Assessment of Serum Melatonin Levels, Sleep Patterns, and Clinical Symptoms in Children With Autism Spectrum Disorder: A Case-Control Study
Article information
Abstract
Objective
This study aimed to investigate the relationship between serum melatonin levels, sleep habits, and clinical features in children with autism spectrum disorder (ASD) compared to healthy controls.
Methods
A total of 71 children, aged 2–8 years, including 38 with ASD diagnosed according to Diagnostic and Statistical Manual of Mental Disorders-Fifth Edition criteria and 33 age- and sex-matched healthy controls, were evaluated. Serum melatonin, vitamin D, ferritin, serum iron, and iron-binding capacity were analyzed. Sleep habits were assessed using the Pediatric Sleep Habits Questionnaire, while the Autism Behavior Checklist and Modified Checklist for Autism in Toddlers were administered to the ASD group. Relationships between biochemical markers and questionnaire scores were analyzed.
Results
The mean age was 44.4±20.4 months in the ASD group and 51.2±20.0 months in the control group (p=0.104). The ASD group exhibited higher “bedwetting” scores, while the control group had higher “daytime sleepiness” scores (p=0.008, p=0.036, respectively). Serum melatonin levels were significantly elevated in the ASD group (823.2±237.9 U/L) compared to controls (677.4±254.7 U/L, p=0.027), with this difference significant in males (p=0.020) but not in females (p=0.608). No significant correlations were observed between melatonin levels and questionnaire scores.
Conclusion
Elevated daytime melatonin levels and altered sleep patterns in children with ASD suggest potential melatonin receptor desensitization. Sex-specific variations underline the importance of personalized melatonin-based interventions. These findings provide insights into developing tailored therapeutic strategies for managing sleep and behavioral challenges in ASD. However, future studies are needed to explore these findings further with larger and more diverse populations.
INTRODUCTION
Sleep disorder affects the daily lives of the cases negatively due to cognitive functions such as working memory and attention. One of the most frequent complains of individuals with autism spectrum disorder (ASD) and their families is sleep disorders, ranging from 40% to 86% of the patients [1]. In ASD cases, it was shown that the serum melatonin level and the level of melatonin metabolites in the urine are low [2]. The concentration of melatonin is regulated by the circadian rhythm under physiological conditions, which is low during the day and high at night [3]. The peak secretion is usually around 2:00 AM and is also called the hormone of darkness. Polymorphism in acetylserotonin O-methyltransferase may be related to low melatonin in ASD [4]. The melatonin receptor agonists can treat insomnia and behavioral symptoms by reducing nitrosative/oxidative stress and inflammation in ASD. Melatonin levels also differ by sex, usually 2–3 times higher in girls than boys. Therefore, ASD may be less common in girls (as it prevents oxidative damage to DNA). In a study conducted with 14 ASD cases, the 24-hour circadian rhythm concentration of serum melatonin was examined, and it was found to be lower at night in ASD compared to healthy controls. No circadian variation was detected in the 10–14 age range. In the 4–14 age range, inverted rhythm, e.g., the opposite of normal rhythm, was found [5].
The aim of this study is to evaluate the relationship between serum melatonin levels, sleeping habits, and clinical symptoms in children with ASD through a case-control design. We selected this design to allow direct comparisons between children with ASD and age-matched typically developing peers, controlling for key confounding factors. Additionally, by focusing on both sleep patterns and melatonin levels alongside clinical severity, we aim to address gaps in the current literature. While previous studies have highlighted the association between melatonin and disrupted sleep in ASD [6,7], relatively few have employed a robust case-control framework to clarify the role of melatonin in modulating core ASD symptoms. Our approach not only extends prior findings but also provides a novel perspective by incorporating detailed clinical assessments and age-specific evaluations, thus offering new insights into the interplay between melatonin, sleep, and ASD symptomatology.
