A risk prediction model for autism spectrum disorder integrating biopsychosocial factors: a systematic review and meta-analysis with multicenter validation

Introduction

Autism spectrum disorder (ASD) is a complex neurodevelopmental condition characterized by persistent deficits in social communication and interaction, along with restricted, repetitive patterns of behavior. Global prevalence of ASD has risen markedly in recent decades, with current World Health Organization estimates indicating that it affects approximately 1% of the population worldwide, posing substantial public health challenges and highlighting the urgent need for effective early detection strategies (1). The clinical presentation of ASD is highly heterogeneous, a feature formally recognized in diagnostic systems such as the International Classification of Diseases (ICD)-11, which classifies the disorder as a continuum of manifestations ranging from subtle social-communicative impairments to marked functional limitations accompanied by prominent repetitive behaviors (2-4). This pronounced heterogeneity directly translates into substantial variability in the predictive efficacy of existing screening tools, particularly when applied to high-risk populations.

Advances in neuroscience have reshaped our understanding of ASD etiology, revealing it as a condition arising from the dynamic interplay of genetic, environmental, and neurobiological factors (5). Neuroimaging studies have identified atypical brain organization—including cortical thickening, altered functional connectivity, and changes in neural circuit dynamics—particularly in regions associated with social and sensory processing (6-8). Genetic research has further uncovered risk genes influencing synaptic function and neural plasticity (9), while cognitive studies highlight differences in social, executive, and sensory domains (10-12). Despite these insights, research has often progressed within isolated domains, with limited integration across biological, psychological, and social dimensions.

This etiological and phenotypic heterogeneity complicates early detection and contributes to substantial variability in the predictive efficacy of existing screening instruments. Widely used tools such as the Modified Checklist for Autism in Toddlers (M-CHAT), though valuable in certain settings, demonstrate inconsistent sensitivity—ranging between 75% and 95% across pediatric populations—and fail to systematically incorporate emerging evidence on environmental, epigenetic, and psychosocial risk factors (13,14). Current screening approaches predominantly focus on genetic or perinatal factors in isolation, neglecting the assessment of interactions between childhood psychosocial stressors—such as adverse childhood experiences (ACEs)—and biomarkers. Moreover, they do not adequately quantify cumulative environmental exposures or incorporate emerging risk factors such as perinatal antibiotic use and prenatal antidepressant exposure.

While several studies have developed genetic risk models or single-biomarker predictors, these approaches exhibit three critical limitations: (I) a unidimensional focus that neglects multifactorial interactions; (II) limited capacity to interpret biopsychosocial mechanisms; and (III) insufficient clinical applicability for personalized risk stratification or longitudinal monitoring (15-18). Recent efforts to identify biological subtypes of ASD, along with growing interest in the gut-brain axis (19), immune dysregulation (20), and oxidative stress (21), underscore the need for integrative frameworks that bridge neurobiological insight with clinically actionable prediction.

To address these gaps, this study aims to develop and validate a clinically applicable, multidimensional risk prediction model for ASD that systematically integrates biopsychosocial factors. By synthesizing evidence from systematic reviews and meta-analyses and conducting multicenter validation, we seek to provide a more robust tool for early ASD detection, ultimately facilitating timely and targeted interventions. We present this article in accordance with the PRISMA (22) and TRIPOD reporting checklists (available at https://tp.amegroups.com/article/view/10.21037/tp-2025-620/rc).

Methods

All stages of study identification, selection, quality assessment, and data extraction were carried out independently by two reviewers. Any discrepancies were resolved through consultation with a third reviewer.

Model development

We developed an ASD risk prediction model synthesizing evidence from systematic reviews/meta-analyses. Multivariate logistic regression with subgroup/sensitivity analyses quantified risk factor contributions, weighted by regression coefficients {β = ln [odds ratio (OR)]} (23). Risk characteristics of patients from multiple centers affiliated to West China Medical University were analyzed. Continuous variables with normal distribution were expressed as mean ± standard deviation (SD), skewness variables were described as median [interquartile range (IQR)], and categorical variables were presented as frequency (percentage).

