Prenatal and perinatal risk factors and the clinical implications on autism spectrum disorder

Abstract

Prenatal and perinatal factors may increase the risk of autism spectrum disorder. However, little is known about whether unaffected siblings of probands with autism spectrum disorder also share the phenomenon and whether the prenatal/perinatal factors are related to the clinical severity of autistic symptoms. We compared the frequency of prenatal and perinatal factors among 323 probands with autism spectrum disorder (mean age ± standard deviation, 10.7 ± 3.5 years; males, 91.0%), 257 unaffected siblings (11.7 ± 4.5; 42.8%), and 1504 typically developing controls (8.9 ± 1.6 years; 53.1%); and investigated their effects on the severity of autistic symptoms. We found that probands with autism spectrum disorder and their unaffected siblings had more prenatal/perinatal events than typically developing controls with higher numbers of prenatal/perinatal factors in probands than in unaffected siblings. The prenatal/perinatal events were associated with greater stereotyped behaviors, social-emotional problems, socio-communication deficits, and overall severity. We also found that six prenatal/perinatal factors (i.e. preeclampsia, polyhydramnios, oligoamnios, placenta previa, umbilical cord knot, and gestational diabetes) were associated with the severity of autistic symptoms, particularly stereotyped behaviors and socio-communication deficits. Our findings suggest that prenatal and perinatal factors may potentially moderate the clinical expression of autism spectrum disorder. The underlying mechanism warrants further research.

Keywords

autism spectrum disorder perinatal prenatal sibling symptom

Introduction

Autism spectrum disorder (ASD) is a multi-factorial neurodevelopmental disorder that afflicts approximately 1 in 68 children (Centers for Disease Control and Prevention, 2014) in the United States and 1 in 100 children in Taiwan (Chen et al., 2017). The development of ASD is associated with a sophisticated collection, both genetic and environmental variables (Meek et al., 2013). Evidence suggests an apparent genetic predisposition to autism with concordance rates of classical autism as 60% and of the broader autistic spectrum as 92% in monozygotic twins (Bailey et al., 1995). In the absence of complete concordance, environmental factors such as prenatal and perinatal risk factors are also likely involved. In a recent twin study, shared environmental factors were significant (30%; 95% confidence interval [CI], 8%–47%), though smaller than genetic influences (56%; 95% CI, 37%–82%), and could explain a substantial part of variance (Colvert et al., 2015). A new approach including a comprehensive assessment of environmental factors is necessary to elucidate the mixture of this complex disorder.

Previous studies regarding prenatal/perinatal factors in ASD revealed controversial results. Suboptimal conditions for pregnancy and delivery have been addressed to increase the risk of developing ASD (Bryson et al., 1988; Gillberg and Gillberg, 1983). Gardener et al. (2009) provided the first quantitative review and meta-analysis of maternal pregnancy-related factors in autism. Among more than 50 factors, vaginal bleeding, gestational diabetes, medication exposure, advanced parental age, and being the first born were considered to be associated with a risk for autism. Their subsequent meta-analysis focused on perinatal and neonatal factors indicated that the risk for autism was also related to abnormal presentation, maternal hemorrhage, birth injury, congenital malformation, low birth weight (LBW), small for gestational age, hyperbilirubinemia, and so on (Gardener et al., 2011). Nevertheless, there was not sufficient evidence to implicate any single perinatal or neonatal factor in autism etiology. Guinchat et al. (2012) strongly suggested that some risk factors (i.e. maternal preeclampsia, diabetes, and infections) had not been sufficiently studied to draw any conclusion.

The prenatal/perinatal factors may provide clues to the etiology of ASD. Literature suggests that the association between obstetrical complications and autism may signal a shared etiology rather than a causal etiology (Deykin and MacMahon, 1980), corresponding to the concept of shared genetic risk (Bolton et al., 1997; Gillberg and Gillberg, 1983). Unaffected siblings of probands with ASD showed similar profiles to the probands but different from controls (Bolton et al., 1997; Bryson et al., 1988; Glasson et al., 2004), suggesting that autism is not necessarily caused by the complications but rather the complications may be the epiphenomenon of a disorder with strong genetic loading (Zwaigenbaum et al., 2002). On the other hand, multiple risk factors during pregnancy appear to be associated with an elevated risk of ASD in the offspring, supporting the hypothesis that environmental influences in conjunction with genetic vulnerability contribute to the risk of developing ASD (Maramara et al., 2014). Although prenatal/perinatal factors have been reported to increase the risk of ASD, how these factors impact the clinical severity is largely unknown. It is interesting that individuals with typical autism suffered from more maternal complications during pregnancy or delivery than those with Asperger syndrome or atypical autism (Glasson et al., 2004), suggesting a potential impact of these factors on clinical subtypes of ASD. Based on this finding, it is questionable whether these factors can moderate the clinical presentation of ASD, resulting in various levels of severity on core symptoms.

