Condition

Autism Spectrum Disorders

Editors: Jill D. Morrow-Gorton MD, MBA; Esther Jolanda van Zuuren MD; Alexander Rae-Grant MD, FRCPC, FAAN

Overview and Recommendations
Background Information
History and Physical
Diagnosis
Management
Complications and Prognosis
Prevention and Screening
Guidelines and Resources
Patient Decision Aids
Patient Information
References

Next Section >

Background Information

Description

Autism spectrum disorders is a continuum of neurodevelopmental disorders characterized by early-onset impairments in social communication and interaction, as well as by restricted, repetitive interests and behaviors (BMJ 2018 May 21;361:k1674).

Also Called

ASD
Autism
Autistic disorder
Infantile autism
Childhood autism
Kanner syndrome
"Autism spectrum disorder" in the Diagnostic and Statistical Manual of Mental Disorders, Fifth edition (DSM-5) includes symptoms previously classified in the DSM-IV as "autistic disorder," "Asperger syndrome," or "pervasive developmental disorder not otherwise specified" (J Psychopharmacol 2018 Jan;32(1):3, Am Fam Physician 2016 Dec 15;94(12):972)

Epidemiology

Incidence/Prevalence

The world-wide prevalence of autism spectrum disorders (ASD) is reported to be 1%-1.5% (BMJ 2018 May 21;361:k1674).
Reported prevalence varies by geographic location.
- Based on limited data from the United Kingdom with many adults who never received a formal diagnosis, the estimated prevalence of ASD in adults was 1% (J Psychopharmacol 2018 Jan;32(1):3).
- In South Korea, the prevalence for children aged 7-12 years was 2.64%, but reporting included the period from 2005 to 2009 only (Annu Rev Public Health 2017 Mar 20;38:81).
- In Denmark, the prevalence was 1.5% (Annu Rev Public Health 2017 Mar 20;38:81).
- In Finland and Sweden, the prevalence was 1% (Annu Rev Public Health 2017 Mar 20;38:81).
- In China, the prevalence was 0.7% (Annu Rev Public Health 2017 Mar 20;38:81).
In the United States:
- Delayed diagnosis among persons of color, those with Hispanic ethnicity, and those with low socioeconomic status is associated with a lower ASD prevalence (Annu Rev Public Health 2017 Mar 20;38:81).
- STUDY SUMMARY
  prevalence of ASD among 8-year old children living in New Jersey, United States in 2016 reported to be 3.2% and about one-third of children with ASD have intellectual disability
  CROSS-SECTIONAL STUDY: Pediatrics 2023 Feb 1;151(2):e2022056594
  based on cross-sectional study
  
  4,661 children aged 8 years (81% male, 45% non-Hispanic White, 26% Hispanic, 20% non-Hispanic Black) with ASD from New Jersey Autism Study in the United States in 2000-2016 were assessed
  
  4,263 children with available data were classified according to cognitive status (88% had intelligence quotient [IQ] test scores and 19% were classified based on documentation regarding cognition)
  
  35% had intellectual disability, defined as IQ score ≤ 70 points or documented cognitive impairment by professional, discontinuation of IQ test due to nontestable status, or special education classification of cognitive impairment at school
  
  prevalence of ASD per 1,000 children comparing 2000 vs. 2016
  ASD overall 9.6 vs. 31.8
  
  ASD with intellectual disability 2.9 vs. 7.3
  
  ASD without intellectual disability 3.8 vs. 18.9
  
  prevalence of ASD has increased from 2000 to 2016 in children with low or mid socioeconomic status but has remained stable in children with high socioeconomic status
  
  compared to low socioeconomic status in analysis of 3,762 children who had IQ test scores available, prevalence of ASD without intellectual disability was higher among children living in
  mid-socioeconomic status (adjusted rate ratio 2, 95% CI 1.7-2.3)
  
  high socioeconomic status (adjusted rate ratio 1.9, 95% CI 1.6-2.3)
  
  Reference - Pediatrics 2023 Feb 1;151(2):e2022056594
- STUDY SUMMARY
  higher prevalence of autism spectrum disorders in children born in northeastern United States compared to elsewhere in continental United States
  CROSS-SECTIONAL STUDY: Am J Epidemiol 2017 Oct 1;186(7):834
  based on cross-sectional study
  
  116,430 female nurses were surveyed for autism spectrum disorders (ASD) diagnosis in their children born between 1989 and 1999 in continental United States
  
  509 children with ASD diagnosis were compared to 32,806 children without ASD diagnosis
  
  compared to children born elsewhere in continental United States, 1.5-fold increased prevalence of ASD diagnosis in children born in northeast United States
  
  compared to children aged 6 years living elsewhere in the continental United States,
  higher prevalence of ASD diagnosis in children aged 6 years living in northeast United States
  
  lower prevalence of ASD diagnosis in children aged 6 years living in southeast United States
  
  being born or living in portions of Indiana at age 6 years associated with greater ASD diagnosis in sensitivity analysis
  