Derived cohort

This predictive model was developed through systematic analysis of 37 meta-analyses and systematic reviews, incorporating evidence from 14 prospective cohort studies (24-37), 12 retrospective studies (38-49), 9 case-control investigations (50-58), and 2 cross-sectional surveys (59,60). Comprehensive literature searches were conducted in PubMed [1966–2023], Embase [1974–2023], and the Cochrane Library [1993–2023] through December 2023. Our search strategy combined controlled vocabulary (e.g., MeSH terms: “autism spectrum disorder”, “risk factor”) with free-text keywords (e.g., “cohort study”) using appropriate Boolean operators. The final pooled cohort comprised 158,217 children and adolescents (aged 2–18 years) from diverse populations across Asia (China, Japan), Europe (United Kingdom, Germany), and North America (USA), with male predominance (76.3%) and representation from both high-income (78.4%) and middle-income (21.6%) settings.

Literature retrieval strategy

The database-specific search syntax included:

• For PubMed: (“autism spectrum disorder” [Title/Abstract] OR ASD [Title/Abstract]) AND (“pretermbirth” [Title/Abstract] OR “antibiotic*” [Title/Abstract]) AND (2010:2024 [pdat]) AND (meta-analysis OR systemic review);
• For Cochrane Library: (“autism spectrum disorder”:ti,ab) AND (“Systematic Review”:pt);
• For Embase: (‘autism spectrum disorder’/exp OR ‘asperger syndrome’/exp).

Inclusion criteria: (I) systematic reviews/meta-analyses; (II) examination of biopsychosocial ASD risk factors; (III) reported risk ratios [OR/relative risk (RR)] with 95% confidence interval (CI); and (IV) English publications (after 2019).

Exclusion criteria: (I) exposure factors with inadequate evidence quality; (II) absence of quantitative risk estimates; and (III) biomarkers requiring excessively complex detection technologies.

From an initial identification of 327 articles through keyword searches, 92 duplicates were removed, and 198 records were excluded during title/abstract screening. Following full-text evaluation of 37 potentially eligible articles, 4 studies met all criteria for quantitative synthesis (Figure 1). Two independent researchers performed both initial screening and full-text assessment, with disagreements resolved through consultation with a third senior investigator.

Figure 1 Literature selection flow chart.

Data extraction and quality assessment

Data extraction encompassed four domains: (I) study characteristics (author, publication year, country); (II) study design (sample size, follow-up duration with units and calculation methods); (III) effect measures RR with 95% CI for cohort studies, OR with 95% CI for case-control studies); and (IV) predictive performance metrics [area under the curve (AUC) values specifying validation approach, sensitivity/specificity reported with corresponding thresholds].

Validation cohort

Inclusion criteria: children evaluated at the Child Health Clinics of West China Second Hospital, Sichuan University (including affiliated clinical centers) and Luzhou Maternal and Child Health Hospital (Luzhou Second People’s Hospital) between January 2016 and March 2023.

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Luzhou Maternal and Child Health Hospital (Luzhou Second People’s Hospital) (No. 2024030) and informed consent was obtained from all patients’ legal guardians. West China Second Hospital, Sichuan University was also informed and agreed to the study.

Exclusion criteria: children with congenital heart disease, major congenital malformations, or severe nutritional disorders.

A total of 3,150 children were included, among whom 563 were under 3 years old or over 14 years old, 715 had follow-up periods less than 24 months, and 697 had incomplete baseline data. The final validation cohort comprised 1,175 children, of whom 175 met Diagnostic and Statistical Manual of Mental Disorders (DSM)-5/ICD-10 diagnostic criteria for ASD (Figure 2).

Figure 2 Patient selection process in validation cohort.

Model performance evaluation

Predictive accuracy was assessed via receiver operating characteristic (ROC) analysis (sensitivity, specificity, optimal cutoff, AUC). AUC values [range: 0.5 (no discrimination) to 1.0 (perfect discrimination)] reflect discrimination capacity. Calibration curves evaluated model fit, while decision curve analysis (DCA) quantified clinical utility. Risk scores were derived by scaling normalized β coefficients (risk score = Σβ × 10) (61). Using Youden index maximization, patients were stratified into four risk groups based on ASD prevalence: low risk: score <15.0; moderate risk: 15.0≤ score <18.5; high risk: 18.5≤ score <22.0; very high risk: ≥22.0.

