The association between post-term births and autism spectrum disorders: an updated systematic review and meta-analysis

Background This study aimed to conduct a meta-analysis to determine whether post-term birth has an increased risk of ASD. Materials and methods To retrieve eligible studies regarding the effect of post-term and ASD in children, major databases including PubMed, Scopus, and Web of Science were searched. A random effect model was used for meta-analysis. For assessing the quality of included studies, the GRADE checklist was used. Results In total, 18 records were included with 1,412,667 sample populations from 12 countries. The pooled estimates of RR and OR showed a significant association between post-term birth and ASD among children, respectively (RR = 1.34, 95% CI 1.10 to 1.58) and (OR = 1.47, 95% CI 1.03 to 1.91). There was no heterogeneity among the studies that reported the risk of ASD among children based on RR (I2 = 6.6%, P = 0.301). There was high heterogeneity in the studies reported risk of ASD based on OR (I2 = 94.1%, P = 0.000). Conclusion Post-term births still occur relatively frequently (up to 5–10%) even in developed countries. Our results showed that post-term birth is an increased risk of ASD, although high heterogeneity was found among the studies reported based on adjusted and crude forms, however, after subgroup analysis by gender, this heterogeneity disappeared among males.


Introduction
Autism spectrum disorder (ASD) is defined by severe social communication deficits and stereotyped, repetitive behaviors.Recognizing that the social problems that characterize ASD can occur in different forms, depending on language skills, general ability level, the severity of symptoms, context, and coexistence disorders, has led to a dramatic increase in the number of children with ASD [1].
The pathophysiology of ASD is not fully understood, ASD was previously thought to be mainly genetic, however, genetic factors alone have been found to account for 20-30% of ASD cases, whereas the remaining 70-80% are due to a complex interaction between environmental risk factors such as prenatal and postnatal environments and genetic predisposition [2].
A post-term birth extends beyond 42 weeks (294 days) from the first day of the last menstrual period, although the routine use of ultrasound to confirm the date of pregnancy can reduce the incidence of post-term birth, however, post-term birth still occur relatively frequently (up to 5-10%) even in developed countries [3] and associated with neonatal morbidity and mortality [4].
A prior systematic review has reported an overall prevalence of 0.6% of ASD in people with post-term birth [5].Some studies have reported post-term birth as a pregnancy risk factor for ASD [6,7], while other studies have not shown this relationship [8,9], also previously, metaanalysis found that post-term birth is not a risk factor for ASD [10].Therefore, this study aimed to conduct a meta-analysis to determine whether post-term birth has an increased risk of ASD.

Materials and methods
This meta-analysis was performed considering the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines [11].The number of retrieved sources, the number of excluded sources, and the reason for exclusion are given in the PRISMA diagram (see Fig. 1).

Search procedure
To retrieve eligible studies regarding the effect of postterm birth on ASD, major databases including PubMed, Scopus, and Web of Science were searched by the following keywords: post-term OR postterm OR postmature*) and (autism OR autistic OR autism spectrum disorders OR ASD).

Inclusion and exclusion criteria
Only original research articles were considered and other types of articles including letters to the editor, case reports, and other systematic reviews were excluded.However, the reference list of such papers was investigated carefully for finding missed sources during database searching.Studies focused on the causal relationship between ASD and post-term birth including cases was included.There was no restriction on the language or the publication date of retrieved sources.Furthermore, only peer-reviewed articles (in published or in-press status) were considered.

Study selection and data extraction
Two independent authors (A.S and E.J) performed the search procedures.Any conflict regarding the retrieved sources was resolved by discussion between authors.A PECOS (Participant, Exposure, Comparison, Outcome and Study design) model was used for finding eligible studies.The Participants were all newborns, the Exposure was post-term birth, the Comparison was term birth, the Outcome was ASD and the Study design was observational study.After the title, abstract and full-text screening and adding the retrieved sources found by manual searching or found by checking the reference list of eligible studies, an electronic form was used for extracting information embedded in each source.The information included the first author's name, publication date, the country that the study was performed, study design, sample size, the measure that was used for effect size estimation, the age range of participants, diagnostic criteria that were used for autism, and quality of each study.

Quality assessment
The evaluation of the studies that were included in the analysis was conducted using a tool called the Grades of Recommendation, Assessment, Development, and Evaluation (GRADE) [12].The assessment considered several important factors including limitations, consistency, directness, imprecision, reporting bias, strength, gradient, and confounding.

Sensitive analysis
In situations where there was high between-study heterogeneity, we employed a sequential algorithm [13] to determine the source of the heterogeneity.This involved removing one study from the calculations at a time and identifying which study had the greatest impact on decreasing I 2 .This process was repeated for a new set of n-1 studies, and continued until I 2 dropped below the intended threshold of 50%.If there were multiple studies that could potentially result in I 2 dropping below the threshold, we reported the minimum I 2 .