  Reference - Am J Epidemiol 2017 Oct 1;186(7):834
- ASD prevalence is reported to be higher in boys than in girls.
  - Based on parent-reported data from the 2014-2016 National Health Interview Survey in the United States, the prevalence is 3.63% in boys aged 3-17 years and 1.25% in girls of the same age (CDC 2017 Nov)
  - STUDY SUMMARY
    prevalence of autism spectrum disorders 1.68% among children aged 8 years in United States during 2014
    CROSS-SECTIONAL STUDY: MMWR Surveill Summ 2018 Apr 27;67(6):1
    based on cross-sectional study
    
    325,483 children aged 8 years in United States across 11 sites were included in surveillance data for 2014 (representing 8% of total population of children aged 8 years in United States) collected by the Autism and Developmental Disabilities Monitoring (ADDM) Network
    
    autism spectrum disorders (ASD) identified with Diagnostic and Statistical Manual of Mental Disorders criteria (DSM-IV-TR or DSM-5)
    
    prevalence of ASD 1.68% in children aged 8 years in 2014 (16.8 children per 1,000 children)
    
    reported characteristics of children with ASD in 2014
    estimated 4 times more boys aged 8 years with ASD reported than girls aged 8 years
    
    median age of earliest known ASD diagnosis 52 months
    
    42% had comprehensive developmental evaluation by age 3 years
    
    Reference - MMWR Surveill Summ 2018 Apr 27;67(6):1
  - STUDY SUMMARY
    prevalence of parent-reported autism spectrum disorders in children in United States increased from 1.16% in 2007-2008 to 2% in 2011-2012, and increase might be restricted to boys
    CROSS-SECTIONAL STUDY: Natl Health Stat Report 2013 Mar 20;(65):1
    based on cross-sectional analysis of data collected in 2007-2008 and 2011-2012
    
    91,642 random households in United States with children (≤ 17 years old) were telephoned in 2007-2008 and 95,677 were telephoned in 2011-2012 and asked about autism spectrum disorders (ASD) status of child as part of Centers for Disease Control and Prevention National Survey of Children’s Health (NSCH)
    
    parent-reported ASD defined as parent/guardian reporting that healthcare provider informed them that child had "autism, Asperger disorder, pervasive developmental disorder, or other autism spectrum disorder"
    
    prevalence of ASD in children 0-17 years in 2007-2008 vs. 2011-2012
    in all children, 1.16% vs. 2% (reported as "significant" but no p value reported)
    
    in boys, 1.8% vs. 3.23% (reported as "significant" but no p value reported)
    
    in girls, 0.49% vs. 0.7% (reported as "not significant")
    
    authors suggest that 0.61% prevalence in 2011-2012 among children aged 6-17 years who were diagnosed after first survey indicates that increased overall prevalence may be due to new diagnoses
    
    Reference - Natl Health Stat Report 2013 Mar 20;(65):1PDF
In Canada, reported increases in the prevalence of ASD reported in 2008 and 2010 among children aged 2-14 years in Newfoundland, Labrador, Prince Edward Island, and Southeastern Ontario may be due to a variety of factors, including:
- Internal measurement effects; changes in how prevalence is measured
- External population identification effects; changes in awareness, diagnostic criteria, access to diagnostic or screening instruments, and special education placement resulting in
  - Earlier age of diagnosis
  - Inaccurate diagnosis given to children with other developmental disorders
  - Increased diagnosis among patients with high cognitive functioning
  - Increased awareness leading to "catch-up" diagnoses of children previously not considered for autism spectrum disorders
- Increases in genetic susceptibility to ASD and/or exposure to environmental factors
- Reference - J Autism Dev Disord 2014 Jan;44(1):120

Risk Factors

Family History

About 74%-93% of autism spectrum disorders (ASD) risk is heritable (Lancet 2018 Aug 11;392(10146):508).
Sibling studies report that ASD occurs in 7%-20% of subsequent children after the older child is diagnosed with ASD, and the prevalence is increased in children with 2 older siblings with ASD (Lancet 2018 Aug 11;392(10146):508).
- STUDY SUMMARY
  risk of autism spectrum disorders increased in siblings of affected children with greatest risk in monozygotic twins
  COHORT STUDY: JAMA 2014 May 7;311(17):1770
  based on retrospective cohort study
  
  2,049,973 children born from 1982 to 2006 in Sweden with follow-up through 2009 were analyzed
  