Model validation

External cohort data validated the prediction model. Discrimination was assessed via ROC analysis [AUC (95% CI), sensitivity, specificity, optimal cutoff]. Patients were stratified into four risk tiers: low, moderate, high, and very high using the optimal cutoff. Cumulative morbidity was analyzed with Kaplan-Meier curves. Calibration used Hosmer-Lemeshow tests (P>0.05 indicating good fit), and between-group differences were assessed by the log-rank test. Statistical analyses utilized SPSS 23.0 for multivariate regression modelling and R 4.2.1 for survival analysis and visualization.

Statistical analysis

Meta-analysis method studies the OR with 95% CI of the included articles were quantitatively pooled by RevMan 5.4 software. Inter-study heterogeneity was double assessed by Cochran’s Q test (significance threshold P<0.10) and I² statistic (I²>50% judged significant heterogeneity). DerSimonian-Laird was applied when significant heterogeneity existed; otherwise, fixed-effects models (Mantel-Haenszel) were used. Sensitivity analyses employed sequential exclusion of included studies. Publication bias was evaluated through funnel plot visualization and Egger’s regression (α=0.05) with P>0.05 were interpreted as lack of substantial bias.

Results

Cohort characteristics

We analyzed baseline profiles of participants in the cohort studies. The derived cohort comprised approximately 158,217 participants aged 2 to 18 years (76.3% male, 23.7% female). Eighteen risk factors were examined, categorized as follows: maternal/pregnancy factors (rheumatoid arthritis, multivitamin supplement use during pregnancy, thyroid dysfunction, preterm birth, low birth weight, abnormal gestational weight gain, and antidepressant use); childbirth factors (oxytocin administration and perinatal antibiotic exposure); childhood factors (asthma, early antibiotic exposure, excessive screen time exposure, and ACEs, the latter being significantly associated with environmental and psychosocial factors); biological markers(abnormal peripheral blood homocysteine (Hcy) levels and impaired facial recognition); comorbidities/related diseases [attention-deficit/hyperactivity disorder (ADHD) and congenital heart disease prevalence warranted particular focus]. The influence of other factors, such as breastfeeding, remained unclear. These factors may increase the risk of ASD through potential interactions involving genetic, immune, metabolic, and environmental pathways.

Results meta-analysis

Model derivation

Through systematic literature review, 18 potential autism risk factors were initially included in this study. After meta-regression analysis, 14 factors were excluded due to low research quality or data deficiencies (Table 1). Four bio-psychosocial core factors were retained. ACEs preterm birth (gestational age <37 weeks); antidepressant exposure during pregnancy [selective serotonin reuptake inhibitors (SSRIs)]; perinatal antibiotic use (within 7 days of birth). Risk prediction model for autism based on multivariate logistic regression analysis: logit(P) = 0.82 × ACEs + 1.19 × premature birth + 0.42 × antidepressants during pregnancy + 0.21 × perinatal antibiotics, internally (Table 2).

Validation cohort

The validation cohort for this study was derived from two tertiary hospitals: West China Second Hospital, Sichuan University (including affiliated clinical centers) and Luzhou Maternal and Child Health Hospital (Luzhou Second People’s Hospital). Between January 2018 and December 2023, a total of 1,175 children aged 3–14 years were diagnosed with ASD. This group comprised 751 males (64.0%) and 424 females (36.0%), with a mean ± SD age of 6.1±2.6 years and a median age of 6 years (IQR, 4–8 years). Within this cohort, 175 children received their ASD diagnosis within 2 years of presentation/assessment (according to the DSM-5 criteria). This subgroup consisted of 108 males (61.7%) and 67 females (38.3%), with a mean ± SD age of 3.5±1.8 years and a median age of 3.2 years (IQR, 2.5–4.5 years).

Model validation

The validation cohort comprised 1,175 children aged 3–14 years from West China Second Hospital, Sichuan University (including affiliated clinical centers) and Luzhou Maternal and Child Health Hospital (Luzhou Second People’s Hospital), including 175 diagnosed with ASD. The model’s discrimination, calibration, and clinical utility were evaluated using an independent cohort via ROC curve analysis. The model achieved an AUC of 0.78 (95% CI: 0.75–0.81), significantly outperforming random prediction. The optimal cutoff, determined by the maximum Youden index, corresponded to a score of 17 points, yielding a sensitivity of 81% and specificity of 80%.