METHODS
This study was approved by the Non-Invasive Clinical Research Ethics Committee of Çukurova University Faculty of Medicine at its meeting dated July 9, 2021, with decision number 17. Informed consent was obtained from all subjects involved in the study. The study was evaluated and deemed to comply with ethical principles and relevant regulations. Our study included; the cases aged 2–8 years who applied to our outpatient clinic between August 2021 and January 2022 and were diagnosed with ASD according to Diagnostic and Statistical Manual of Mental Disorders-Fifth Edition (DSM-5), and a healthy control group matched for sex and age. Exclusion criteria are those who have medical diseases known to affect cognitive and brain function, cancer, bipolar disorder, psychotic disorders, epilepsy, having any diagnosis of neurological diseases, having received electroconvulsive therapy, intraocular surgery or laser therapy in the last 3 months, ocular diseases such as retinitis pigmentosa, traveling between continents in the last 1 month, sleep apnea and sleep disorders due to a medical condition, autoimmune diseases, severely impaired hearing or severe mental retardation, who use psychotropic, neurotropic, selective serotonin reuptake inhibitor, anti-psychotic, serotonin-norepinephrine reuptake inhibitor, mood stabilizer, stimulant and benzodiazepine drugs and other drugs known to change sleep patterns.
Inclusion criteria for the ASD group were children aged 2–8 years with a clinical diagnosis of ASD based on DSM-5 criteria, without any additional uncontrolled or severe comorbid neurological or psychiatric disorders (e.g., active epilepsy, severe anxiety disorder) that could interfere with sleep or cognitive assessments. For the control group, age- and sexmatched children with no history of ASD or any significant neurological or psychiatric disorders were included. Additional exclusion criteria for both groups consisted of chronic medical conditions affecting the central nervous system, ongoing treatment with medications known to alter sleep patterns or cognitive function, and any acute or uncontrolled chronic illnesses. Specifically, children diagnosed with epilepsy or significant anxiety disorders under ongoing pharmacological treatment were excluded to minimize confounding factors.
Psychometric measurements
The diagnostic criteria of DSM-5 were used along with the Autism Behavior Checklist (ABC), the Modified Checklist for Autism in Toddlers (M-Chat), and the Children’s Sleep Habits Questionnaire (CSHQ).
Sleep assessment
In this study, sleep patterns were evaluated using the CSHQ. CSHQ is a widely used parent-reported instrument designed to assess common sleep problems in children, covering domains such as bedtime resistance, sleep duration, and night wakings. It consists of 33 items rated on a 3-point scale (usually, sometimes, rarely) [8]. The CSHQ has been utilized in various pediatric populations, including children with ASD, and has demonstrated acceptable reliability and validity in this group [9,10]. For our study, parents or primary caregivers of the participants filled out the CSHQ to document the child’s sleep behaviors over the previous week. Prior to data collection, the research team provided guidance on how to complete the questionnaire to ensure consistent understanding and minimize response bias.
Laboratory data
Vitamin D level, ferritin, serum iron and iron binding capacity, as well as endogenous melatonin levels, which are routinely requested from patients diagnosed with ASD, were examined. Blood samples were collected in ACD-A tubes (Becton Dickinson) between 8:30–10:30 in the morning (background light intensity <2 lx, participants were allowed to sleep between 23:00 and 6:00 in the evening). In addition, the function of the pineal gland can be roughly evaluated in studies where melatonin metabolites are evaluated non-invasively. Therefore, it was decided to measure blood melatonin levels in our study. For the measurement of serum melatonin level, 5 mL blood sample was taken from the brachial vein of the forearm and was centrifuged. The obtained serum samples were stored as 2 mL at -80°C. All patient blood samples and control group blood samples were taken into Becton Dickinson vacuum-gel tubes. Serum was obtained by centrifuging them at 2,000 rpm for 20 minutes. Serum melatonin (SUNRED brand BT LAB Cat.NO EA0013Ge) level was measured by enzyme-linked immunosorbent assay in accordance with kit protocols. Serum samples were diluted according to the kit procedure. Serial dilutions were made according to the standards of kit procedure. Fifty microliters of standard serum or 40 μL of patient and control serum were added to each determined well. Ten microliters of Biotin Antigen was added to the samples. They were incubated at 37°C at the specified time intervals in accordance with the procedure. Fifty milliliters of stop solution was added. It was immediately read at 450 nm. The resulting concentration values were multiplied by the dilution coefficient. Serum melatonin level concentration was found. Serum melatonin levels were measured as pg/mL. In addition, chemiluminescence method was used with DXI 800 device (Beckman Coulter) for ferritin, photometric method was used with AU5800 device (Beckman Coulter) for iron and iron binding capacity, high-performance liquid chromatography method was used with Thermo Dionex device (Thermo Fisher Scientific) for vitamin D. Melatonin levels of patients diagnosed with ASD and healthy controls were compared.
