The WHO estimates that 1 in every 160 children in the world have ASD (2), whereas, according to CDC, about 1 in every 36 children who are 8 years old in the United States were diagnosed with ASD in 2020, a prevalence that was as 1 in 150 in 2000 (3). These increases are attributed to advancements in diagnostic criteria, increased awareness, and enhanced screening/surveillance systems (4, 5). The 2013 shift from DSM-IV to DSM-5 revised diagnostic thresholds by grouping disorders like Asperger’s syndrome and PDD-NOS into ASD, potentially contributing to increased prevalence (6, 7). However, these changes also posed diagnostic difficulties because females and adults are often paucisymptomatic and are underdiagnosed (8, 9). Past criteria have proved to be wanting when identifying high-functioning persons and females, as they are likely to feign or exhibit symptoms dissimilar to those of male patients (10).

Scientific information has strengthened the evidence for the role of genetic factors in the development of ASD, with twin studies suggesting a heritability index of between 50%-80% (11). Studies have pointed out that ASD has connections with several gene mutations and copy number variations like the SHANK3, CHD8, NRXN1, and FMR1 (12, 13). These mutations impact synaptic transmission, neuronal growth, and structural plasticity, essential for cognitive and behavioral control. DNA methylation and histone modification facilitate gene-environment interactions (epigenetic mechanisms), mediating the impact of environmental factors on genetic predispositions during neural development, potentially contributing to ASD’s connectivity and learning deficits (14). Other research indicates that epigenetic mechanisms may play a role in the disruption of connectivity and learning deficits in children with ASD (15).

Environmental risks are also very influential in the development of this ASD. Previous prenatal factors, including greater maternal age, paternal age, Gestational Diabetes Mellitus (GDM), preterm birth, cesarean section, and low birth weight, have been categorized as prenatal and perinatal risk factors for ASD (16–18). Airborne pollutants such as nitrogen dioxide and particulate matter, metals including mercury and lead, pesticides, and endocrine-disrupting chemicals are other environmental factors that have been shown to adversely affect neurodevelopment and influence the likelihood of ASD development (19–21). One notable factor is Maternal Immune Activation (MIA), which occurs when the mother’s immune system is triggered by increasing factors such as infection, autoimmune diseases, or inflammation during pregnancy. MIA has been demonstrated in animal and human studies to disrupt fetal brain development (21, 22). Children with ASD may have low vitamin D, B6, B12, folates, zinc, magnesium, and omega-3 fatty acids, which are crucial for neurotransmitter synthesis and immune function (23).

Emerging trends indicate that Maternal Metabolic Outcomes (MMO) such as GDM, obesity, and preeclampsia augment the odds of ASD in offspring (24). These disruptions are believed to trigger inflammatory responses, oxidative stress, and hormonal imbalances, which can negatively impact fetal brain development (25). For example, maternal diabetes, especially during the third trimester, may affect brain plasticity and synaptogenesis due to hyperglycemia and insulin resistance (2).

There are striking differences in ASD diagnosis and further management of the disorder regarding the region and socioeconomic status (SES). ASD is often undiagnosed in low- and middle-income countries (LMICs) because the healthcare systems are scanty; there is a dearth of specialists to attend to these patients. Society has a nondescriptive attitude towards mental health disorders (26). For instance, research on sub-Saharan African and South-Asian countries makes lower prevalence rate estimates, which might have been influenced by under-diagnosis instead of under-incidence (27, 28). In many such regions, children can end up being diagnosed with intellectual disability or psychiatric disorders when, in truth, they have ASD because of inadequate culturally sensitive screening tools (29). Racial and ethnic minorities, in particular, are documented to receive a diagnosis and are provided early intervention services later than others in high-income countries (30, 31). These disparities are attributed to language solutions that act as barriers, disparate healthcare access, and institutional racism that hinder good diagnosing and early intervention.

New systems of picturing autism have proposed concepts such as the attribute LAZY that encompasses Late-diagnosed, Atypical, Zero-yield screen patients with ASD forms that are less likely to be diagnosed at an early age or are high-functioning or female (25). These may not present classical diagnostic signs during initial assessments performed during childhood, especially when the diagnostic instruments are not sensitive to gender or cultural differences. Reluctance to acknowledge “My child is Autistic,” especially in Arab countries, widens this gap. Recent studies indicate Qatar has a high ASD prevalence, with 2019 research estimating 1.4% among school-aged children (32) and another reporting 1.14% among children aged 6 to 11 years (33). Some communities attribute ASD behaviors to possession by evil spirits or witchcraft, delaying diagnosis and treatment.