Based on ASD risk scores, the 1,175 validation cohort participants were stratified into four risk groups: low-risk (score <15.0), medium-risk (15.0≤ score <18.5), high-risk (18.5≤ score <22.0), and extreme-risk (≥22.0). Kaplan-Meier curves were used to calculate cumulative risk for each group. Compared to the low-risk reference group, the high-risk and extreme-risk groups showed significantly elevated risk ratios: high-risk: RR =5.99 (95% CI: 2.17–16.6; P<0.01); extreme-risk: RR =20.89 (95% CI: 7.91–55.18; P<0.01). Within the validation cohort, the ROC curve AUC was 0.78 (95% CI: 0.75–0.81), with consistent sensitivity (81%) and specificity (80%) (Figures 3-5). Furthermore, DCA confirmed substantial clinical utility, with the model providing a net benefit over default strategies at threshold probabilities exceeding 15%.

Figure 3 ROC curve was used to evaluate the model for predicting ASDs. The results showed that the AUC was 0.78 (95% CI: 0.75–0.81), indicating high discrimination of the model. ASD, autism spectrum disorder; AUC, area under the curve; CI, confidence interval; ROC, receiver operating characteristic.

Figure 4 Cumulative incidence of ASD endpoints by risk group analyzed by Kaplan-Meier curves: high-risk group vs. low-risk group: HR of 5.99 (95% CI: 2.17–16.6), P<0.01; very high-risk group vs. low-risk group: HR 20.89 (95% CI: 7.91–55.18), P<0.01. ASD, autism spectrum disorder; CI, confidence interval; HR, hazard ratio.

Figure 5 Four risk groups stratified by risk score in validation cohort prevalence of ASD: low-risk group (score <15.0), medium-risk group (15.0≤ score <18.5), high-risk group (18.5≤ score <22.0), and extreme-risk group (score ≥22.0). ASD, autism spectrum disorder.

Discussion

In recent years, the global prevalence of ASD has risen to 1.6% (WHO, 2023), yet early diagnosis rates remain below 20%. Substantial clinical heterogeneity contributes to the suboptimal performance of conventional screening tools; for example, the M-CHAT (62), demonstrates limited sensitivity (65–78%) in population studies (2). This evidence highlights an urgent need for multidimensional prediction models integrating biological, psychological, and social determinants.

To address the limitations of existing tools in terms of interpretability and clinical utility, we developed a novel ASD prediction model grounded in a systematic synthesis of recent evidence [2019–2024]. The model integrates four key biopsychosocial factors—ACEs, preterm birth, and prenatal antidepressant and perinatal antibiotic exposures—into a single formula: logit(P) = 0.82 × ACEs + 1.19 × preterm birth + 0.42 × prenatal antidepressants + 0.21 × perinatal antibiotics. The final model demonstrated strong predictive performance upon external validation, achieving an AUC of 0.78, which supports its potential as a practical tool for improving early ASD risk stratification.

The four independent risk factors identified in this study are hypothesized to contribute to ASD risk through multidimensional and potentially interrelated biological pathways. ACEs may influence neurodevelopment through persistent activation of the hypothalamic-pituitary-adrenal axis, which has been associated with structural alterations in brain regions critical for social communication, such as reduced hippocampal volume and atypical myelination in the prefrontal cortex (62); preterm birth, often accompanied by hypoxic conditions, may promote microglial activation and oxidative stress, which could disrupt the developmental timing and specificity of synaptic pruning (63); exposure to antidepressants during pregnancy—particularly SSRIs—is theorized to interfere with fetal serotonin signaling, a system integral to neural crest migration and cortical organization (60); similarly, perinatal antibiotic exposure is thought to potentially alter gut microbiome composition, which may in turn influence neurodevelopment via the gut-brain axis through changes in immune function and microbial metabolite production, such as short-chain fatty acids (33). However, it is important to note that these mechanisms remain areas of active investigation. The observed associations may also reflect shared underlying vulnerabilities—such as genetic predispositions or maternal inflammatory states—that contribute both to these perinatal risk factors and to ASD etiology. Further longitudinal and mechanistic studies are needed to clarify whether these factors play causal roles or represent early markers of shared etiological pathways.