Statistical analysis
Statistical analysis of the data was performed using the SPSS 25.0 (IBM Corp.) program. Normality test of numerical variables was determined by Kolmogorov–Smirnov test. Numerical variables were summarized as mean (±standard deviation). Categorical variables were given as numbers and percentages. When comparing numerical variables between the two groups, the Mann–Whitney U test was used because the variables did not show normal distribution. Chi-square and Fisher’s exact tests were used when comparing categorical variables. The correlation between the measurement variables was compared with the Pearson correlation test and their correlation coefficients were calculated. Statistical significance level was taken as 0.05 for each test. To minimize potential confounding variables, we clearly defined the exclusion criteria related to medication use (e.g., psychotropic, neurotropic, or sleep-altering drugs) and compared key laboratory parameters (e.g., melatonin, vitamin D) among ASD cases classified by severity (mild, moderate, severe). Additionally, other factors such as age, sex, and active comorbid conditions were controlled through matching in the control group or exclusion where applicable. This multifaceted approach allowed us to evaluate the relationships between melatonin levels, sleep patterns, and ASD clinical symptoms in a valid and reliable manner. In our study, G*Power software (version 3.1.9.7; Heinrich Heine University) was used for sample size planning to detect a specific effect size (Cohen’s d) between the ASD (n=38) and control (n=33) groups. Assuming a two-tailed t-test for two independent groups, with α=0.05 and Cohen’s d=0.8 (large effect size), the sample sizes of 38 and 33 participants in our study were calculated to provide a statistical power of 91.2%. This result indicates that our sample size is sufficient to detect the anticipated effect size.
RESULTS
Thirty eight patients with ASD and 33 healthy children evaluated in the control group were included in the study. In this study, the mean age of children diagnosed with ASD was 44.4±20.4 months, which did not differ significantly from the control group’s mean age of 51.2±20.0 months (p=0.104). The ASD cohort exhibited a higher proportion of males (86.8%) compared to the control group (75.8%), although this difference was not statistically significant (p=0.228). Notably, serum melatonin levels were significantly elevated in the ASD group, averaging 823.2±237.9 U/L compared to 677.4± 254.7 U/L in the control group (p=0.027). This elevation was particularly evident among male participants, with ASD males showing melatonin levels of 820.8±243.5 U/L versus 649.0± 262.8 U/L in control males (p=0.020), while no significant difference was observed among females (838.6±221.8 U/L in ASD versus 766.2±254.7 U/L in controls, p=0.608). Additionally, children with ASD reported lower daytime sleepiness scores (3.2±1.2) compared to the control group (3.8±1.3, p=0.036) and exhibited higher bedwetting scores (3.3±1.3 versus 2.5±0.8, p=0.008). These findings highlight significant differences in melatonin levels and specific sleep-related behaviors between children with ASD and their healthy counterparts (Table 1).
