Background
Autism Spectrum Disorder (ASD) is a complex neurodevelopmental disorder that seriously endangers human health. Deficits in social communication and repetitive and stereotyped interests and behaviors are its characteristics [
1]. The latest report in the United States shows that the prevalence was 1/54 in 2020 [
2]. In China, the prevalence has risen from 0.35% in 2018 [
3] to 0.7% in 2020 [
4]. It is one of the fastest-growing pediatric diseases and has attracted worldwide attention. Furthermore, ASD often accompanies sleep difficulties. Less than 50% [
5‐
8] of typically developing children experience sleep disorders, but 50–80% [
9‐
12] of ASD children. Common sleep problems are trouble falling asleep, decreased sleep duration, sleep onset delay, and night waking [
13‐
15]. Other researchers have found high rates of bedtime resistance and daytime sleepiness [
16].
These sleep problems can cause a series of effects. Sleep disorders may adversely affect children’s daily function, affecting behavior, learning, memory regulation and cognition [
17‐
20]. It may also cause emotional problems such as aggression, irritability, over-reactivity and depression [
21,
22]. Moreover, disorders also negatively impact ASD symptoms. For example, ASD children with sleep problems showed more severe social skills deficits, and they scored lower on social tests [
10,
15,
23,
24]. Other studies have found that core symptoms also affect sleep, such as communication difficulties may exacerbate sleep disorders. All these problems not only reduce the quality of life, but also influence the effect of intervention. These connections all illustrate the severity of sleep problems in ASD. To date, most of these connections have focused on the general concept of “sleep disorders” or “duration of sleep”, and few have linked other dimensions of sleep problems to ASD symptoms. The same problem exists when discussing the relationship between developmental level and sleep disorders. Many children with ASD also suffer from intellectual disability (33% of individuals with ASD have IQ scores of 70 or less), and 24% in the borderline range [
2]. Research results on whether developmental level or intelligence affect sleep are also inconsistent [
15,
25‐
29].
Behavioral insomnia in children is considered the most common cause of sleep disorders in children. Also in ASD. So far, behavioral interventions, sleep education, environmental changes and exogenous melatonin are the most effective ways to promote sleep in ASD [
30]. The Autism Treatment Network (ATN) proposes to behavioral strategies, parent education about environmental modification, and positive bedtime routines as the first-line approach [
31]. However, in China, there are few epidemiological studies on sleep in ASD children. Due to cultural, geographic, and resource constraints, do Chinese children with ASD have more sleep behavior or hygiene problems? Therefore, for the first time, we conducted a large sample, multi-center cross-sectional survey, combined with a case-control study. The main purpose of this study has two, one is to investigate the prevalence of sleep problems and common sleep problems in ASD children in China, the other is to explore the correlation between common sleep problems and ASD core symptoms and developmental levels in Chinese ASD children. Our research helps to understand the sleep conditions of Chinese ASD children and provides a basis for precise intervention in the future.