Some studies calculated the optimality score of all contributing perinatal factors and tested its relationship with severity of autistic symptoms. One of them indicated that the optimality score was significantly associated with the number of autistic symptoms (Bolton et al., 1997). The heterogeneity of autism may offer some insight into the potential modifiability of autistic symptoms. It is therefore recommended to take an endophenotypic approach by exploring the degree of impairment within each of the core symptoms (McBride et al., 1996). Wallace et al. (2008) have examined the association between obstetric complications and autistic symptoms using the Autism Diagnostic Interview–Revised (ADI-R) domain scores. They found that hypertension, preeclampsia, and generalized edema were associated with a higher level of communication deficits, and hypertension, albuminuria, and generalized edema were associated with more severe repetitive behaviors, in support with associations between hypertension-related obstetric conditions and symptom severity. They also suggested assessing the symptom severity using instruments designed specifically for ASD, such as the Social Responsiveness Scale (SRS; Constantino et al., 2003).

In this study, we hypothesized that the prenatal/perinatal factors might be associated with the severity of ASD core symptoms. We first adopted an unaffected sibling design to compare the rates of prenatal/perinatal factors for probands with ASD and their unaffected siblings, relative to that of typically developing controls (TDC). We further tested the relationship between the total number of the prenatal/perinatal factors and the severity of autistic symptoms using several instruments developed for ASD research. In addition, we attempted to identify the specific prenatal/perinatal factors that predict the severity of autistic symptom domains.

Methods

Participants

There were 323 probands with ASD (ages, mean ± standard deviation, 10.7 ± 3.5 years; male, 91.0%) recruited from the Department of Psychiatry of two medical centers from Northern Taiwan. The diagnoses of ASD were made by child psychiatrists according to the Diagnostic and Statistical Manual of Mental Disorders (4th ed.; DSM-IV) diagnostic criteria of autistic disorder (n = 186) and Asperger’s disorder (n = 137) and were confirmed by the ADI-R (Lord et al., 1994) interview with parents. The patients with ASD who had known chromosomal disease (e.g., Fragile X syndrome and Rett’s disorder) were excluded. A total of 273 siblings agreed to participate in the study; 16 of them fulfilled the diagnosis of ASD. The remaining 257 siblings (11.7 ± 4.5 years old; male, 42.8%) were clinically evaluated and demonstrated no evidence of having ASD symptoms or diagnosis. We recruited 1504 TDC (8.9 ± 1.6; 53.1%) from the community through advertisement or from regular classes in primary or high schools and ensured their absence of mental problems by questionnaire screening. The three groups showed comparable maternal gestational age at delivery. Probands with ASD had lower birth body weight than TDC. Parental ages at child delivery were slightly older in the ASD group than in the TDC group (Supplementary Table 1).

Procedure

This study was approved by the Research Ethics Committee of the National Taiwan University Hospital, Taiwan (ASD and siblings: 9561709027, ClinicalTrials.gov: NCT00494754; TDC: 201005041R, ClinicalTrials.gov: NCT01247662) before study implementation. Written informed consent was obtained from the participants after the research goals and procedures were well explained.

The parents were then interviewed with the ADI-R by well-trained research assistants to validate the clinical diagnosis of ASD and with semi-structured questions about prenatal/perinatal factors during pregnancy and delivery. The National Health Insurance in Taiwan covers most medical expenses during routine prenatal examinations, and the findings at prenatal visits are recorded on the Maternal Health Booklet kept by each expectant mother according to Taiwan’s national policy. Therefore, the mothers can answer the questions regarding prenatal/perinatal events based on the records. During the interview, we first asked open questions to explore any prenatal and perinatal events; then, we asked closed questions to check each risk factor detailed below. For any positive report, we clarified the diagnosis procedure and the required intervention to ensure the validity of the diagnosis. Afterward, parents completed the SRS (Constantino et al., 2003; Gau et al., 2013) and Social Communication Questionnaire (SCQ; Gau et al., 2011; Rutter et al., 2003).

Prenatal and perinatal factors

We collected prenatal and perinatal events including preeclampsia (defined as blood pressure higher than 140/90 mmHg, proteinuria, and/or edema after gestational age 20 weeks), gestational diabetes (confirmed by 100 g glucose tolerance test: two of the four sugar levels higher than the thresholds: 95 (0 h), 180 (1 h), 155 (2 h), and 140 (3 h) mg/dL), abnormal bleeding, edema, Rubella infection, Cesarean section (CS), LBW (<2000 g), cardiopulmonary resuscitation (CPR), incubation, and phototherapy for jaundice. In ASD probands and unaffected siblings, a more detailed history was obtained, containing emergent CS, induction for delivery, Herpes infection, placenta abruptia and previa, polyhydramniosis and oligoamniosis (by routine ultrasound examinations), prolonged labor (>24 h), umbilical cord knot, and cord entanglement (after the delivery). For those with a prenatal/perinatal event, medical records were checked when available (around 15%) to validate the reports from the interview. The maternal reports were generally consistent with medical records. We also calculated the total number of prenatal/perinatal factors for each participant for further analysis.