Statistical analysis
A random effect model was used for meta-analysis.The effect size was calculated using the odds ratio (OR) and relative risk (RR) with a 95% confidence interval (where the risk increases the occurrence of a desired outcome, the OR and RR will be greater than 1).All analyses were performed using STATA, version 13 (StataCorp, College Station, TX, USA).Furthermore, for all statistical analyses, the significance level was adjusted to 0.05.To investigate the source of heterogeneity, subgroup analyses for the type of study design, gender, and quality of studies were also performed.
The between-study heterogeneity was calculated based on I 2 statistic [14] and Cochrane Q-test.The value obtained for I 2 determines the heterogeneity between studies in a way that for I 2 > 50% a significant betweenstudy heterogeneity exists; otherwise, when I 2 merges to zero, between-study heterogeneity is non-significant.The Cochrane Q-test determines the statistical significance of the heterogeneity.For assessing publication bias, the funnel plot was used as a graphical tool.The symmetrical distribution of studies in a funnel plot implies the absence of publication bias.Furthermore, Egger's [15] and Begg's [16] tests were used as quantitative measures for assessing publication bias.

Main analysis
The association between post-term birth and ASD is presented in Fig. 2. The pooled estimates of RR and OR showed a significant association between post-term birth and ASD, respectively (RR = 1.34, 95% CI 1.10 to 1.58) and (OR = 1.47, 95% CI 1.03 to 1.91).There was no heterogeneity among the studies that reported the risk of ASD based on RR (I 2 = 6.6%,P = 0.301).There was high heterogeneity in the studies reported risk of ASD based on OR (I 2 = 94.1%,P = 0.000).
The funnel plot is presented in Fig. 3.There was no evidence of publication bias among studies.The P values were 0.910 and 0.341 based on Begg's and Eggerʼs regression, respectively.

Quality of the studies
The results of this evaluation can be found in Table 2, where it shows that the quality of 5 studies was considered moderate, while 13 studies were assessed as having low quality (Table 2).

Sensitive analysis
To ensure between-study homogeneity, we conducted a sensitivity analysis using the sequential algorithm.By omitting the two studies [20,23] from the meta-analysis examining the association, we were able to achieve the minimum desired I 2 threshold of 50% (OR = 1.39, 95% CI 0.76, 2.02, I 2 = 30.7%).

Subgroup analysis
The subgroup analysis was performed based on the design of the studies, the quality of the studies, the child's gender, publication year, continent, and autism diagnosis criteria.OR in case-control and cohort studies were reported 0.94 (0.88, 1.0) and 1.30 (1.23, 1.37), respectively.A significant association was found between postterm birth and ASD among cohort studies (Table 3).In addition, there was a significant association between post-term birth and ASD in the studies with high quality (OR = 1.63, 95% CI 1.11, 2.15).However, this association had high heterogeneity (I 2 = 97.5%).There was no significant association between post-term birth and ASD in the studies with low quality (OR = 0.94, 95% CI 0.88, 1.00).The subgroup analysis based on the child's gender showed that a significant association was found between post-term birth and ASD among male children (OR = 1.37, 95% CI 1.19, 1.56).This association was homogenous, while there was no significant association between post-term birth and ASD among female children (OR = 1.13, 95% CI 0. 84, 1.43).There was a significant association between post-term birth and ASD in the studies that publication year was after 2000 (OR = 1.61, 95% CI 1.13, 2.09).There was not a significant association between post-term birth and ASD in the studies based on continent and autism diagnosis criteria (Table 3).There was significant difference in ASD risk between study design in cohort and case-control studies, quality of the studies between low and moderate quality, child's gender between male and female and autism diagnoses criteria between ICD and DSM (P interaction = 0.001, 0.001, 0.024 and 0.026), but there was no significant difference in publication year between before and after 2000 and continent between Europe/USA and Asian (all P interac- tion = 0.383 and 0.778).