  14,516 children (0.7%) were diagnosed with autism spectrum disorders (ASD); 5,689 children had autistic disorder
  
  familial risk of autism was assessed by adjusted relative recurrence risk (RRR) (defined as relative risk in child with sibling or cousin with diagnosis compared to risk in child with no diagnosed family member)
  
  recurrence risk of ASD at age 20 years increased in
  monozygotic twins (RRR 153, 95% CI 56.7-412.8)
  
  dizygotic twins (RRR 8.2, 95% CI 3.7-18.1)
  
  full siblings (RRR 10.3, 95% CI 9.4-11.3)
  
  maternal half siblings (RRR 3.3, 95% CI 2.6-4.2)
  
  paternal half siblings (RRR 2.9, 95% CI 2.2-3.7)
  
  cousins (RRR 2, 95% CI 1.8-2.2)
  
  findings also significant for autistic disorder
  
  Reference - JAMA 2014 May 7;311(17):1770, editorial can be found in JAMA 2014 May 7;311(17):1738, commentary can be found in JAMA 2014 Sep 17;312(11):1154
- STUDY SUMMARY
  male sex and presence of > 1 older, affected sibling each associated with increased risk of autism spectrum disorders
  COHORT STUDY: Pediatrics 2011 Sep;128(3):e488
  based on prospective cohort study
  
  664 infants at maximum age of 18 months with an older biological sibling with autism spectrum disorders (ASD) followed to age 36 months, then classified as having or not having ASD
  
  ASD developed in 18.7% (103 males, 29 females)
  
  factors associated with increased risk of ASD
  male sex (relative risk [RR] 2.8, 95% CI 1.9-4)
  
  > 1 older sibling with ASD (RR 2.2, 95% CI 1.4-3.3)
  
  Reference - Pediatrics 2011 Sep;128(3):e488, commentary can be found in Evid Based Ment Health 2012 May;15(2):31
Family history of autoimmune disease is associated with increased ASD risk.
- STUDY SUMMARY
  family history of autoimmune diseases associated with increased risk of autism spectrum disorders in children
  SYSTEMATIC REVIEW: Behav Brain Res 2016 Jan 1;296:61SYSTEMATIC REVIEW: Neurosci Biobehav Rev 2015 Aug;55:322
  based on 2 systematic reviews of observational studies
  
  systematic review of 10 studies (9 case-control and 1 cohort) evaluating for children with autism spectrum disorders (ASD) in 9,775 mothers with autoimmune disease and 952,211 control mothers
  diagnosis of ASD by ICD (5 studies), Diagnostic and Statistical Manual of Mental Disorders, fourth edition [DSM-IV] (2 studies), autism diagnostic interview-revised/autism diagnostic observation schedule (1 study), questionnaire (1 study), and East Tennessee Autism Society of America data (1 study)
  
  0.7-2.64% lifetime prevalence of ASD reported in children
  
  maternal autoimmune diseases associated with increased ASD in offspring (pooled odds ratio 1.34, 95% CI 1.23-1.46)
  
  Reference - Behav Brain Res 2016 Jan 1;296:61
  
  systematic review of 11 studies (3 cohort studies, 6 case-control studies, and 2 cross-sectional studies) evaluating children with autism ASD in families with family history of autoimmune diseases
  family history associated with increased ASD in children
  all autoimmune disease (odds ratio [OR] 1.28, 95% CI 1.12-1.48) in analysis of 8 studies
  
  hypothyroidism (OR 1.64, 95% CI 1.07-2.5) in analysis of 6 studies
  
  type 1 diabetes (OR 1.49, 95% CI 1.23-1.81) in analysis of 6 studies
  
  rheumatoid arthritis (OR 1.51, 95% CI 1.19-1.91) in analysis of 4 studies
  
  psoriasis (OR 1.59, 95% CI 1.28-1.97) in analysis of 3 studies
  
  Reference - Neurosci Biobehav Rev 2015 Aug;55:322
  
  DynaMed Commentary
  Out of the 11 studies included in this systematic review, 4 studies were included in the systematic review above, and 7 studies were not discussed.

Genetic Risk Factors

Genetic risk for ASD is most likely polygenic involving multiple single nucleotide polymorphisms (SNPs), each of minor effect. Breakpoints consistently associated with autism spectrum disorders (ASD) include SHANK3 deletion, 1q21, 3q29, 7q11.23, 15q11.2-13.1, 15q12, 15q13, 16p11, 17q12, 22q11.2 and Xq (J Psychopharmacol 2018 Jan;32(1):3).
De novo suspected single gene loss of function mutations and copy number variants (CNVs) such as microdeletions or microduplications spanning multiple genes have been reported to increase risk of ASD and intellectual disability (J Psychopharmacol 2018 Jan;32(1):3).
- In a meta-analysis of 3,613 cases of ASD, the prevalence of 16p11.2 deletion was 0.50% (95% CI, 0.31%-0.82%) and the prevalence of duplication was 0.25% (96% CI, 0.14%-0.56%) (Transl Psychiatry 2019 Jan 16;9(1):8).