Instruments

The ADI-R (Lord et al., 1994) is a standardized, semi-structured interview with a caregiver for children aged from 18 months on. It covers most developmental and behavioral aspects of ASD, including reciprocal social interaction, communication, and stereotyped/repetitive behaviors. The ratings were based on assessment under current conditions and under the most severe state at 4–5 years. The Chinese version of the ADI-R was approved by the Western Psychological Services and has been applied in several studies (Gau et al., 2011).

The SCQ is a parent-report questionnaire to screen autistic symptoms for individuals above 4 years old. It contains 40 yes-or-no items that correspond to the three core symptoms (impairment in social development, communication, and stereotyped/repetitive behaviors), with good convergent validity to the ADI-R. The Chinese version of the SCQ has satisfactory reliability and validity (Gau et al., 2011) and has been widely used in genetic and neuroimaging research in Taiwanese ASD populations.

The SRS is a 65-item 4-point Likert-type questionnaire for quantifying autistic traits by the frequency of each behavior (Constantino et al., 2003). It has good reliability and concurrent validity to the ADI-R and the SCQ. The SRS scores demonstrate a continuous distribution in the general population. The psychometric properties of its Chinese version have been established, with a valid four-factor structure including social awareness, social communication, social emotion, and stereotyped behaviors (Gau et al., 2013). The Chinese SRS has been widely used in ASD research.

Statistical analysis

The frequency of each prenatal/perinatal factor were compared between probands with ASD, unaffected siblings, and TDC using logistic regression controlling for sex because fetal sex has a major effect on maternal and obstetric outcomes (Al-Qaraghouli and Fang, 2017). Bonferroni correction was adopted in post hoc comparison (Bonferroni correction significant level p < 0.0167 in Table 1). The total number of prenatal/perinatal factors was compared between probands and unaffected siblings using the analysis of covariance, with a more stringent control strategy over maternal age upon delivery together with sex. The reason to control for maternal age is based on recent evidence showing that advanced maternal age may increase the risk for several complications (Lean et al., 2017). To examine the associations between each prenatal/perinatal factor and the symptom subscores in probands, we used a general linear model with adjustment for sex, age, and maternal age upon delivery, because autistic symptom severity may change with sex and age in individuals with ASD (Van Wijngaarden-Cremers et al., 2014). Each symptom subscore and the total scores of the SCQ and SRS were treated as dependent variables, with each prenatal/perinatal factor as independent variable. We used the “six factors” that were associated with at least one symptom domain to categorize probands and compare the clinical profile between probands with any of the six factors and without. To establish a predictive model for each symptom subscore, regression analysis with backward selection was applied to select the significant correlates from the models including all six prenatal/perinatal factors that showed statistically significant effects on autistic symptoms in univariate analyses. Factors with p < 0.05 remained in the model. We used SAS v.9.2 for all statistical analyses. The significance level was p < 0.05 (except the post hoc comparison in Table 1).

Table 1.

Comparison of the frequency of prenatal and perinatal risk factors in probands with autism spectrum disorders (ASD), unaffected siblings (Sib), and typically developing controls (TDC).

	ASD (n = 323)		Unaffected sibling (n = 257)		TDC (n = 1504)		ASD versus TDC		Sib versus TDC		ASD versus Sib
	N	%	N	%	N	%	OR	95% CI	OR	95% CI	OR	95% CI
Perinatal factors
C/S	130	40.25	103	40.08	437	31.28	1.40*	(1.08–1.82)	1.50**	(1.14–1.97)	0.12	(0.01–1.20)
LBW < 2000	9	2.83	3	1.2	12	0.88	3.46*	(1.32–9.09)	1.30	(0.36–4.66)	3.94	(0.82–18.92)
CPR	30	9.29	12	4.67	23	1.55	4.86***	(2.70–8.74)	3.34**	(1.63–6.83)	1.65	(0.75–3.61)
Incubation	65	20.12	23	8.95	81	5.47	3.84***	(2.63–5.60)	1.74	(1.07–2.82)	2.21**	(1.23–3.95)
Phototherapy	138	77.53	82	68.33	353	23.84	10.36***	(7.09–15.14)	7.05***	(4.71–10.56)	1.42	(0.79–2.56)
Prenatal factors
Abnormal bleeding	72	22.29	46	17.9	127	8.57	2.73***	(1.95–3.84)	2.42***	(1.67–3.51)	1.50	(0.17–13.60)
Gestational diabetes	4	1.24	1	0.39	11	0.74	1.78	(0.52–6.12)	0.52	(0.07–4.05)	3.43	(0.57–20.82)
Preeclampsia	8	2.48	2	0.78	8	0.58	5.43**	(1.76–16.77)	1.20	(0.25–5.69)	1.51	(0.74–3.10)
Rubella infection	3	0.93	1	0.39	4	0.27	4.59	(0.83–25.27)	1.38	(0.15–12.50)	0.75	(0.07–8.32)
Herpes infection	3	0.95	0	0							–	–
Emergent C/S	56	17.7	31	12.6							1.35	(0.79–2.32)
Labor induction	82	25.9	54	21.9							1.08	(0.69–1.70)
Prolonged labor (>24 h)	15	4.7	5	2.0							2.43	(0.74–7.89)
Umbilical cord knot	2	0.63	0	0							–	–
Placenta previa	17	5.36	8	3.24							1.18	(0.46–2.99)
Hydramnios	2	0.63	2	0.81							0.39	(0.05–2.82)
Oligohydramnios	3	1.51	2	0.81							0.59	(0.10–3.58)

C/S: Cesarean section; LBW: low birth weight; CPR: cardiopulmonary resuscitation; OR: odds ratio; CI: confidence interval.