M-CHAT and ADOS-2: Current large-scale screening methods used globally to identify the early signs of ASD include the Modified Checklist for Autism in Toddlers (M-CHAT) and the Autism Diagnostic Observation Schedule (ADOS-2) (34).

However, the source of ASD is still not well understood due to its multi-faceted nature and the significant impact of the disorder on children all over the world; therefore, there is a need to compile knowledge about different patterns of prevalence and factors related to the occurrence of ASD as well as interventions for the disorder. There is growing interest in the use of educational approaches in the management of ASD, especially those that are individualized. One such is Montessori training, which is peaceful, systematic, touch-tangible, and student-centered. A couple of former studies highlight that Montessori approaches, such as autonomous discovery, handling things, and group work for a group, are consistent with learning styles preferred by children with ASD, especially regarding communication and behavior improvements (38). However, this promising approach has been used in systematic reviews but not explored in detail.

Hence, this review will compile a comprehensive synthesis of the global prevalence rate of ASD and explore genetic, epigenetic, and environmental factors that underlie the condition and existing intervention approaches with a special emphasis on the Montessori approach. This synthesis is drawn to help clinicians, educators, and policymakers make sense of the current state of ASD research and to identify future directions of care and research.

This systematic review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) criteria to guide their conduct to be clear and precise. This review aimed to review literature documenting the prevalence of ASD worldwide, the associated genetic and environmental causes, and the effectiveness of the existing intervention methods with specific reference to Montessori education. No quantitative meta-synthesis was performed due to the imbalance and heterogeneity in some variables employed in the discerned studies that ranged in match utilities, interventions and outcomes measures, and region environment.

The study selection process adhered to the PRISMA framework, involving four stages: identification, screening, eligibility, and inclusion. A total of 1,740 records were identified through 5 significant databases. The following selection criteria were applied: 1) answered the study question, 2) written in English, were published in peer-reviewed journals, 3) used rigorous methodology, 4) research indicating people with ASD of different age groups and genders from different geographic and socio-economic backgrounds, 5) Previous research on screening approaches to MMO in children and adults categorized the assessments into behavioral and pharmacological treatment perquisites, and/or policy-based and educational solutions were included, 6) researches that compare checklists, including the Autism Diagnostic Observation Schedule (ADOS-2) and Modified Checklist for Autism in Toddlers (M-CHAT), 7) targeted population was clearly defined with sample size greater than 50 cases and controls to give enough power to the study, and 8) the review only examined the articles published from the year 2000 up to 2024 to enhance the chances of covering the eventuality of capturing the recent knowledge in autism. Five analytical and epidemiological study types were considered for inclusion: cohort, cross-sectional, case-control, RCTs, and systematic reviews. The review also focused on those works reporting attempts to apply new or unique methods or adapt them to the regional particulars for those with autism, e.g., Montessori education. After applying the selection criteria, 480 were found to be duplicates, and 1,260 records were screened by title and abstract. Of these, 980 were excluded for irrelevance, leaving 280 full-text articles for eligibility assessment. Following the application of PICOS criteria, 265 articles were excluded due to lack of primary data, non-peer-reviewed status, small sample sizes (<50), or methodological limitations. Ultimately, 15 studies were included for qualitative synthesis and risk of bias assessment. The PRISMA flow chart, presented in the Results section, visually summarizes this process. The title and abstract screening steps were performed independently and blinded by two investigators to minimize bias. Only the full-text articles that met the initial criteria were evaluated in the second step.

The excluded studies were: 1) non-peer-reviewed conferences, editorials, or opinion pieces and did not contain research data; 2) studies that had articles that were specific to comorbidity and did not research AUTISM or had it as one of the co-primary diagnoses were excluded, 3) Any study conducted before the year 2000 was excluded unless it based research on autism in any way, 4) Studies with a sample size (<50) participants were also not considered due to low statistical power, 5) The studies conducted in animals without any direct association with clinical diagnosis and management of ASD, and 6) studies that appeared in journals, books, or were otherwise published in a language other than English were excluded if no official translation was available.