This study advances beyond traditional screening tools and single-dimensional models through three key innovations:

• Multidimensional integration: we are the first to unify psychosocial stress (ACEs), perinatal medical exposures (antibiotics/antidepressants), and classical biological markers (preterm birth) within a single framework, overcoming the oversimplified genetic-perinatal dichotomy prevalent in existing models.
• Enhanced interpretability: standardized β coefficients quantify factor contributions within the model. ACEs demonstrated the strongest association (β=0.82, accounting for 37.3% of the predictive weight), providing a mechanistic basis for targeted interventions.
• Improved generalizability: external validation across diverse populations (78.4% high-income; 21.6% upper-middle-income countries, World Bank classifications) demonstrated robust performance. The calibration slope of 0.92 (95% CI: 0.87–0.97) indicates consistent cross-cultural applicability.

However, some limitations of this systematic review should be considered. The predominantly Asian derivation cohort (78.4%) may reduce applicability to populations with differing perinatal practices, despite Western participant inclusion. Model validation relied solely on cross-sectional data, precluding assessment of risk factor dynamics, which limited the model’s sensitivity to early ASD detection. Future work will include validation studies in Western countries to assess the model’s performance and robustness across different populations.

Conclusions

In summary, this study sought to advance ASD risk prediction by developing and validating an integrated bio-psycho-social model. Building upon existing research, this model synthesizes factors such as ACEs, preterm birth, and perinatal medication exposures to provide a preliminary framework for future refinement and clinical assessment.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the PRISMA and TRIPOD reporting checklists. Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-620/rc

Data Sharing Statement: Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-620/dss