Post hoc comparison: *p < 0.0167 (significant level of Bonferroni correction: 0.05/3 = 0.0167), **p < 0.01, ***p < 0.001.

Results

Prenatal/perinatal factors among ASD, unaffected siblings, and TDC

Compared to TDC, probands with ASD had a higher frequency of CS, LBW, CPR, phototherapy, and in need of incubation after birth. Their mothers also experienced more abnormal bleeding and preeclampsia during pregnancy than TDC. The unaffected siblings showed the same pattern of more prenatal/perinatal factors, except for LBW, incubation, and preeclampsia. The comparisons between probands and unaffected siblings were not statistically significant, except that probands were more likely to use incubation than unaffected siblings (Table 1).

Multiplex versus singleton ASD probands

Among the probands, those who had siblings with ASD (multiplex, n = 11) did not show higher frequencies of any prenatal/perinatal factors than those without (Supplementary Table 2). The total number of prenatal/perinatal factors was not higher in multiplex probands than singletons, either (F_(4,309) = 0.17, p = 0.681). However, the symptom severity of multiplex probands was higher than that of singletons on the ADI-R nonverbal (regression coefficient estimate, β = 3.99, standard error (SE) = 1.35, p = 0.003) and verbal communication (β = 2.86, SE = 0.92, p = 0.002), stereotyped/repetitive behaviors (β = 2.16, SE = 0.81, p = 0.008) at the most severe state at 4–5 years, and more severe current stereotyped/repetitive behaviors on the ADI-R (β = 2.07, SE = 0.74, p = 0.006) and SCQ (β = 1.78, SE = 0.79, p = 0.025).

The total number of prenatal/perinatal factors

Figure 1 presents the number of risk factors for probands and unaffected siblings. There were significantly higher proportion of probands (n = 264, 83.3%) than unaffected siblings (n = 178, 72.1%) who had at least one prenatal/perinatal factors (x² = 10.30, p = 0.001). In probands, around 29.3% had one factor, and 55.8% had two or more factors. In general, probands had more factors (1.94 ± 1.55) than unaffected siblings (1.38 ± 1.29). To compare the overall frequency, probands were reported with more prenatal/perinatal factors than unaffected siblings (x² = 21.22, p < 0.0001). The significance remained when controlled for sex and maternal age upon delivery (F = 7.41, p = 0.007).

Figure 1.

The total number of prenatal/perinatal factors in probands with autism spectrum disorders (ASD) and their unaffected siblings. The total number of prenatal/perinatal factors are the sum of the following factors: preeclampsia, herpes infection, vaginal bleeding, gestational diabetes, edema, placenta abruption, placenta previa, hydramnios, oligohydramnios, emergent Cesarean section, labor induction, prolonged labor (>24 h), umbilical cord entanglement, umbilical cord knot, low birth body weight (<2000 g), acute resuscitation, usage of incubation, and phototherapy for jaundice.

Prenatal/perinatal factors and autistic symptoms

The number of the factors in probands was correlated with overall symptom severity on SCQ and SRS. After controlling for sex, age, and maternal age at delivery, the significance remained (Table 2). The same pattern was demonstrated on symptom subscores, particularly the stereotyped behaviors (SCQ and SRS), the social-emotional problems, and socio-communication deficits (SRS).

Table 2.

The association of the total number of prenatal/perinatal factors and the autistic symptoms (overall severity and specific symptom domains) in probands with autism spectrum disorders, by linear regression model with adjustment for sex, age, and maternal age upon delivery.

	β	SE	t	p
Social Communication Questionnaire
Total scores	0.56	0.25	2.24	0.025
Social deficits	0.23	0.16	1.43	0.152
Communication deficits	0.11	0.06	1.62	0.106
Stereotyped behaviors	0.25	0.08	3.11	0.002
Social Responsiveness Scale
Total scores	3.54	1.03	3.29	0.001
Social communication deficits	1.63	0.44	3.61	0.000
Stereotyped behaviors	0.94	0.26	3.53	0.001
Social awareness deficits	0.34	0.21	1.52	0.129
Social-emotional problems	0.64	0.21	2.87	0.004

SE: standard error.

Among the factors, we found that six of them were significantly associated with symptom severity by univariate analysis adjusting for sex and age. These factors were preeclampsia, polyhydramnios, oligoamniosis, placenta previa, umbilical cord knot, and gestational diabetes (Supplementary Table 3). Probands who had at least one of these six factors (n = 34), compared to those without, had more stereotyped behaviors across the assessments by the ADI-R, SCQ, and SRS; and more social and communication deficits consistently in the SCQ and SRS (Table 3).