A comprehensive and systematic literature search was conducted across five significant databases: PubMed, Scopus, Web of Science, Cochrane Library, and Google Scholar. The search spanned publications from January 2000 to March 30, 2024. Boolean operators (AND, OR) and Medical Subject Headings (MeSH) were used to construct the search queries. Key terms included: “Autism Spectrum Disorder,” “ASD prevalence,” “autism diagnosis,” “genetic factors in ASD,” “environmental risk factors for autism,” “early intervention for ASD,” “Montessori education and autism,” and “behavioral therapy for autism.” The search strategy was refined repeatedly to maximize sensitivity and specificity. Grey literature and preprints were not included due to the scope of the review. The study was not preregistered on PROSPERO or any other protocol registry, which is acknowledged as a methodological limitation.

The primary research question guiding this systematic review was: “What are the global prevalence rates of ASD, what genetic and environmental factors contribute to its variability, and what intervention strategies—particularly Montessori education—are reported as effective in existing literature?” This question was formulated using the PICOS framework in Table 1 below to ensure a structured evidence retrieval and appraisal approach.

Data collection was done using data extraction forms to make the process more accurate and uniform. These consisted of the study title, authors, year of publication, and the source of the journal. Further information included the type of studies, the sample size, participants’ age and/or sex, prevalence rates by geographical location, and diagnostic methods used (DSM-IV, DSM-5, or ICD-10). Other aspects used systematically included information about the genetic/hereditary and environmental factors, the details of the intervention procedures, and the success rate of the various treatment options, including educational impact. Guidelines and policy proposals for managing ASD were retrieved when formulating healthcare frameworks were provided. Data extraction was completed by two authors individually, and disagreements were adjudicated with consensus through a review of primary sources.

The key findings of this systematic review are: 1—the rates of ASD in samples; 2—the identification of genetic and environmental risk factors for ASD; 3—the behavioral, pharmacological, and educational approaches adopted in the study, including those associated with the Montessori system. The review sought to situate these outcomes in various cultural and socio-economic environments.

The methodological quality of the included studies was assessed according to the Newcastle-Ottawa Scale (NOS) for observational studies and the Cochrane Risk of Bias for the RCTs. The NOS evaluates the studies according to selection bias, proper comparability, and outcome assessment. In contrast, the Cochrane tool assesses factors including random sequence generation, allocation concealment, blinding participants and personnel, the issue of incomplete outcome data, and reporting bias. The Newcastle-Ottawa Scale and the Cochrane risk bias assessment identified high-quality studies when assigned 7 points and above. Moderate-risk studies were considered cautiously, while this review did not include high-risk studies.

At the study level, potential sources of bias included selection, performance, attrition, reporting, and other types of bias. Bias was further checked through recruitment procedures and sample selection, and performance bias was checked by blinding the participants and implementing the intervention. Bias on detection was assessed through the relevance and the applicability of the outcome assessment methods. Attrition bias was investigated based on the completion rate of the studies while reporting bias was assessed based on the selective reporting of outcomes. Meta-analysis was also done with the exclusion of either moderate or high-risk publications to see the difference in the results.

The literature used in this narrative synthesis consists of 15 studies providing information about the prevalence of ASD globally, while geography is not uniformly covered. Studies showed the prevalence rate of tying across all age groups; differences were observed when it came to diagnostic criteria, health care facilities, availability of screening, culture, etc.; there is sufficient data for North America, specifically the United States and Canada, the Precise prevalence percent ranges from 2.8% to 1.0%. The most recent statistics from the U.S. Centers for Disease Control and Prevention show that the incidence is about 2.8 percent or one in thirty-six children of 8 years old (1). The included studies used surveys based on DSM-IV or DSM-5 enhanced surveillance systems in the USA, leading to precise real-life prevalence estimates (2–6).

In Europe, the rate of ASD is 1 in 150 children in a study conducted in the United Kingdom and the use of standard diagnostic tools, including the DSM-IV and ICD-10 (7). However, there are not enough studies in European literature published in recent years concerning intervention and education, including Montessori

In Asia, studies were mainly done in China and India. Another publication from China indicates underdiagnosis due to the lack of early detection programs and little professional education, with prevalence rates as low as 1:250 in rural areas (8). In other regions, such as India, there are no available data because people with special needs are fewer, services offered are scarce, and there is a stigma in the society. However, there is evidence that it was close to 1% in the urban areas (9).

In Africa, especially in sub-Saharan regions, studies reporting the incidence of ASD are limited. A scoping review in sub-Saharan Africa reported prevalence rates as low as 1 in 500 or unknown due to limited surveillance (41). Due to cultural beliefs that consider autism as an effect of endowment by evil spirits, spirits, or witchcraft, children with ASD suffer delayed diagnosis and treatment (11).