Peer Review File: Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-620/prf

Funding: The project was funded by the Medicine &Engineering & Informatics Fusion and Transformation Key Laboratory of Luzhou City (No. XGY202411), the Southwest Medical University Natural Science Foundation (No. 2023QN018), and the Sichuan Maternal and Child Medical Science and Technology Innovation Project (No. FXYB02).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tp.amegroups.com/article/view/10.21037/tp-2025-620/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Luzhou Maternal and Child Health Hospital (Luzhou Second People’s Hospital) (No. 2024030) and informed consent was obtained from all patients’ legal guardians. West China Second Hospital, Sichuan University was also informed and agreed to the study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Salari N, Rasoulpoor S, Rasoulpoor S, et al. The global prevalence of autism spectrum disorder: a comprehensive systematic review and meta-analysis. Ital J Pediatr 2022;48:112. [Crossref] [PubMed]
2. Sanchez-Jimeno C, Blanco-Kelly F, López-Grondona F, et al. Attention Deficit Hyperactivity and Autism Spectrum Disorders as the Core Symptoms of AUTS2 Syndrome: Description of Five New Patients and Update of the Frequency of Manifestations and Genotype-Phenotype Correlation. Genes (Basel) 2021;12:1360. [Crossref] [PubMed]
3. Herrera ML, Paraíso-Luna J, Bustos-Martínez I, et al. Targeting epigenetic dysregulation in autism spectrum disorders. Trends Mol Med 2024;30:1028-46. [Crossref] [PubMed]
4. Gillett G, Leeves L, Patel A, et al. The prevalence of autism spectrum disorder traits and diagnosis in adults and young people with personality disorders: A systematic review. Aust N Z J Psychiatry 2023;57:181-96. [Crossref] [PubMed]
5. Willsey HR, Willsey AJ, Wang B, et al. Genomics, convergent neuroscience and progress in understanding autism spectrum disorder. Nat Rev Neurosci 2022;23:323-41. [Crossref] [PubMed]
6. Buch AM, Vértes PE, Seidlitz J, et al. Molecular and network-level mechanisms explaining individual differences in autism spectrum disorder. Nat Neurosci 2023;26:650-63. [Crossref] [PubMed]
7. Rasero J, Jimenez-Marin A, Diez I, et al. The Neurogenetics of Functional Connectivity Alterations in Autism: Insights From Subtyping in 657 Individuals. Biol Psychiatry 2023;94:804-13. [Crossref] [PubMed]
8. Shen L, Zhang J, Fan S, et al. Cortical thickness abnormalities in autism spectrum disorder. Eur Child Adolesc Psychiatry 2024;33:65-77. [Crossref] [PubMed]
9. Nisar S, Bhat AA, Masoodi T, et al. Genetics of glutamate and its receptors in autism spectrum disorder. Mol Psychiatry 2022;27:2380-92. [Crossref] [PubMed]
10. Lage C, Smith ES, Lawson RP. A meta-analysis of cognitive flexibility in autism spectrum disorder. Neurosci Biobehav Rev 2024;157:105511. [Crossref] [PubMed]
11. Baumeister S, Moessnang C, Bast N, et al. Processing of social and monetary rewards in autism spectrum disorders. Br J Psychiatry 2023;222:100-11. [Crossref] [PubMed]
12. Gonçalves AM, Monteiro P. Autism Spectrum Disorder and auditory sensory alterations: a systematic review on the integrity of cognitive and neuronal functions related to auditory processing. J Neural Transm (Vienna) 2023;130:325-408. [Crossref] [PubMed]
13. Sturner R, Howard B, Bergmann P, et al. Autism screening at 18 months of age: a comparison of the Q-CHAT-10 and M-CHAT screeners. Mol Autism 2022;13:2. [Crossref] [PubMed]
14. Bandara D, Riccardi K. Graph Node Classification to Predict Autism Risk in Genes. Genes (Basel) 2024;15:447. [Crossref] [PubMed]
15. Ben-Sasson A, Guedalia J, Nativ L, et al. A Prediction Model of Autism Spectrum Diagnosis from Well-Baby Electronic Data Using Machine Learning. Children (Basel) 2024;11:429. [Crossref] [PubMed]
16. Dawson G, Rieder AD, Johnson MH. Prediction of autism in infants: progress and challenges. Lancet Neurol 2023;22:244-54. [Crossref] [PubMed]
17. Ma Z, Xu L, Li Q, et al. Prediction Model for Sensory Perception Abnormality in Autism Spectrum Disorder. Int J Mol Sci 2023 25;24:2367.
18. Keifer CM, Mikami AY, Morris JP, et al. Prediction of social behavior in autism spectrum disorders: Explicit versus implicit social cognition. Autism 2020;24:1758-72. [Crossref] [PubMed]
19. Petropoulos A, Stavropoulou E, Tsigalou C, et al. Microbiota Gut-Brain Axis and Autism Spectrum Disorder: Mechanisms and Therapeutic Perspectives. Nutrients 2025;17:2984. [Crossref] [PubMed]