Table 3.

The association of the presence of six prenatal factors and clinical autistic symptoms in probands with autism spectrum disorders, by a general linear model with adjustment for sex and age.

Autistic symptoms	Probands with at least one factor		Probands without any of the six factors		B	SE	F	p
Autistic symptoms	Mean	SD	Mean	SD	B	SE	F	p
Autism Diagnostic Interview–Revised (severe)
A: Social reciprocal interaction	20.67	5.13	20.30	6.19	0.33	1.13	0.09	0.770
Bv: Communication (verbal)	15.42	3.21	15.40	4.52	−0.02	0.81	0	0.979
Bn: Communication (non-verbal)	7.64	2.13	7.93	3.10	−0.34	0.56	0.37	0.543
C: Restricted interests/stereotyped behaviors	8.91	2.57	7.00	2.58	1.84	0.47	15.23	0.000
Autism Diagnostic Interview–Revised (current)
A: Social reciprocal interaction	10.12	4.97	10.43	4.55	0.34	0.84	0.17	0.684
C: Restricted interests/stereotyped behaviors	6.36	2.78	5.25	2.39	1.06	0.45	5.63	0.018
Social Communication Questionnaire
Total scores	22.21	5.57	18.07	7.00	4.03	1.24	10.61	0.001
Social deficits	12.56	5.14	10.30	5.37	2.17	0.94	5.30	0.022
Communication deficits	4.56	1.94	4.11	1.93	0.43	0.35	1.21	0.226
Stereotyped behaviors	6.12	3.07	4.72	2.48	1.38	0.46	8.86	0.003
Social Responsiveness Scale
Total scores	99.24	29.16	88.50	26.03	10.60	4.71	5.07	0.025
Social communication deficits	38.21	14.01	32.46	12.29	5.65	2.24	6.34	0.012
Stereotyped behaviors	22.03	7.83	18.70	6.91	3.33	1.25	7.08	0.008
Social awareness deficits	20.82	5.44	21.04	4.86	−0.21	0.89	0.06	0.810
Social-emotional problems	18.18	6.31	16.30	6.06	1.83	1.08	2.88	0.091

SD: standard deviation; SE: standard error; ADI-R: Autism Diagnostic Interview–Revised; ASD: autism spectrum disorder.

The six prenatal factors included preeclampsia, gestational diabetes, placenta previa, hydramnios, oligohydramnios, and umbilical cord knot. The ADI-R interviews of the most severe condition around 4–5 years old revealed that ASD probands scored higher than cut-off in the “qualitative abnormalities in reciprocal social interaction” (cut-off = 10), “qualitative abnormalities in communication, verbal” (cut-off = 8), “qualitative abnormalities in communication, nonverbal” (cut-off = 7), and “restricted, repetitive, and stereotyped patterns of behaviors” (cut-off = 3).

Final predictive models of autistic symptoms

For the six significant prenatal/perinatal factors, the severity of stereotyped behaviors of probands can be predicted by oligoamniosis, placenta previa, polyhydramniosis, and preeclampsia. Preeclampsia and oligoamniosis were the two most significant factors to predict socio-communication deficits, stereotyped behaviors, and total scores of SRS (Table 4).

Table 4.

Predictive models to select significant determinants (from the six prenatal/perinatal factors) for autistic symptoms by backward selection.

	Social Communication Questionnaire									Social Responsiveness Scale
	Social deficits			Stereotyped behaviors			SCQ total scores			Communication deficits			Stereotyped behaviors			SRS total scores
	β	SE	F	β	SE	F	β	SE	F	β	SE	F	β	SE	F	B	SE	F
Age	0.36	0.08	19.1**				0.28	0.11	6.71	0.53	0.20	7.25*	0.35	0.11	10.04**	1.39	0.41	11.31**
Preeclampsia	4.65	1.99	5.49							12.15	4.71	6.65	6.14	2.64	5.42	23.14	9.89	5.47
Oligohydramnios				3.62	1.47	6.08				17.33	7.16	5.85	9.56	4.01	5.7	35.84	15.03	5.69
Placenta previa	2.76	1.29	4.58	1.29	0.63	4.16	4.49	1.72	6.8*
Hydramnios				4.79	1.80	7.1*

SCQ: Social Communication Questionnaire; SRS: Social Responsiveness Scale; SE: standard error.

The variables for selecting predictors included sex, age, and the six prenatal/perinatal factors (i.e., preeclampsia, gestational diabetes, umbilical cord knot, oligohydramnios, placenta previa, and hydramnios).

p < 0.01; **p < 0.005.

Discussion

As one of few studies examining the prenatal/perinatal factors for ASD using a sibling design and correlating these factors with ASD core symptoms, we have the following findings. First, similar to probands with ASD, unaffected siblings had higher frequencies of prenatal/perinatal factors than TDC. Second, probands with ASD had higher total numbers of prenatal/perinatal factors than unaffected siblings. Third, in ASD probands, the total number of prenatal/perinatal factors was associated with overall symptom severity, as well as specific symptoms such as stereotyped behaviors.