Discussion
At present, this is the first large-scale study on sleep conditions in Chinese ASD children. The 13 cooperative centers have the ability to standardize the diagnosis and treatment of ASD, which can better reflect the sleep conditions of children with ASD. Primarily, the prevalence of sleep difficulties in ASD was 67.4%, which is consistent with previous reports, significantly higher than that in TD (51%), and the scores and prevalence of sleep problems in all dimensions were also higher. These emphasized the severity of sleep problems in ASD. Further analysis found that the first four were bedtime resistance (25.6%), anxiety (22.7%), sleep onset delay (17.9%) and daytime sleepiness (14.7%). Except for these four dimensions, in previous reports, night waking is also a common sleep problem reported by parents, but it was the lowest in our study (5%). This sleep pattern is similar to a recent survey in China and India. China investigated the sleep conditions of 475 preschool children with ASD and found that the prevalence of sleep problems was 81.7%. The four with the highest were sleep resistance (90.9%), sleep anxiety (91.7%), daytime sleepiness (60.7%) and sleep onset delay (59.1%), and the lowest two were night waking (25.4%) and Sleep disordered breathing (19.8%) [
42]. In India, the prevalence in 2 to 6 years old children with ASD was 93%. The four with the highest prevalence were sleep resistance (95%), sleep anxiety (85%), sleep duration (81%) and sleep onset delay (66%), the lowest two were night awakening (50%) and daytime sleepiness (27%) [
43]. Although they have similar sleep patterns to ours, the prevalence is significantly different. By comparison, these studys defined these types of sleep problems as a sub-scale score above the respective cut off, but did not indicate the cut off value clearly. We used M ± 2SD as the cutoff value, which has been used in other articles. Combined with domestic and foreign research, we found that ASD has a higher detection rate of sleep problems, but the results of specific dimensions are different. This may be related to the survey environment, scale, living habits, survey methods, etc. In fact, there is no consensus on the definition and classification of sleep difficulties in the world and few studies have calculated the prevalence of each dimension in CSHQ, so it is not convenient for us to compare each dimension here. But in general, the most common sleep problems are all difficulties in falling asleep, shortened sleep time, delayed sleep onset and awakening at night. It reminds us that we should pay more attention to these problems when discussing sleep conditions of Chinese ASD children, and we can also explore the sleep patterns of ASD more specifically and comprehensively in future studies.
There are currently a large number of studies on the overall sleep status and autism symptoms of ASD, so we do not separately analyze and discuss it, but focus the research on different dimensions. For this study, it is the four dimensions of with the highest prevalence [
44‐
46].
The sleep problem with the highest prevalence of ASD in this survey was bedtime resistance (25.6%). Children with it had higher SRS scores, which were mainly manifested in social awareness, cognition and communication. Hollway [
47] analyzed data of 1583 ASD children and found that social interaction deficits were related to bedtime resistance, and Asians had a higher risk of sleep resistance than other races. Our results echo it. At present, behavioral insomnia of childhood is recognized as the most common cause of sleep disorders in children. It may be related to poor sleep training or environmental restrictions imposed by parents or caregivers, including sleep onset association type (children’s specific dependence on stimuli, people, objects, or settings to start or return to sleep) and limit setting type (behavior of delay or refusal before going to bed due to the difficulty of setting the limit by the caregiver). Some limit setting insomnia of childhood can be manifested as resistance at bedtime [
48‐
51], and a recent study found that sleep hygiene was positively correlated with the resistance [
52]. Therefore, due to the behavioral characteristics of ASD children and cultural differences, ASD children in China and even Asia have more sleep hygiene or sleep behavior problems, which leads to a high prevalence of sleep resistance. For example, co-sleeping is common in many Asian countries [
43,
53,
54]. Although children had lower ABC language and CARS scores when there was bedtime resistance, there was a trend of higher scores in other sub-scales. CARS are used to measure ASD severity which is more related to biological mechanisms, while bedtime resistance is more likely to be caused by sleep behavior and sleep hygiene. It may be the reason for the opposite trend here of CARS. Anyway, their relationship needs further study.
In this study, ASD children with sleep onset delay and sleep anxiety had higher ABC, SRS and autism warning behavior scores, which were specifically manifested in higher social interaction, communication, cognition, language and stereotyped behavior scores. Tudor [
55] found that autism symptoms and severity were associated with short sleep duration and sleep onset delay. Communication symptoms can be predicted by sleep anxiety, and autism severity, stereotyped behaviors and social interaction deficits can be predicted by sleep onset delay. Hollway’s [
47] research showed that sleep anxiety was related to abnormal taste and smell perception. Our research further found that these two may be related to social interaction, communication, cognition, language and stereotyped behavior. Special attention should be paid to the sleep problems caused by language disability. Children with autism are often unable to express some demands or discomfort due to their language barriers, which may indirectly aggravate their sleep problems. By this, for ASD children with language impairment, we must pay more attention to their sleep. Earlier we mentioned that there is no consensus on the definition of many sleep problems in the world, one of which is sleep anxiety. Sleep anxiety in CSHQ including afraid of sleeping in dark, afraid of sleeping alone, trouble sleeping away and needs parents in the room to sleep. These cannot fully reflect the anxiety state of children when they fall asleep. A survey of ATN that included 1784 from childhood to adolescence ASD showed that with age increase, the prevalence of sub-items related to sleep anxiety in CSHQ showed a downward trend [
56]. Therefore, in future research, we need to figure out whether children have sleep anxiety and may need to observe it by age.