Advanced Parental Age

Advanced maternal age (≥ 40 years) and paternal age (≥ 50 years) are reported to be independently associated with autism spectrum disorders (ASD) risk in offspring (Lancet 2018 Aug 11;392(10146):508).
STUDY SUMMARY
age of either parent ≥ 35 years old at time of birth associated with increased risk of autism spectrum disorders in children in Denmark
COHORT STUDY: Ann Epidemiol 2012 Mar;22(3):143
based on retrospective cohort study

1,311,736 children born (as singletons) between 1980 and 2003 in Denmark (Danish Medical Birth Registry) were assessed via medical records for autism spectrum disorders (ASD) and demographic factors including parental age

ASD in 0.74%

parental ages ≥ 35 years compared to < 35 years associated with increased risk of ASD in multivariate analysis adjusted for gestational age, birth weight, birth order, gender, and parental psychiatric history at time of birth
for both parents aged 35-39 years, adjusted hazard ratio 1.21 (95% CI 1.1-1.34)

for both parents ≥ 40 years old, adjusted hazard ratio 1.55 (95% CI 1.28-1.89)

Reference - Ann Epidemiol 2012 Mar;22(3):143
STUDY SUMMARY
maternal age > 35 years associated with increased risk of having a child with autism
META-ANALYSIS: J Am Acad Child Adolesc Psychiatry 2012 May;51(5):477
based on meta-analysis of cohort studies

25,687 autism cases and 8,655,576 controls from 16 epidemiological studies comparing mothers age and risk of autism spectrum disorders (ASD)

adjusted relative risk for ASD compared to mothers aged 25-29 years
in mothers ≥ 35 years old 1.52 (95% CI 1.12-1.92)

in mothers < 20 years old 0.76 (95% CI 0.6-0.97)

adjusted for paternal age and other potential confounders

Reference - J Am Acad Child Adolesc Psychiatry 2012 May;51(5):477, editorial can be found in J Am Acad Child Adolesc Psychiatry 2012 May;51(5):461

Obstetric Factors

Interpregnancy intervals shorter than 12-18 months or longer than 60 months may be associated with increased risk of autism spectrum disorders (ASD):
- STUDY SUMMARY
  interpregnancy intervals of < 12 months or ≥ 60 months may be associated with increased risk of autism spectrum disorders
  SYSTEMATIC REVIEW: Pediatrics 2016 May;137(5):e20153482
  based on systematic review of observational studies
  
  systematic review of 12 observational studies (6 cohort studies, 5 cross-sectional studies, and 1 case-control study) evaluating association between interpregnancy interval (IPI) and neurodevelopmental disabilities in 1,334,569 children
  
  7 studies evaluated association between IPI and risk of autism spectrum disorders (ASD)
  
  IPI defined as time between birth of immediate older sibling and conception of younger sibling
  
  risk of ASD was significantly increased with
  IPI < 12 months compared to
  IPI ≥ 36 months (odds ratio [OR] 1.58, 95% CI 1.02-2.45) in analysis of 5 studies with 438,703 children
  
  IPI 24-59 months (OR 1.83, 95% CI 1.33-2.54) in analysis of 4 studies with 448,362 children
  
  IPI 12-23 months compared to IPI 24-59 months (OR 1.32, 95% CI 1.09-1.59) in analysis of 4 studies with 534,135 children
  
  IPI ≥ 60 months compared to IPI 24-59 months (OR 1.37, 95% CI 1.02-1.86) in analysis of 4 studies with 320,528 children
  
  compared to IPI ≥ 36 months, risk of ASD was not significantly increased with
  IPI 12-23 months in analysis of 5 studies with 577,143 children
  
  IPI 24-35 months in analysis of 5 studies with 453,574 children
  
  Reference - Pediatrics 2016 May;137(5):e20153482
- STUDY SUMMARY
  interpregnancy intervals of < 18 months or ≥ 60 months may be associated with increased risk of autism spectrum disorders
  CASE-CONTROL STUDY: Autism Res 2018 Jan;11(1):81
  based on multi-site case-control study
  
  356 children with autism spectrum disorders (ASD), 627 children with non-ASD developmental disorders, and 524 unaffected children were assessed for maternal IPI of < 18 months (short IPI), or ≥ 60 months (long IPI), or 18-59 months (normal IPI)
  
  IPI defined as time from most recent older sibling birth (stillbirth or livebirth) to estimated conception of younger sibling
  
  IPIs associated with increased risk of ASD in second and later-born children
  short IPI < 18 months (adjusted odds ratio [OR] 1.5, 95% CI 1.1-2.2)
  
  long IPI ≥ 60 months (adjusted OR 1.5, 95% CI 0.99-2.4)
  
  shorter IPIs < 12 months and 12-18 months similarly associated with increased ASD in second and later-born children (data not provided)
  
  no significant association between short IPI and other non-ASD developmental disorders
  