Consistent with previous studies, unaffected siblings demonstrated a higher frequency of prenatal/perinatal factors than TDC yet without a significant difference from the probands (Bolton et al., 1997; Bryson et al., 1988; Glasson et al., 2004), suggesting that they have a similar profile of obstetric complications to probands. Our finding supports Glasson et al.’s (2004) hypothesis that obstetric complications may represent a compromised gestational environment to which all offspring are exposed; however, autism develops only in some offspring, possibly because of their particular genotype that results in less tolerance to the prenatal insults. In other words, ASD risk genes caused the vulnerability of the probands in response to prenatal adversities, implying a possibility for gene–environment interaction on fetal brain development.

Considering that the rare events in our cohort may render a false-negative bias in comparing the frequency of each factor, we further compared the total number of prenatal/perinatal events between probands and unaffected siblings and found a higher total number in probands, consistent with a previous finding that multiple risk factors during pregnancy appeared to be associated with a higher risk of ASD in the offspring (Maramara et al., 2014). These findings also support Glasson et al.’s (2004) explanation that unaffected siblings share part of the genes with probands, thus only show fewer symptoms and fewer complications.

Our novel finding that the number of prenatal/perinatal events in probands was associated with the severity of overall autistic symptoms and stereotyped behaviors is of particular interest. This “dose effect” of prenatal/perinatal events is compatible with a previous finding that the optimality scores were associated with symptom numbers in autism (Bolton et al., 1997), suggesting a relationship between the extent of environmental adversities and the severity of clinical symptoms. Whether and how the accumulation of factors relate to the clinical expression of ASD warrants further investigation.

Among the prenatal/perinatal factors, neonatal jaundice that required phototherapy was found with the highest odds ratio (OR), consistent with a meta-analysis that showed jaundice was associated with a risk for ASD (Amin et al., 2011). However, we do not find its association with ASD symptom severity as revealed in a study from China (Duan et al., 2014). Instead, the other six factors are in association with autistic symptoms (i.e. preeclampsia, placenta previa, gestational diabetes, polyhydramnios, oligoamniosis, and umbilical cord knot). These factors involve the mechanisms of placentation and fetal ingestion and metabolism of amniotic fluid. Placental abnormalities (Anderson et al., 2007), umbilical cord complications (Gardener et al., 2011), and gestational diabetes (Gardener et al., 2009) have been reported to occur more frequently in children who develop autism, while the relationship between the excess or insufficiency in amniotic fluid and the risk of ASD was scarcely investigated. Among the six factors, preeclampsia was known to be associated with fetal growth restriction (Gruslin and Lemyre, 2011) and subsequent cognitive deficits, motor development problems (Kronenberg et al., 2006) and communication deficits (Wallace et al., 2008). Several studies also found an increased risk of ASD among children exposed to preeclampsia (OR = 1.24–2.36; Dodds et al., 2011; Mann et al., 2010) and proposed that enhanced ASD risk primarily involved severe preeclampsia and placental insufficiency (Walker et al., 2015). Future studies may elucidate the moderating mechanism underlying these associations.

The findings of the correlations between the total number of prenatal/perinatal events (or the presence of six specific factors) and symptom severity, particularly stereotyped behaviors and communication deficits, are in line with a previous study showing that hypertension, preeclampsia, albuminuria, and generalized edema were significantly associated with the severity of communication deficits and repetitive behaviors (Wallace et al., 2008). Stereotyped or repetitive behaviors are not uncommon in children suffering from birth brain insults, in that hypoxia may lead to long-term changes in dopaminergic function and eventually relate to the repetitive behaviors observed in patients with autism (Boksa and El-Khodor, 2003). Other than dopamine, the role of serotonin has also been discussed in several studies focused on the clinical efficacy of selective serotonin reuptake inhibitor in treating stereotyped behaviors of ASD (Sugie et al., 2005). As a major source of serotonin that is essential for brain development (Petraglia et al., 2010), the placenta may play a potential role in modulating the clinical phenotype of the fetus later in the development. How these alterations work on the sensitive periods of neurodevelopment and subsequently result in more severe clinical phenotypes are intriguing and worth further research.

The results of this study should be interpreted cautiously. The data of prenatal/perinatal events reported by mothers were subject to recall bias. Mothers having a child with ASD diagnosis might overestimate the nonspecific complications such as vaginal bleeding or dysfunctional labor (Ronsmans et al., 1997). However, the report on major obstetric complications such as preeclampsia was more reliable (O’Callaghan et al., 1990; Ronsmans et al., 1997) because these diagnoses need to be established by obstetricians based on laboratory examinations. A similar question is whether the farther from the date of delivery (the older the children), the memory was less reliable. We had tested this possibility but found that children’s age had no significant effect on the total number of prenatal/perinatal factors in either probands or unaffected siblings. Older children (⩾ mean age 11.2) had comparable total numbers of prenatal/perinatal risk factors for younger children (F_(4,309) = 1.86, p = 0.173). Meanwhile, older mothers (⩾ mean age 42) reported similar numbers of prenatal/perinatal factors compared to younger mothers (F_(4,309) = 0.05, p = 0.082). There was neither a significant age nor group interaction on the total number of prenatal/perinatal factors, suggesting that the difference between probands and unaffected siblings did not increase in older children. Hence, it is unlikely that recall bias significantly influences our results. Second, several obstetric complications were too rare for us to conduct group comparisons. Thus, a false-negative result should be considered. A larger cohort is necessary to detect a small-effect difference. Nevertheless, a representative sample with sibling design as well as carefully controlled covariates such as maternal age upon delivery constitute the strengths of this study. Our analysis of clinical symptoms based on comprehensive assessment provides a thorough view of the possible implications of prenatal/perinatal factors on ASD psychopathology.