The daytime state is closely related to the study and function during the day. Our research found that daytime sleepiness was related to the core symptoms of autism and children showed higher scores in ABC, SRS, and CARS. A recent study in India found that the severity of autism was positively correlated with daytime sleepiness in male children between 2 and 6 years old [
43]. Another study found that daytime sleepiness in toddlers was related to autism characteristics. Toddlers with autism characteristics had a higher prevalence of daytime sleepiness than those without autism characteristics [
57]. Hodge [
58] reported that the typical developing children of 3–17 years old had a significant decrease in daytime sleepiness with age, but the daytime sleepiness of ASD children still increased. Therefore, daytime sleepiness may be a symptom that can detect children’s ASD early, and we should pay attention to children’s daytime sleepiness. In addition to poor sleep behavior and sleep hygiene, sleep disturbance involves more complex neurobiological mechanisms, especially the abnormal melatonin level. Compared with normal children, ASD had significantly lower levels of melatonin and metabolites, and there were decreased at night, increased during the day, and delayed circadian rhythm [
59‐
65]. This partly explains their delayed onset of sleep, night waking and daytime sleepiness. The underlying biological rhythm and behavioral characteristics of ASD may make them more vulnerable to sleep disorders.
In addition, Gender-stratified analysis showed that ASD children with common sleep problems exhibited higher ASD symptom scores both in male and female, but some results were not statistically significant in female. This may be caused by the small sample size of the girls group. The analysis found that the sample size of boys was 1077 and that of girls was 233.
These four types of sleep problems were found to be related to core symptoms, but not related to the neurodevelopmental level. Previous studies on the relationship between developmental level and core symptoms are inconsistent. Several studies reported that the neurodevelopmental level or intelligence may affect sleep [
26,
27,
29]. One research found that intelligence has an effect on sleep anxiety to a certain extent, but has no effect on bedtime resistance and sleep duration [
47]. Another study found that intelligence only affects night waking [
66]. In contrast, sleep disorders can occur at all neurodevelopmental levels in most studies [
15,
25,
27,
28]. The reason for the inconsistency of these results may be that most sleep assessments are based on subjective reports, lack of objective sleep data support, and there are certain differences in the age of children included in different studies. More research is needed to clarify the relationship between sleep disorders and developmental trajectories.
In addition to helping to understand the sleep conditions of Chinese children with ASD, our research also provides directions for precise intervention in the future. Bedtime resistance was a sleep problem closely related to sleep behavior and sleep hygiene, and the most prominent sleep problem in this survey. Daytime sleepiness and sleep onset delay were likely to be related to abnormal melatonin levels. This suggests that in the future intervention, the cultivation of sleep hygiene and behavior is as important as the intervention of biological factors, especially starting from the actual situation of our country, combining our country’s culture and resources to formulate intervention strategies. Moreover, there is still a clinical phenomenon that children with ASD often come to the doctor because of language delay. Parents generally pay more attention to the core symptoms of the child, thus ignoring the co-occurring sleep problems which are related to core symptoms. Therefore, we suggest that in the future diagnosis and treatment, sleep assessment should be routinely included to achieve early prevention, detection and intervention of sleep disorders in ASD children.
This study has limitations. First, Children are recruited online rather than randomly sampled, which may make the sample less representative, and it was designed as a cross-sectional study when comes to the relationship between sleep disorders and core symptoms, which could not establish a causal relationship. Moreover, there was no objective measure of sleep. The CSHQ was completed by parents. This could led to a response bias. So there may be deviations between the real sleep situation and the survey results. It is better to combine subjective and objective assessments when possible. Although the scale assessment was completed by trained professionals, this study involved 13 cities in China, and the quality control was not mature enough, which may also be the reason for the contrary results in the analysis of sleep resistance. Therefore, further objective research is needed.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.