20. Pangrazzi L, Balasco L, Bozzi Y. Oxidative Stress and Immune System Dysfunction in Autism Spectrum Disorders. Int J Mol Sci 2020;21:3293. [Crossref] [PubMed]
21. Liu M, Chen Y, Sun M, et al. Auts2 regulated autism-like behavior, glucose metabolism and oxidative stress in mice. Exp Neurol 2023;361:114298. [Crossref] [PubMed]
22. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Syst Rev 2021;10:89. [Crossref] [PubMed]
23. Horner D, Jepsen JRM, Chawes B, et al. A western dietary pattern during pregnancy is associated with neurodevelopmental disorders in childhood and adolescence. Nat Metab 2025;7:586-601. [Crossref] [PubMed]
24. Kaas TH, Vinding RK, Stokholm J, et al. Association between childhood asthma and attention deficit hyperactivity or autism spectrum disorders: A systematic review with meta-analysis. Clin Exp Allergy 2021;51:228-52. [Crossref] [PubMed]
25. Jenkinson R, Milne E, Thompson A. The relationship between intolerance of uncertainty and anxiety in autism: A systematic literature review and meta-analysis. Autism 2020;24:1933-44. [Crossref] [PubMed]
26. Alemany S, Avella-García C, Liew Z, et al. Prenatal and postnatal exposure to acetaminophen in relation to autism spectrum and attention-deficit and hyperactivity symptoms in childhood: Meta-analysis in six European population-based cohorts. Eur J Epidemiol 2021;36:993-1004. [Crossref] [PubMed]
27. da Motta TP, Owens J, Abreu LG, et al. Malocclusion characteristics amongst individuals with autism spectrum disorder: a systematic review and meta-analysis. BMC Oral Health 2022;22:341. [Crossref] [PubMed]
28. Amnuaylojaroen T, Parasin N, Saokaew S. Exploring the association between early-life air pollution exposure and autism spectrum disorders in children: A systematic review and meta-analysis. Reprod Toxicol 2024;125:108582. [Crossref] [PubMed]
29. Sun CK, Cheng YS, Chen IW, et al. Impact of parental rheumatoid arthritis on risk of autism spectrum disorders in offspring: A systematic review and meta-analysis. Front Med (Lausanne) 2022;9:1052806. [Crossref] [PubMed]
30. Yao H, Fu Y, Weng X, et al. The Association between Prenatal Per- and Polyfluoroalkyl Substances Exposure and Neurobehavioral Problems in Offspring: A Meta-Analysis. Int J Environ Res Public Health 2023;20:1668. [Crossref] [PubMed]
31. Yin F, Wang H, Liu Z, et al. Association between peripheral blood levels of C-reactive protein and Autism Spectrum Disorder in children: A systematic review and meta-analysis. Brain Behav Immun 2020;88:432-41. [Crossref] [PubMed]
32. Uddin MS, Azima A, Aziz MA, et al. CNTNAP2 gene polymorphisms in autism spectrum disorder and language impairment among Bangladeshi children: a case-control study combined with a meta-analysis. Hum Cell 2021;34:1410-23. [Crossref] [PubMed]
33. Lee E, Cho J, Kim KY. The Association between Autism Spectrum Disorder and Pre- and Postnatal Antibiotic Exposure in Childhood-A Systematic Review with Meta-Analysis. Int J Environ Res Public Health 2019;16:4042. [Crossref] [PubMed]
34. Laverty C, Surtees A, O'Sullivan R, et al. The prevalence and profile of autism in individuals born preterm: a systematic review and meta-analysis. J Neurodev Disord 2021;13:41. [Crossref] [PubMed]
35. Tioleco N, Silberman AE, Stratigos K, et al. Prenatal maternal infection and risk for autism in offspring: A meta-analysis. Autism Res 2021;14:1296-316. [Crossref] [PubMed]
36. Boshoff K, Gibbs D, Phillips RL, et al. A meta-synthesis of how parents of children with autism describe their experience of advocating for their children during the process of diagnosis. Health Soc Care Community 2019;27:e143-57. [Crossref] [PubMed]
37. Li H, Liu H, Chen X, et al. Association of food hypersensitivity in children with the risk of autism spectrum disorder: a meta-analysis. Eur J Pediatr 2021;180:999-1008. [Crossref] [PubMed]
38. Nayeri T, Sarvi S, Moosazadeh M, et al. Relationship between toxoplasmosis and autism: A systematic review and meta-analysis. Microb Pathog 2020;147:104434. [Crossref] [PubMed]
39. Zhu Z, Tang S, Deng X, et al. Maternal Systemic Lupus Erythematosus, Rheumatoid Arthritis, and Risk for Autism Spectrum Disorders in Offspring: A Meta-analysis. J Autism Dev Disord 2020;50:2852-9. [Crossref] [PubMed]
40. Guo BQ, Li HB, Ding SB. Blood homocysteine levels in children with autism spectrum disorder: An updated systematic review and meta-analysis. Psychiatry Res 2020;291:113283. [Crossref] [PubMed]
41. Chien MC, Huang CY, Wang JH, et al. Effects of vitamin D in pregnancy on maternal and offspring health-related outcomes: An umbrella review of systematic review and meta-analyses. Nutr Diabetes 2024;14:35. [Crossref] [PubMed]