ASD prevalence studies in Latin America are limited, and most of the research is not published in indexed journals. In Brazil and Argentina, data shows increased awareness, and increasing diagnosis rates are estimated to be between 1 and 150 people. However, as far as I am aware of a level 1 literature review (12), there are not many well-controlled epidemiological surveys or investigations in this area temporarily.

In the Middle East, there have been increasing trends in the consistency of the prevalence ratio estimates. For instance, recent prevalence studies in Qatar showed that the prevalence rates range from 1.14% to 1.4% among school-going children due to enhanced efforts in early diagnosis and treatment in the region (13, 14). Still, knowledge of ASD is somehow restricted, and specific education and therapy for this disorder are also scarce in many Gulf and Arab countries.

Of the 15 investigated studies, 10 were case-control, 3 cross-sectional studies, a single prospective cohort study, and one RCT. The sample numbers were between 80 and 2256 participants. These features were genetic factors such as SHANK3, CHD8, and NRXN1 mutations, epigenetic factors such as DNA methylation, and environmental factors like prenatal toxins, maternal infections, and parental advanced age.

For instance, some papers looked into air pollution during pregnancy, maternal fever or influenza, and perinatal nutrition, such as folic acid use, and their relationship with the risk of ASD. For example, Schmidt et al. (2012) reported a reduced risk of ASD when the mother had taken high levels of folic acids early in the pregnancy (4), while Zerbo et al. (2013) established that fever experienced by the mother during the second trimester of pregnancy was associated with a higher risk of ASD (5).

Regarding the interventions, most of the studies compared the effectiveness of behavioral therapy, such as Applied Behavior Analysis, Speech and Language therapy, pharmacological treatment like Risperidone and aripiprazole, and newer approaches, which include artificial intelligence and virtual reality. Most of these investigations revealed improved results regarding communication, behavior, impulse control, and cognitive skills, where child-based interventions were used and applied early (2, 6, 9).

while the review focuses on the results of studies highlighting Montessori education, only one of the cited articles alluded to using Montessori-based treatments. This research (Lillard, 2012) compared students’ achievement results in classic Montessori, supplemented Montessori, and control groups. It stated that children in a Montessori environment showed better prosocial and regulatory skills than children in conventional learning settings (39). Still, this study was not ASD-specific, and children with autism were included in the analysis as a group, which makes the generalization of the results more challenging. There is a dearth of literature concerning Montessori education among students with ASD, and therefore, future research should specifically address it. There is little research-based information, however, about Montessori education and specifically on children with ASD – even though some tenets in the Montessori approach seem pretty applicable to children with ASD: emphasis on sensory learning, structure, and child-controlled pace of learning. However, it can be concluded that Montessori-based education is promising for students with autism, but more research is needed for a definitive understanding.

There was variability in the methodological quality and risk of bias in the included studies. Applying the Newcastle-Ottawa Scale (NOS) and Cochrane Risk of Bias Tool, nine were considered low risk, five had moderate risk, and one was at high risk. Thus, Samples in such studies selected arbitrarily failed to incorporate a population-based sampling framework. For example, several cross-sectional studies included participants from hospitals or clinics and are not necessarily generalizable to other ASD samples.

Performance bias was considerably low, especially in randomized and blinded trials. In contrast, detection bias depended on the systematic use of standardized assessment tools and instruments such as the DSM-IV, DSM-5, or ICD-10 diagnostic systems. Dropout rates were either low or adequately addressed in most studies; thus, attrition bias could be controlled accordingly. Bias in reporting was noticeable in cross-sectional studies that did not report ns-findings or did not report on the CIs and ESs.

Overall, methods categorized as moderate risk included low sample size, regional restriction, such as using data collected within a single city, and failure to control variables such as socio-economic status. These methodological variations are among the key reasons a quantitative meta-analysis could not be done since combining such data would dilute the analysis.

The results of this review agree with this index, identifying the differences in the prevalence of ASD worldwide and demonstrating how genetic, epigenomic, and environmental influences interact. Interventions are often of a different type and availability and are most popular in behavioral therapies available worldwide. Thus, even pharmacological and technological advances are more supportive, secondary to early behavioral and education interventions.

However, the review highlights an area in ASD education that lacks much research, namely Montessori-based learning. While presented in the included literature as limited, Montessori educational principles of structure, independent learning, and use of multiple senses reflect the requirements of children with ASD (39). This aligns well with the need to conduct more research to establish its usefulness in children with ASD.