Demographic, melatonin levels, laboratory data and children’s sleep habits questionnaire among groups
Patients with ASD had a mean score of 5.3±2.9 on the risky questions section of the M-Chat, with a total M-Chat score averaging 13.2±4.8. The ABC assessment revealed that these patients scored 11.2±5.9 in the sensorial domain, 15.9±8.1 in building relationships, and 16.6±9.5 in the use of body and object. Language skills were assessed with a mean score of 14.1±5.8, while social and self-care abilities scored 12.6±4.7. The overall ABC score for the ASD patients was 70.4±25.3. These results indicate significant challenges in multiple domains of behavior and development in children with ASD (Table 2).
M-Chat and ABC assessment results of patients with autism spectrum disorder
Patients with severe ASD exhibited a significantly lower iron binding capacity, with a mean value of 153.5±181.2 g/dL, compared to 296.2±94.7 g/dL in the mild group and 299.3±92.9 g/dL in the moderate group (p=0.043). Additionally, vitamin D levels were markedly lower in the severe ASD group, averaging 3.3±4.3 ng/mL, as opposed to 26.2±15.7 ng/mL in the mild group and 17.9±8.6 ng/mL in the moderate group (p=0.005). Although melatonin levels and other hematological parameters (such as erythrocyte/red blood cell [RBC], white blood cell [WBC], hemoglobine [HGB], hematocrit [HCT], mean corpuscular volume [MCV], and platelet [PLT]) did not show statistically significant differences across the clinical severity groups, the trend indicates varying degrees of impact on these laboratory markers depending on the severity of ASD (Table 3).
Comparison of laboratory data of patients with ASD according to their clinical status
The correlation analysis between melatonin levels and various clinical and laboratory parameters in children with ASD revealed no significant correlations. Specifically, melatonin levels showed a weak positive correlation with hemoglobin levels (r=0.196) and HCT (r=0.201), although these were not statistically significant. The analysis also indicated negligible correlations between melatonin and vitamin D (r=-0.062), ferritin (r=0.010), and the scores from the M-Chat (r=0.063) and ABC (r=-0.066). These findings suggest that melatonin levels are not strongly associated with these parameters in the ASD population studied (Table 4).
Correlation analysis of melatonin and other parameters with autism spectrum disorders
When examining melatonin levels, some values in the control and ASD groups were observed to be relatively higher or lower compared to each other. These outliers may stem from individual differences in participants’ circadian rhythms, the timing of melatonin measurement, or the heterogeneity of the sample population. To minimize the impact of outliers on statistical analysis, normality checks (Kolmogorov–Smirnov test) were conducted, and appropriate non-parametric tests, such as the Mann–Whitney U test, were applied for non-normally distributed data. Additionally, inclusion and exclusion criteria for participants helped to partially limit the effect of extreme outliers.
DISCUSSION
In the ASD patients included in our study, the male sex ratio was observed as 86.8%, and the male sex ratio was 6.6 times higher. The differences in ASD according to sex are thought to be due to genetic characteristics [11]. In our study, RBC count, WBC count, HGB, HCT, MCV, and PLT values were found to be quite similar between children with ASD and healthy children (p>0.05). Vitamin D level and iron binding capacity; were found to be significantly lower in the severe ASD group compared to the other groups (p=0.005 and p=0.043, respectively). In our study, when healthy children and children with ASD were compared in terms of sleep problems, it was observed that the total sleep duration was similar between the two groups (p=0.479). In terms of staying awake at night, it was determined that patients with ASD had significantly longer awake times, but this difference was not statistically significant (p=0.052). There were no significant differences between the ASD and control groups in terms of difficulty in waking up in the morning, sleep fragmentation, sleep anxiety, sleep disturbance, parasomnia, morning awakening style, sleep duration, sleep transition, and the need to sleep with others (p>0.05). It was determined that the sleepiness problem experienced during the day was observed less frequently in ASD patients compared to the control group, but bedwetting problems were observed more frequently (p=0.036 and p=0.008, respectively). It was determined that melatonin levels are high in ASD patients and melatonin levels decrease as the clinical condition worsens. In addition, in this study, although melatonin levels were lower in measurements made at night than in the control group, it was found that melatonin levels were higher in measurements made during the day than in healthy children [12].