  Reference - Autism Res 2018 Jan;11(1):81
- STUDY SUMMARY
  short interpregnancy interval associated with increased risk of autism in second child
  COHORT STUDY: Pediatrics 2011 Feb;127(2):246COHORT STUDY: Epidemiology 2013 Nov;24(6):906
  based on 2 cohort studies
  
  662,730 second-born children of singleton sibling pairs in California, United States were assessed for autism spectrum disorders (ASD) and IPI (time between birth of first child and conception of second)
  shorter interpregnancy intervals associated with increased risk of ASD compared to interpregnancy interval ≥ 36 months
  interpregnancy interval < 12 months (odds ratio [OR] 3.39, 95% CI 3-3.82)
  
  interpregnancy interval 12-23 months (OR 1.86, 95% CI 1.65-2.1)
  
  interpregnancy interval 24-35 months (OR 1.26, 95% CI 1.1-1.45)
  
  Reference - Pediatrics 2011 Feb;127(2):246
  
  223,476 second-born children of singleton sibling pairs in Norway were assessed for ASD and interpregnancy interval
  sibling pairs in which first-born children were diagnosed with ASD were excluded
  
  autistic disorder (according to DSM-IV) in 0.14%, ASD (including Asperger syndrome or pervasive developmental disorder not otherwise specified [PDD-NOS]) in 0.43%
  
  shorter interpregnancy intervals associated with increased risk (relative to interval of 36 months) of developing autistic disorder (but not Asperger syndrome or PDD-NOS)
  < 9 months (adjusted odds ratio [OR] 2.18, 95% CI 1.42-3.26)
  
  9-11 months (adjusted OR 1.71, 95% CI 1.07-2.64)
  
  intervals > 11 months not associated with increased risk
  
  Reference - Epidemiology 2013 Nov;24(6):906
STUDY SUMMARY
meconium aspiration, birth injury or trauma, congenital malformation, anemia in infant, and ABO or Rh incompatibility are among obstetric factors associated with increased risk of developing autism spectrum disorders
SYSTEMATIC REVIEW: Pediatrics 2011 Aug;128(2):344
based on systematic review of observational studies

review of 40 cohort studies assessing perinatal and neonatal risk factors for developing autism spectrum disorders (ASD)

studies excluded if case-series, autism prevalence studies, studies of other psychiatric conditions, or no nonautistic comparison group

factors associated with increased risk of developing ASD
presentation
abnormal presentation in general (risk ratio [RR] 1.44, 95% CI 1.07-1.94) in analysis of 15 studies

breech presentation (RR 1.81, 95% CI 1.21-2.71) in analysis of 4 studies

fetal distress (RR 1.52, 95% CI 1.09-2.12) in analysis of 4 studies

birth injury or trauma (RR 4.9, 95% CI 1.41-16.94) in analysis of 6 studies

multiple births (such as twins) (RR 1.77, 95% CI 1.23-2.55) in analysis of 10 studies

maternal hemorrhage (RR 2.39, 95% CI 1.35-4.21) in analysis of 4 studies

birth during the summer (RR 1.14, 95% CI 1.02-1.26) in analysis of 3 studies

birth size
birth weight < 2,500 g (RR 1.63, 95% CI 1.19-2.33) in analysis of 15 studies

small size for gestational age (RR 1.35, 95% CI 1.13-1.61) in analysis of 10 studies

clinical characteristics of infant
congenital malformation (RR 1.8, 95% CI 1.42-2.82) in analysis of 11 studies

low 5-minute Apgar score (RR 1.67, 95% CI 1.24-2.26) in analysis of 8 studies

general status of infant
feeding difficulties in infant (RR 3.35, 95% CI 1.33-8.4) in analysis of 3 studies

meconium aspiration by infant (RR 7.34, 95% CI 2.3-23.47) in analysis of 3 studies

medical status of infant
anemia (RR 7.87, 95% CI 1.43-43.36) in analysis of 4 studies

ABO or Rh incompatibility (RR 3.7, 95% CI 1.9-7.23) in analysis of 5 studies

hyperbilirubinemia (RR 1.87, 95% CI 1.01-3.47) in analysis of 6 studies

factors not associated with increased ASD risk include
features of labor (such as duration), delivery procedures (such as use of anesthesia or Cesarean procedure), unclear amniotic fluid, birth during nonsummer months, and gestational age

infant conditions/features - head circumference, soft neurological signs (including irritability, floppiness, or convulsions), low 1-minute Apgar score, weakness or lack of crying, respiratory distress, infection or elevated temperature, jaundice, phototherapy, seizures, microcephaly, heart defect, or medical treatment