42. Wu Y, Cao H, Baranova A, et al. Multi-trait analysis for genome-wide association study of five psychiatric disorders. Transl Psychiatry 2020;10:209. [Crossref] [PubMed]
43. Tsai TY, Chao YC, Hsieh CY, et al. Association Between Atopic Dermatitis and Autism Spectrum Disorder: A Systematic Review and Meta-analysis. Acta Derm Venereol 2020;100:adv00146. [Crossref] [PubMed]
44. Hartley G, Sirois F, Purrington J, et al. Adverse Childhood Experiences and Autism: A Meta-Analysis. Trauma Violence Abuse 2024;25:2297-315. [Crossref] [PubMed]
45. Andreadou MT, Katsaras GN, Talimtzi P, et al. Association of assisted reproductive technology with autism spectrum disorder in the offspring: an updated systematic review and meta-analysis. Eur J Pediatr 2021;180:2741-55. [Crossref] [PubMed]
46. Jenabi E, Seyedi M, Bashirian S, et al. Is there an association between labor induction and attention-deficit/hyperactivity disorder among children? Clin Exp Pediatr 2021;64:489-93. [Crossref] [PubMed]
47. Salgado-Cacho JM, Moreno-Jiménez MDP, de Diego-Otero Y. Detection of Early Warning Signs in Autism Spectrum Disorders: A Systematic Review. Children (Basel) 2021;8:164. [Crossref] [PubMed]
48. Aishworiya R, Ma VK, Stewart S, et al. Meta-analysis of the Modified Checklist for Autism in Toddlers, Revised/Follow-up for Screening. Pediatrics 2023;151:e2022059393. [Crossref] [PubMed]
49. Rech JP, Irwin JM, Rosen AB, et al. Comparison of Physical Activity Between Children With and Without Autism Spectrum Disorder: A Systematic Review and Meta-Analysis. Adapt Phys Activ Q 2022;39:456-81. [Crossref] [PubMed]
50. Shayestehfar M, Nakhostin-Ansari A, Memari A, et al. Risk of autism spectrum disorder in offspring with parental schizophrenia: a systematic review and meta-analysis. Nord J Psychiatry 2023;77:127-36. [Crossref] [PubMed]
51. Jenabi E, Bashirian S, Salehi AM, et al. Not breastfeeding and risk of autism spectrum disorders among children: a meta-analysis. Clin Exp Pediatr 2023;66:28-31. [Crossref] [PubMed]
52. Panagiotou G, Taylor PN, Rees DA, et al. Late offspring effects of antenatal thyroid screening. Br Med Bull 2022;143:16-29. [Crossref] [PubMed]
53. Ma D, Huang JL, Xiong T. Association between congenital heart disease and autism spectrum disorders: A protocol for a systematic review and meta-analysis. Medicine (Baltimore) 2023;102:e33247. [Crossref] [PubMed]
54. Mitchell RA, Barton SM, Harvey AS, et al. Factors associated with autism spectrum disorder in children with tuberous sclerosis complex: a systematic review and meta-analysis. Dev Med Child Neurol 2021;63:791-801. [Crossref] [PubMed]
55. Griffin JW, Bauer R, Scherf KS. A quantitative meta-analysis of face recognition deficits in autism: 40 years of research. Psychol Bull 2021;147:268-92. [Crossref] [PubMed]
56. Friel C, Leyland AH, Anderson JJ, et al. Prenatal Vitamins and the Risk of Offspring Autism Spectrum Disorder: Systematic Review and Meta-Analysis. Nutrients 2021;13:2558. [Crossref] [PubMed]
57. Ophir Y, Rosenberg H, Tikochinski R, et al. Screen Time and Autism Spectrum Disorder: A Systematic Review and Meta-Analysis. JAMA Netw Open 2023;6:e2346775. [Crossref] [PubMed]
58. Jenabi E, Bashirian S, Khazaei S. Association between neonatal jaundice and autism spectrum disorders among children: a meta-analysis. Clin Exp Pediatr 2020;63:8-13. [Crossref] [PubMed]
59. Hegvik TA, Klungsøyr K, Kuja-Halkola R, et al. Labor epidural analgesia and subsequent risk of offspring autism spectrum disorder and attention-deficit/hyperactivity disorder: a cross-national cohort study of 4.5 million individuals and their siblings. Am J Obstet Gynecol 2023;228:233.e1-233.e12. [Crossref] [PubMed]
60. Uguz F. Neonatal and Childhood Outcomes in Offspring of Pregnant Women Using Antidepressant Medications: A Critical Review of Current Meta-Analyses. J Clin Pharmacol 2021;61:146-58. [Crossref] [PubMed]
61. Ren Q, Chen D, Liu X, et al. Derivation and Validation of a Prediction Model of End-Stage Renal Disease in Patients With Type 2 Diabetes Based on a Systematic Review and Meta-analysis. Front Endocrinol (Lausanne) 2022;13:825950. [Crossref] [PubMed]
62. Braconnier ML, Siper PM. Neuropsychological Assessment in Autism Spectrum Disorder. Curr Psychiatry Rep 2021;23:63. [Crossref] [PubMed]
63. Crump C, Sundquist J, Sundquist K. Adverse Pregnancy Outcomes and Long-Term Mortality in Women. JAMA Intern Med 2024;184:631-40. [Crossref] [PubMed]