In the seven different studies compiled in a meta-analysis study examining the relationship between ASD and melatonin, melatonin levels were found to be lower in ASD patients than in healthy individuals [13]. In addition, four different studies have shown that melatonin levels decrease as the clinical condition of ASD gets worse [11]. There are studies claims that the potential to lead to an effective and relatively simple treatment and maybe prevention procedure for ASD, using exogenous melatonin even for infants [14,15]. However, almost all of these studies measured melatonin levels at night. In the study of Ritvo et al. [16], it was observed that melatonin levels measured during the day were higher in patients with ASD. In our study, it was observed that serum melatonin levels were significantly higher in ASD patients than in healthy controls (p=0.027). In studies where melatonin levels were measured at night, lower melatonin levels were found in pa-tients with ASD, while elevated melatonin levels were observed in patients with ASD in daytime melatonin levels [17]. This is an indication that the circadian rhythm of melatonin is impaired in patients with ASD. Although there are still missing data today, some genes that cause disruption of melatonin synthesis are thought to play a role in the etiology of ASD. In addition, the presence of some gene mutations that play a role in melatonin receptor desensitization is also accused in the etiology of ASD [18].
Previous studies report that melatonin levels in individuals with ASD are generally irregular, negatively impacting the sleep-wake cycle [19,20]. Furthermore, reductions in endogenous melatonin levels or disruptions in its rhythm have been shown to play a significant role in initiating and maintaining sleep in individuals with ASD [21,22]. Meta-analyses have revealed that melatonin supplementation has particularly beneficial effects on sleep duration, sleep latency, and sleep efficiency [23]. Similarly, the literature reports that 67%–89% of children with ASD experience various sleep disorders, such as short or fragmented sleep, prolonged sleep onset latency, and frequent nighttime awakenings, which in turn exacerbate ASD symptoms like daytime irritability and stereotypic behaviors [24,25]. Additionally, it has been suggested that low maternal melatonin levels during the intrauterine period may increase the fetus’s susceptibility to ASD [13] and that sex-based differences in melatonin levels (e.g., relatively higher levels in females) may contribute to the higher prevalence of ASD in males [26]. However, environmental factors, such as seasonal transitions (particularly the transition from winter to spring), are reported to have the potential to exacerbate sleep disturbances, including late bedtimes and fragmented sleep patterns [27]. In our study, while total sleep duration in the ASD group was found to be similar to that of the control group, significant differences were observed in specific subdomains, particularly daytime sleepiness and nocturnal enuresis. These findings highlight the impact of endogenous melatonin levels and circadian rhythm on the specific sleep habits of children with ASD. Furthermore, the observation that melatonin levels in male ASD participants were higher compared to the control group appears to align with previous findings regarding sex-based melatonin differences. In conclusion, our findings support the view emphasized in the literature that “melatonin deficiency or disruptions in its rhythm may exacerbate sleep problems accompanying ASD” and demonstrate that sleep disturbances are a significant modifying factor in the clinical course of ASD.
Melatonin receptor desensitization, which is thought to be observed in ASD patients, may also play a role in the ASD clinic. The fact that the daytime melatonin levels were increased in our study is in parallel with the literature. There is no existing study examining the sex-related melatonin values in ASD patients in the literature. Studies in healthy people have shown that women’s melatonin levels are higher than men, and the difference can sometimes be up to 3 times higher. In our study, melatonin levels were found to be significantly higher in male patients between ASD patients and the control group (p=0.020). It was observed that melatonin levels were higher in girls with ASD, but this difference was not significant (p=0.608). High melatonin levels observed in women are thought to be a protective factor against ASD. In particular, it has been found as a data that has been revealed in previous studies that the circadian rhythm of melatonin works differently in ASD patients and that it is an inverted rhythm.