results may be limited by significant heterogeneity

Reference - Pediatrics 2011 Aug;128(2):344
Other possible obstetric-related risk factors for ASD include Caesarean delivery (though evidence is inconsistent), maternal epidural analgesia, obese maternal prepregnancy weight, and gestational diabetes mellitus diagnosed at ≤ 26 weeks gestational age.
- There is inconsistent evidence for Caesarean delivery as a risk factor for ASD.
  - STUDY SUMMARY
    Caesarean section delivery may be associated with autism spectrum disorders in children
    SYSTEMATIC REVIEW: J Child Psychol Psychiatry 2015 May;56(5):500
    based on systematic review of observational studies
    
    systematic review of 25 studies (17 case-control and 8 cohort studies) evaluating 1,192,806 children and young adults for association between autism spectrum disorders (ASD) or attention-deficit/hyperactivity disorder (ADHD) and birth by Caesarean section delivery compared to vaginal delivery
    12,897 children had ASD, defined as diagnosis of autistic disorder, Asperger disorder, or pervasive developmental disorder-not otherwise specified
    
    13,115 children and young adults had ADHD, defined as diagnosis of ADHD, attention-deficit disorder, or hyper kinetic disorder
    
    compared to vaginal delivery, Caesarean section delivery associated with increased odds of ASD (adjusted odds ratio 1.23, 95% CI 1.07-1.4) in analysis of 13 studies, with data limited by heterogeneity
    
    Caesarean section delivery not significantly associated with increased odds of ADHD in analysis of 2 studies
    
    Reference - J Child Psychol Psychiatry 2015 May;56(5):500
  - STUDY SUMMARY
    modes of delivery (vaginal or Caesarean) do not appear to be associated with autism spectrum disorders or attention-deficit/hyperactivity disorder in children
    COHORT STUDY: J Autism Dev Disord 2016 Feb;46(2):603
    based on cohort study
    
    13,141 singleton children in national United Kingdom cohort were assessed for mode of delivery in birth and association with diagnoses of autism spectrum disorders (ASD) or attention-deficit/hyperactivity disorder (ADHD) by age 7 years
    
    birth modes of delivery were spontaneous vaginal delivery (VD), assisted VD, and induced VD, or emergency Caesarean section (CS), planned CS, and induced CS
    
    209 cases of ASD (1.5%) and 173 cases of ADHD (1.3%)
    
    no significant associations between various modes of vaginal or Caesarean delivery in birth and either ASD or ADHD in children
    
    Reference - J Autism Dev Disord 2016 Feb;46(2):603
  - STUDY SUMMARY
    Caesarean delivery does not appear to be associated with autism spectrum disorders in children
    COHORT STUDY: JAMA Psychiatry 2015 Sep;72(9):935
    based on cohort study
    
    2,697,315 singleton live births in Sweden (1989-2010) from national registry data were assessed for obstetrical mode of delivery and diagnosis of autism spectrum disorders (ASD)
    
    modes of delivery were unassisted VD, assisted VD, elective CS, and emergency CS
    
    28,290 children (1.0%) had ASD with median age at diagnosis of 12 years
    
    adjusted analysis used nonaffected siblings as matched controls
    
    compared with unassisted VD, with sibling control analysis CS not significantly associated with ASD
    elective CS (adjusted odds ratio [OR] 0.89, 95% CI 0.76-1.04)
    
    emergency CS (adjusted OR 0.97, 95% CI, 0.85-1.11)
    
    Reference - JAMA Psychiatry 2015 Sep;72(9):935
- STUDY SUMMARY
  maternal epidural analgesia during labor associated with autism spectrum disorders in children
  COHORT STUDY: JAMA Pediatr 2020 Dec 1;174(12):1168
  Family_Medicine Obstetric_and_Gynecologic_Conditions Pediatricsmaternal epidural analgesia during labor associated with autism spectrum disorder in children (JAMA Pediatr 2020 Oct 12 early online)11/02/2020 08:37:39 AM
  based on retrospective cohort study
  
  medical records for 147,895 children who were delivered vaginally in singleton pregnancies (mean gestational age 38.9 weeks, 50.3% boys) at Kaiser Permanente hospitals in California, United States from 2005 to 2012 were reviewed
  
  74.2% were exposed to maternal epidural analgesia during labor
  exposure for < 4 hours in 29.6%
  
  exposure for 4-8 hours in 45.8%
  
  exposure for > 8 hours in 24.6%
  
  incidence of autism spectrum disorders (ASD)
  1.9% in children exposed to epidural analgesia
  
  1.3% in children not exposed to epidural analgesia
  
  analyses adjusted for birth year, medical center, maternal age, parity, race/ethnicity, educational level, household income, history of comorbidity, diabetes during pregnancy, smoking during pregnancy, preeclampsia or eclampsia, prepregnancy body mass index, gestational weight gain, gestational age at delivery, and birth weight
  
  compared to no exposure, increased risk of ASD associated with epidural analgesia exposure
  overall (adjusted hazard ratio [HR] 1.37, 95% CI 1.23-1.53)
  
  for < 4 hours of exposure (adjusted HR 1.33, 95% CI 1.17-1.53)
  
  for 4-8 hours of exposure (adjusted HR 1.35, 95% CI 1.2-1.53)
  
  for > 8 hours of exposure (adjusted HR 1.46, 95% CI 1.27-1.69)
  
  increased duration of exposure associated with increasing risk of ASD (adjusted HR 1.05, 95% CI 1.01-1.09 per 4 hours of exposure)
  
