childhood were associated with psychotic experiences at age 12 years. These findings tentatively suggest that arousal and rapid eye movement forms of sleep disorder might be early indicators of susceptibility to psychotic experiences.” I appreciate their use of the word “tentatively”.

Also, the study does not tell us what having psychotic experiences at age 12 means in regards to future risk of mental health problems. The same authors, using the same cohort data set, actually addressed this in a 2013 study published in the American Journal of Psychiatry. In that paper they found a very poor positive predictive value of non-clinical psychotic experiences at age 12 years. Only 1.7% of 12-year-old subjects reporting psychotic experiences went on to meet criteria for a psychotic disorder at age 18 years. Sadly, only half of the 1.7% had received any mental health services.

Mothers were interviews roughly every 16 months between age 2 and 9 years, which seems like plenty of time for there to be some error in recall to creep in but it could be worse. Children whose mothers reported only one period of nightmares were 16% more likely to have psychotic experiences at age 12 years. Kids with 3 or more periods of nightmares were 56% more likely. Of the 12-year-old children asked about recent parasomnias, those with nightmares were 3.5 times more likely to have psychotic experiences while those reporting night terrors had twice the risk. The specific psychotic experiences they asked about were hallucinations, interrupted thoughts, and delusions and only 4.7% reported having one.

As I said in the intro, this study is interesting but I’m not sure if it adds much to what we already know. And naturally much of the news coverage is over-hyped and at times comes across as implying that somehow nightmares and night terrors might play a causal role in psychotic experiences in older kids, which is completely unfounded. There is also the implication that psychotic experiences at age 12 are a good predictor of psychiatric illness. The authors are absolutely not making that claim. But if this gets parents and healthcare professionals to pay a little more attention to sleep hygiene and pediatric mental health, I’m okay with that.

I feel like if a mother approached a pediatrician about her child having persistent severe nightmares or night terrors, especially if they had aged out of the typical peak years, most of us would consider the possibility that something more was going on, and not just mental illness. We would get a feel for the home and school situation and ask about risk factors for abuse, neglect, exposure to violence and poor sleep hygiene. We would also review potential worrisome indications for mental health issues. And I’d like to think that we would make the appropriate referral if necessary. But I don’t think that we should approach nightmares in the overwhelming majority of young children as a sign of mental illness.

Of course the major issue with pediatric mental health in this country is the abysmal lack of pediatric mental health in this country. There is often nowhere for poor kids to go for care, and often long waits for even those with money. Many get to the point of suicidal ideation or self-harm having never been evaluated. Suicide is an all too common cause of death in teens. Many don’t enter the system until they are forced to sleep in an emergency department for 3 days while waiting for a bed in an inpatient psychiatric facility.

Not surprisingly, the apparent relationship between parasomnias and psychotic experiences was stronger in the 12-year-old subjects who were currently having parasomnias. We already know that there is an association between stress and anxiety with parasomnias, and mental illness, or maybe just worrying about mental illness, certainly could be a strain. Many adolescent patients with schizophrenia for instance will self-medicate with marijuana. Marijuana can cause psychotic experiences. Some kids with schizophrenia are missed because their symptoms are blamed on the pot. As stated above, parasomnias almost always resolve by the time children are this age, so it stands to reason that there might be more to the situation at that point.

I realize that the authors attempted to account for potential psychosocial confounders, but this was based on self-report. I imagine that there might have been some under-reporting of some potential confounders, like marijuana use perhaps, or perhaps even under-reporting of psychotic symptoms. These children weren’t evaluated by a psychiatrist or other mental health professional. They just answered a list of questions.

We also know that there is a genetic predisposition at play with parasomnias, particularly night terrors. This also appears to be the case with many psychiatric conditions. This supports a possible link between the two issues. We also know that poor sleep is a factor in parasomnias. Sleep deprivation also plays a role in a variety of neuropsychiatric complaints. This wasn’t accounted for in the study questions. Neither was the possible use of illicit drugs, which could play a role in both conditions.

Psychotic experiences are actually fairly common in the 9-12 years age group according to the largest systematic review/meta-analysis to date. A median of 17% of kids in that age range compared to 7.5% of kids aged 13 to 18 years. They occur in about 5% of adults. Most kids who have these experiences in early adolescence won’t have them a few years later. Unfortunately, according to the authors of this review we should think more broadly because early psychotic symptoms may be a risk factor for much more than future psychosis. It should probably be considered a warning sign for future depression and other psychopathology. Still, most kids who have some thought interruptions or a hallucination or two at age 10 won’t go on to have any psychiatric condition. It probably represents some remnant of what would have been considered normal behavior a few years earlier.

I hope that parents won’t be losing sleep over their child having a nightmare or a night terror, or even bunch of them. The main author, Professor Dieter Wolke, stated in the press release, and I agree, that “We certainly don’t want to worry parents with this news; three in every four children experience nightmares at this young age. However, nightmares over a prolonged period or bouts of night terrors that persist into adolescence can be an early indicator of something more significant in later life.”


Sleep is an extremely important and often overlooked component of health in all age groups. But it is particularly important in kids because of the long-term implications of early sleep deprivation on development and a variety of other conditions. Pediatric healthcare professionals should be asking parents about their children’s sleep and be prepared to offer solutions for common sleep problems.

There may be a link between some sleep problems in kids, in this case nightmares and night terrors, and self-report of psychotic experiences in young adolescents. But it remains unclear what this means in regards to future diagnosis of true psychiatric illness. I don’t think that the study in question really should change what we do, and it should not lead to increased parental anxiety. Parents have enough to worry about.

There are a number of risk factors, such as family history and early environment, that serve as red flags for potential psychiatric illness in kids. Some of them also likely play a role in difficulties with sleep and the occurrence of parasomnias. And in some kids mental illness itself likely does cause sleep problems. It isn’t beyond possibility that some children with mental illness will present with a complaint of severe or persistent nightmares or night terrors. We should always consider the possibility of mental illness and ask the appropriate follow-up questions.



  • Clay Jones, M.D. is a pediatrician and a regular contributor to the Science-Based Medicine blog. He primarily cares for healthy newborns and hospitalized children, and devotes his full time to educating pediatric residents and medical students. Dr. Jones first became aware of and interested in the incursion of pseudoscience into his chosen profession while completing his pediatric residency at Vanderbilt Children’s Hospital a decade ago. He has since focused his efforts on teaching the application of critical thinking and scientific skepticism to the practice of pediatric medicine. Dr. Jones has no conflicts of interest to disclose and no ties to the pharmaceutical industry. He can be found on Twitter as @SBMPediatrics and is the co-host of The Prism Podcast with fellow SBM contributor Grant Ritchey. The comments expressed by Dr. Jones are his own and do not represent the views or opinions of Newton-Wellesley Hospital or its administration.

Posted by Clay Jones

Clay Jones, M.D. is a pediatrician and a regular contributor to the Science-Based Medicine blog. He primarily cares for healthy newborns and hospitalized children, and devotes his full time to educating pediatric residents and medical students. Dr. Jones first became aware of and interested in the incursion of pseudoscience into his chosen profession while completing his pediatric residency at Vanderbilt Children’s Hospital a decade ago. He has since focused his efforts on teaching the application of critical thinking and scientific skepticism to the practice of pediatric medicine. Dr. Jones has no conflicts of interest to disclose and no ties to the pharmaceutical industry. He can be found on Twitter as @SBMPediatrics and is the co-host of The Prism Podcast with fellow SBM contributor Grant Ritchey. The comments expressed by Dr. Jones are his own and do not represent the views or opinions of Newton-Wellesley Hospital or its administration.