Limitations of the study
One of the biggest limitations of our study is that children’s sleep habits in children were evaluated with a questionnaire method that provides subjective data. In addition, at the time of blood analysis, the number of hours the participant slept, what time he woke up in the morning, or what time he slept at night was disregarded. Since there is less female sex among the applications, in order to conduct future studies with multicenter, cohort and large study groups, with female-weighted participants, and to evaluate the sleep of the children studied more objectively, it is thought that more precise results can be obtained by performing sleep monitoring. Additionally, the relatively small sample size and the cross-sectional design of our study, where participants were assessed at a single time point, limit the ability to establish causal relationships. Potential confounding variables, such as medication use, existing comorbidities, and environmental factors, may have also influenced our results. Although we implemented strict exclusion criteria, it was not possible to fully control for all confounding factors.
Sleep disorders are a common issue in children with ASD and can become more complex due to sensory sensitivities (e.g., light, sound, touch) [28]. In clinical practice, priority should be given to improving sleep hygiene (e.g., maintaining consistent sleep/wake schedules, appropriate lighting, reducing electronic device usage), minimizing sensory stimuli (e.g., earplugs, dark rooms), and providing sleep education for families. Cognitive-behavioral therapies or social story interventions can support the child’s adaptation to a sleep routine. Pharmacological approaches, such as melatonin, may be beneficial in certain cases; however, the presence of comorbid conditions such as depression or anxiety must always be considered. When sleep problems are severe and resistant to treatment, comprehensive evaluation and alternative treatment options (e.g., low-dose medication therapies) should be considered in collaboration with a child psychiatrist and/or pediatrician.
Conclusion
Our study contributes a unique perspective by enrolling participants who do not use medications known to affect sleep or melatonin, thereby offering a clearer view of the endogenous melatonin-sleep interplay in children with ASD. While prior research has established a significant relationship between ASD and melatonin levels—often involving an inverted oscillation rhythm and a potential link between sleep habits and ASD severity—our findings underscore the importance of considering medication-free samples to accurately assess these associations. From a clinical standpoint, recognizing elevated or dysregulated melatonin levels and related sleep disturbances in ASD may prompt earlier interventions— such as tailored sleep hygiene programs, monitoring of circadian patterns, and targeted behavioral therapies— to improve both sleep quality and core ASD symptoms. Additionally, these findings may guide future research toward larger-scale, longitudinal studies that include 24-hour serum melatonin measurements and more sex-balanced cohorts. Such efforts could yield robust insights into the circadian regulation of melatonin in ASD and inform more nuanced treatment strategies, ultimately enhancing the quality of life for individuals with ASD and their families.
Notes
Availability of Data and Material
All the data can be requested from the corresponding author upon reasonable request.
Conflicts of Interest
The authors have no potential conflicts of interest to disclose.
Author Contributions
Conceptualization: Çağlar Charles Daniel Jaicks, Perihan Çam Ray. Data curation: Perihan Çam Ray, Gonca Gül Çelik. Formal analysis: Özlem Görüroğlu Öztürk, Yusuf Döğüş. Funding acquisition: Çağlar Charles Daniel Jaicks. Investigation: Ayşegül Yolga Tahiroğlu, Zeliha Haytoğlu. Methodology: Özlem Görüroğlu Öztürk, Perihan Çam Ray. Project administration: Çağlar Charles Daniel Jaicks. Resources: Zeliha Haytoğlu, Ayşegül Yolga Tahiroğlu. Software: Yusuf Döğüş. Supervision: Çağlar Charles Daniel Jaicks. Validation: Gonca Gül Çelik, Yusuf Döğüş. Visualization: Gonca Gül Çelik. Writing—original draft: Perihan Çam Ray, Çağlar Charles Daniel Jaicks. Writing—review & editing: Özlem Görüroğlu Öztürk, Çağlar Charles Daniel Jaicks.
Funding Statement
This study was supported by Çukurova University Scientific Research Projects Coordination Unit under project number TTU-2022-14121.
Acknowledgments
None