  Reference - JAMA Pediatr 2020 Dec 1;174(12):1168
  
  DynaMed Commentary
  Although these data show an association, they do not demonstrate a causal relationship between epidural anesthesia in labor and ASD. As discussed by authors, the possible effects of unmeasured confounders including both pre and postpartum factors like environmental exposures, paternal factors, genetic predisposition, and infections, may play a role in the observed association.
- STUDY SUMMARY
  obese maternal prepregnancy weight associated with autism-associated behaviors in offspring
  COHORT STUDY: Autism Res 2019 Jan;12(1):80
  based on cohort study
  
  1,238 children aged 19-20 years evaluated for autistic-like traits and assessed for maternal height, weight, and body mass index (BMI) data collected during second- and third-trimester of pregnancy (16-20 weeks)
  
  Autism-Spectrum Quotient (AQ) self-reported questionnaire and AQ Total Score with threshold of 26 used to identify individuals with high levels of autistic-like traits
  
  median AQ Total score 15 (range 1-47)
  
  5% of mothers had prepregnancy BMI ≥ 30 (obese category)
  
  maternal prepregnancy obesity associated with increased risk of high levels of autism-associated behaviors in children (adjusted odds ratio 2.8, 95% CI 1.06-7.43)
  
  no significant association between autism-like behaviors and paternal BMI category, hypertensive diseases of pregnancy, maternal cigarette smoking during pregnancy, maternal alcohol consumption during pregnancy, parity, maternal education, maternal age at conception, threatened abortion, or maternal diabetes
  
  Reference - Autism Res 2019 Jan;12(1):80
- STUDY SUMMARY
  gestational diabetes mellitus diagnosed at ≤ 26 weeks gestational age but not later associated with increased risk of autism spectrum disorders in children
  COHORT STUDY: JAMA 2015 Apr 14;313(14):1425
  based on retrospective cohort study
  
  322,323 singleton children born at gestational age 28-44 weeks between 1995 and 2009 in southern California, United States were assessed
  children born to mothers with type 1 diabetes and children with congenital anomalies were excluded
  
  2.3% were born to mothers with gestational diabetes mellitus (GDM) diagnosed at ≤ 26 weeks gestational age, 5.5% to mothers with later GDM diagnosis, and 2% to mothers with preexisting type 2 diabetes
  
  1.1% were diagnosed with autism spectrum disorders (ASD) during median 5.5-year follow-up
  
  increased risk of ASD independently associated with GDM diagnosed at ≤ 26 weeks gestational age (adjusted hazard ratio 1.42, 95% CI 1.15-1.74)
  
  no significant association between ASD and GDM diagnosed at > 26 weeks gestational age, preexisting type 2 diabetes, or use of antidiabetic medication
  
  Reference - JAMA 2015 Apr 14;313(14):1425, commentary can be found in Evid Based Ment Health 2015 Nov;18(4):113
- STUDY SUMMARY
  individuals with diabetes, hypertension, or obesity may have increased risk of having child with autism or other neurodevelopmental disorders
  CASE-CONTROL STUDY: Pediatrics 2012 May;129(5):e1121
  based on case-control study
  
  689 children aged 2-5 years with autism spectrum disorders and developmental delays, and 315 matched controls were assessed for risk associated with maternal metabolic conditions (diabetes, hypertension, and obesity)
  
  maternal metabolic conditions associated with increased risk (compared to controls) of
  autism spectrum disorders (odds ratio 1.61, 95% CI 1.1-2.37)
  
  developmental delays (odds ratio 2.35, 95% CI 1.43-3.88)
  
  expressive language scores lower in children of individuals with metabolic conditions compared to children of individuals without metabolic conditions (p = 0.01)
  
  Reference - Pediatrics 2012 May;129(5):e1121

Gestational Age and Weight at Birth

Lower gestational age and very low birth weight may be associated with increased risk of autism spectrum disorders (ASD). However, prematurity-related prenatal and neonatal complications and maternal factors may be attenuating factors.
- STUDY SUMMARY
  lower gestational age at birth may be associated with increased risk of autism spectrum disorders
  COHORT STUDY: J Pediatr 2013 Feb;162(2):361
  based on retrospective cohort study
  