In conclusion, this study demonstrated that prenatal/perinatal factors occurred more frequently in both ASD probands and their unaffected siblings. Probands had a higher number of obstetric complications than unaffected siblings, which may be associated with more severe autistic symptoms. The presence of some factors was associated with more severe overall autistic symptoms and/or stereotyped behaviors. Our findings provide evidence to support that prenatal/perinatal factors might modify the phenotype expression of ASD. The underlying mechanism warrants further investigation.

Supplemental Material

AUT772813_Lay_Abstract – Supplemental material for Prenatal and perinatal risk factors and the clinical implications on autism spectrum disorder

Supplemental material, AUT772813_Lay_Abstract for Prenatal and perinatal risk factors and the clinical implications on autism spectrum disorder by Yi-Ling Chien, Miao-Chun Chou, Wen-Jiun Chou, Yu-Yu Wu, Wen-Che Tsai, Yen-Nan Chiu and Susan Shur-Fen Gau in Autism

Supplemental Material

AUT772813_Supplementary_material – Supplemental material for Prenatal and perinatal risk factors and the clinical implications on autism spectrum disorder

Supplemental material, AUT772813_Supplementary_material for Prenatal and perinatal risk factors and the clinical implications on autism spectrum disorder by Yi-Ling Chien, Miao-Chun Chou, Wen-Jiun Chou, Yu-Yu Wu, Wen-Che Tsai, Yen-Nan Chiu and Susan Shur-Fen Gau in Autism

Footnotes

Acknowledgements

Y.-L.C. analyzed and interpreted the data and drafted the manuscript. S.S.-F.G. conceptualized and designed the study, acquired and analyzed the data, interpreted the results, revised the manuscript critically, and approved the final version. M.-C.C., W.-J.C., Y.-Y.W., W.-C.T., and Y.-N.C. acquired the data and revised the article critically for the important intellectual content. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Supplementary material

Supplementary material for this article is available online.

ORCID iD

Susan Shur-Fen Gau

References

Al-Qaraghouli

Fang

YMV

(2017) Effect of fetal sex on maternal and obstetric outcomes. Frontiers in Pediatrics 5: 144.

Amin

Smith

Wang

(2011) Is neonatal jaundice associated with autism spectrum disorders: a systematic review. Journal of Autism and Developmental Disorders 41(11): 1455–1463.

Anderson

Jacobs-Stannard

Chawarska

et al . (2007) Placental trophoblast inclusions in autism spectrum disorder. Biological Psychiatry 61(4): 487–491.

Bailey

Le Couteur

Gottesman

et al . (1995) Autism as a strongly genetic disorder: evidence from a British twin study. Psychological Medicine 25(1): 63–77.

Boksa

El-Khodor

(2003) Birth insult interacts with stress at adulthood to alter dopaminergic function in animal models: possible implications for Schizophrenia and other disorders. Neuroscience and Biobehavioral Reviews 27(1–2): 91–101.

Bolton

Murphy

Macdonald

et al . (1997) Obstetric complications in autism: consequences or causes of the condition? Journal of the American Academy of Child and Adolescent Psychiatry 36(2): 272–281.

Bryson

Smith

Eastwood

(1988) Obstetrical suboptimality in autistic children. Journal of the American Academy of Child and Adolescent Psychiatry 27(4): 418–422.

Centers for Disease Control and Prevention (2014) Prevalence of autism spectrum disorder among children aged 8 years—Autism And Developmental Disabilities Monitoring Network, 11 sites, United States, 2010. MMWR Surveillance Summary 63: 1–21.

Chen

Shen

Gau

(2017) The Mandarin version of the Kiddie-Schedule for Affective Disorders and Schizophrenia–Epidemiological version for DSM-5—a psychometric study. Journal of the Formosan Medical Association 116(9): 671–678.

10.

Colvert

Tick

McEwen

et al . (2015) Heritability of autism spectrum disorder in a UK population-based twin sample. JAMA Psychiatry 72(5): 415–423.

11.

Constantino

Davis

Todd

et al . (2003) Validation of a brief quantitative measure of autistic traits: comparison of the social responsiveness scale with the autism diagnostic interview-revised. Journal of Autism and Developmental Disorders 33(4): 427–433.

12.