Discussion
The existence of cognitive and learning problems in a person with ASD, its negative impact on patient's social interactions with the world around them, and the increasing cost to the health system have caused researchers to pay much attention to this disorder and work on the factors affecting the disease [31].
There is evidence that preterm birth [32], small for gestational age [33], prenatal and antenatal depression [34], and labor induction [35] are risk factors for ASD, also, some studies have shown that low maternal residential greenness exposure increases the risk of mental disorders such as ASD.The neonatal risk factors such as not breast feeding [36], low weight, jaundice, congenital heart disease, low birth weight, very low birth weight, and small size for gestational age are increase risk of ASD [33,[37][38][39].The results of this study showed that post-term birth should be added to the above as a risk factor for ASD.
Although there is insufficient evidence to implicate any one perinatal or neonatal factor in autism etiology, the studies using optimality scales provide some evidence to suggest that exposure to multiple neonatal complications may increase autism risk.It also is important to note that the observed association between perinatal and neonatal complications and the risk of ASD may reflect the consequences of previous prenatal complications [40].
Previously, only a meta-analysis to date has been conducted about the association between post-term birth Fig. 2 Forest plot of the association between post-term birth and ASD and the risk of ASD.They showed that post-term birth was not a risk factor for ASD risk (OR = 1.14; 95% CI 0.58 to 2.24) [10].However, databases were searched until March 2007 and authors had not included the database of the Web of Science.Therefore, the findings of this meta-analysis are required to be updated.In the present meta-analysis, post-term birth is a risk factor for ASD (RR = 1.34, 95% CI 1.10 to 1.58) and (OR = 1.47, 95% CI 1.03 to 1.91).
Post-term birth contributes to severe morbidities for the mother and child, including macrosomia, shoulder dystocia, birth injury, fourth degree perineal laceration, fetal compromise, antenatal and postpartum hemorrhage, fetal dysmaturity, labor > 24 h and newborn respiratory distress syndrome [41].
Post-term birth is considered a complex biological process, which is related to factors such as the duration of pregnancy, the delivery, the fetus status of the intrauterine, and the gender and fetal-placental system.Studies showed its prevalence in Austria, Belgium, Germany, Turkey, and Denmark to be 0.4%, 0.6%, 2.3%, 0.76%, and 8.1%, respectively, this rate varies between 1 and 2.5% in America and Canada and 1.16% in China [42].
In some studies, it has been found that the occurrence of post-term birth is more common in female fetuses [43].However, some studies have not reported a significant difference in the occurrence of post-term birth between genders [42].Differences in the diagnosis of post-term birth and potential racial differences can lead to varying results in studies [42].Additionally, the sex ratio may be influenced by environmental factors such as wars and natural disasters [44,45].Also, due to the high prevalence of ASD in males, it is possible to justify the heterogeneity between studies to some extent.
Post-term birth and ASD have not yet been fully elucidated, and there are several reasons and pathways for it.First, a post-term birth typically has a higher risk for perinatal problems such as prolonged labor, cephalopelvic disproportion, and shoulder dystocia, which are associated with perinatal oxygen deficiency followed by neurobehavioral problems [46].A second explanation is uteroplacental insufficiency: a non-optimal 'old' placenta offers fewer nutrients and less oxygen than a term fetus requires.The lack of nutrients and oxygen may predispose to abnormal fetal development and this may lead to abnormal neurobehavioral development [47].
Thirdly, it is possible that a disturbance of the 'placental clock' , which regulates the duration of pregnancy, is involved.A marker of this clock is the placental secretion of corticotrophin-releasing hormone, which is lower in post-term deliveries than in term deliveries, is a sign of this clock, and its principal regulator of the maternal-fetal hypothalamic-pituitary-adrenal (HPA) axis.It has been suggested that placental endocrine dysfunction or maternal stress at critical times during fetal development may influence the fetal HPA axis, leading to neuroendocrine abnormalities that could increase the child more susceptibility to neurobehavioral disorder later in life [3,48].Also maternal exposure to environmental stressors including noise, air pollution, heat, and traffic density can aggravate the disruption of the placental clock and fetal hypothalamic-pituitary-adrenal axis [49].However, some researchers have raised doubts about the connection between perinatal and neonatal issues and autism, questioning whether it is a direct cause.They suggest that the correlation may be influenced by birth order, as complications during these periods are more common in first-born, fourth-born, and later offspring.It has been noted that individuals in these birth orders have a higher likelihood of developing autism [50].
This study has some limitations.First, a high heterogeneity among the results of our study was the limitation of this meta-analysis.However, with subgroup analysis based on gender, this heterogeneity was resolved in males.This study included 1,412,667 participants who reported that post-term birth was a risk factor for ASD.In addition, publication bias did not occur among the results and these were strengths of the present meta-analysis.

Conclusion
Post-term birth still occurs relatively frequently (up to 5-10%) even in developed countries.Our results showed that post-term birth is an increased risk of ASD, although high heterogeneity was found among the studies reported based on adjusted and crude forms, however, after subgroup analysis by gender, this heterogeneity disappeared among males.

Fig. 1
Fig. 1 Flowchart of the selection process of the studies in the different phases of the meta-analysis

Fig. 3
Fig. 3 Funnel plot of the association between post-term birth and ASD

Table 1
Summary results of the included studies

Table 3
Results of subgroup analysis of post-term birth and autism spectrum disorders (ASD) based on OR