Deykin

MacMahon

(1980) Pregnancy, delivery, and neonatal complications among autistic children. American Journal of Diseases of Children 134(9): 860–864.

13.

Dodds

Fell

Shea

et al . (2011) The role of prenatal, obstetric and neonatal factors in the development of autism. Journal of Autism and Developmental Disorders 41(7): 891–902.

14.

Duan

Yao

et al . (2014) Perinatal and background risk factors for childhood autism in central China. Psychiatry Research 220(1–2): 410–417.

15.

Gardener

Spiegelman

Buka

(2009) Prenatal risk factors for autism: comprehensive meta-analysis. British Journal of Psychiatry 195(1): 7–14.

16.

Gardener

Spiegelman

Buka

(2011) Perinatal and neonatal risk factors for autism: a comprehensive meta-analysis. Pediatrics 128(2): 344–355.

17.

Gau

Lee

Lai

et al . (2011) Psychometric properties of the Chinese version of the Social Communication Questionnaire. Research in Autism Spectrum Disorders 5(2): 809–818.

18.

Gau

Liu

et al . (2013) Psychometric properties of the Chinese version of the Social Responsiveness Scale. Research in Autism Spectrum Disorders 7(2): 349–360.

19.

Gillberg

(1983) Infantile autism: a total population study of reduced optimality in the pre-, peri-, and neonatal period. Journal of Autism and Developmental Disorders 13(2): 153–166.

20.

Glasson

Bower

Petterson

et al . (2004) Perinatal factors and the development of autism: a population study. Archives of General Psychiatry 61(6): 618–627.

21.

Gruslin

Lemyre

(2011) Pre-eclampsia: fetal assessment and neonatal outcomes. Best Practice & Research: Clinical Obstetrics & Gynaecology 25(4): 491–507.

22.

Guinchat

Thorsen

Laurent

et al . (2012) Pre-, peri- and neonatal risk factors for autism. Acta Obstetricia Et Gynecologica Scandinavica 91(3): 287–300.

23.

Kronenberg

Raz

Sander

(2006) Neurodevelopmental outcome in children born to mothers with hypertension in pregnancy: the significance of suboptimal intrauterine growth. Developmental Medicine and Child Neurology 48(3): 200–206.

24.

Lean

Derricott

Jones

et al . (2017) Advanced maternal age and adverse pregnancy outcomes: a systematic review and meta-analysis. PloS One 12(10): e0186287.

25.

Lord

Rutter

Le Couteur

(1994) Autism diagnostic interview-revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. Journal of Autism and Developmental Disorders 24(5): 659–685.

26.

McBride

Anderson

Shapiro

(1996) Autism research: bringing together approaches to pull apart the disorder. Archives of General Psychiatry 53(11): 980–983.

27.

Mann

McDermott

Bao

et al . (2010) Pre-eclampsia, birth weight, and autism spectrum disorders. Journal of Autism and Developmental Disorders 40(5): 548–554.

28.

Maramara

Ming

(2014) Pre- and Perinatal risk factors for autism spectrum disorder in a New Jersey Cohort. Journal of Child Neurology 29: 1645–1651.

29.

Meek

Lemery-Chalfant

Jahromi

et al . (2013) A review of gene-environment correlations and their implications for autism: a conceptual model. Psychological Review 120(3): 497–521.

30.

O’Callaghan

Larkin

Waddington

(1990) Obstetric complications in schizophrenia and the validity of maternal recall. Psychological Medicine 20(1): 89–94.

31.

Petraglia

Imperatore

Challis

(2010) Neuroendocrine mechanisms in pregnancy and parturition. Endocrine Reviews 31(6): 783–816.

32.

Ronsmans

Achadi

Cohen

et al . (1997) Women’s recall of obstetric complications in south Kalimantan, Indonesia. Studies in Family Planning 28(3): 203–214.

33.

Rutter

Bailey

Lord

(2003) Social Communication Questionnaire. Los Angeles, CA: Western Psychological Services.

34.

Sugie

Fukuda

et al . (2005) Clinical efficacy of fluvoxamine and functional polymorphism in a serotonin transporter gene on childhood autism. Journal of Autism and Developmental Disorders 35(3): 377–385.

35.

Van Wijngaarden-Cremers

Van Eeten

Groen

et al . (2014) Gender and age differences in the core triad of impairments in autism spectrum disorders: a systematic review and meta-analysis. Journal of Autism and Developmental Disorders 44(3): 627–635.

36.

Walker

Krakowiak

Baker

et al . (2015) Preeclampsia, placental insufficiency, and autism spectrum disorder or developmental delay. JAMA Pediatrics 169(2): 154–162.

37.

Wallace

Anderson

Dubrow

(2008) Obstetric and parental psychiatric variables as potential predictors of autism severity. Journal of Autism and Developmental Disorders 38(8): 1542–1554.

38.

Zwaigenbaum

Szatmari

Jones

et al . (2002) Pregnancy and birth complications in autism and liability to the broader autism phenotype. Journal of the American Academy of Child and Adolescent Psychiatry 41(5): 572–579.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.43 MB

0.21 MB