  218,110 singleton live births (51% male) in Alberta, Canada from 1998 to 2004 were analyzed
  
  autism spectrum disorders (ASD) in 0.85% of male persons and 0.17% of female persons overall
  Prevalence of Autism Spectrum Disorders by Gestational Age
  Prevalence of Autism Spectrum Disorders by Gestational Age
  Gestational Age Prevalence
  
  Male Persons (%) Female Persons (%)
  
  26 weeks 2.6 0
  
  30 weeks 2 0.43
  
  34 weeks 1 0.57
  
  40 weeks 0.8 0.2
  
  comparing lower to higher gestational age, increased risk of ASD associated with (data estimated from figure)
  gestational age < 30 weeks (risk ratio 2.7, 95% CI 1.6-4.5)
  
  gestational age < 36 weeks (risk ratio 1.5, 95% CI 1.2-1.7)
  
  Reference - J Pediatr 2013 Feb;162(2):361
  
  DynaMed Commentary
  Authors did not control for confounding variables such as maternal age, birth weight, or perinatal complications.
- STUDY SUMMARY
  very low birth weight and preterm birth associated with increased risk of childhood autism spectrum disorders
  CASE-CONTROL STUDY: J Pediatr 2012 Nov;161(5):830
  based on case-control study
  
  1,132 children born between 1987 and 2005 with diagnoses of childhood autism spectrum disorders (ASD) and matched to general population controls (4 controls to 1 affected child) were assessed for gestational age and weight at birth using nationwide registry data in Finland
  
  all diagnoses according to ICD-9 and ICD-10 criteria
  
  prevalence of childhood ASD among children age ≤ 19 years was 9 per 10,000 by year 2005
  
  perinatal characteristics associated with increased risk of childhood ASD
  very low (< 1500 g) birth weight (adjusted odds ratio [OR] 3.1, 95% CI 1.4-6.5)
  
  low (< 2500 g) birth weight (adjusted OR 1.6, 95% CI 1.05-2.3)
  
  very low gestational age (< 32 weeks) (adjusted OR 2.5, 95% CI 1.3-5)
  
  small for gestational age status (adjusted OR 1.7, 95% CI 1.1-2.6)
  
  Reference - J Pediatr 2012 Nov;161(5):830
- STUDY SUMMARY
  small-for-gestational age birth weight appears associated with increased risk of autism in preterm infants, while large-for-gestational age birth weight associated with decreased risk of autism in preterm infants and increased risk in term infants
  COHORT STUDY: Am J Obstet Gynecol 2012 Apr;206(4):314.e1
  based on retrospective population-based cohort study
  
  21,717 children with autism spectrum disorders (ASD) compared to 5,957,888 control children from a birth cohort of California were evaluated according to birth weight percentile into 1 of 3 groups
  small-for-gestational age (SGA), < fifth percentile or fifth to 10th percentile
  
  appropriate size for gestational age (GA), > 10th percentile to < 90th percentile
  
  large-for-gestational age (LGA), either 90th-95th percentile or > 95th percentile
  
  compared to nonaffected preterm infants
  preterm SGA (fifth percentile) infants associated with increased risk of ASD
  23-31 weeks (adjusted odds ratio [OR] 1.60, 95% CI 1.09-2.35)
  
  32-33 weeks (adjusted OR 1.83, 95% CI 1.16-2.87)
  
  term LGA (> 95th percentile) infants 39-41 weeks associated with increased risk of ASD (adjusted OR 1.16, 95% CI 1.08-1.26)
  
  preterm LGA infants 23-31 weeks associated with decreased risk of ASD (adjusted OR, 0.45; 95% CI, 0.21-0.95).
  
  Reference - Am J Obstet Gynecol 2012 Apr;206(4):314.e1
- STUDY SUMMARY
  lower gestational age at birth not associated with increased risk of ASD after adjusting for prematurity-related prenatal and neonatal complications and maternal factors
  CASE-CONTROL STUDY: Pediatrics 2009 Nov;124(5):e817
  based on case-control study
  
  1,216 children with ASD diagnosed at ≤ 10 years old and born in Sweden between 1987 and 2002 were compared with 6,080 gender- and birth year-matched controls
  
  gestational age ≤ 36 weeks compared to 37-41 weeks not associated with increased risk for autism spectrum disorders (ASD) in multivariate analysis adjusted for
  birth characteristics (such as small size for gestational age, congenital malformations, and being a twin)
  
  numerous neonatal complications (including but not limited to low Apgar score, intracranial bleeding, and seizures)
  
  numerous maternal characteristics (including but not limited to age, smoking status, and presence of diabetes)
  
  1,188 observations not included in adjusted analysis due to missing data for some variables such as birth characteristics, neonatal complications, or maternal characteristics
  
  Reference - Pediatrics 2009 Nov;124(5):e817

Prevalence of Autism Spectrum Disorders by Gestational Age
Gestational Age	Prevalence
26 weeks	2.6	0
30 weeks	2	0.43
34 weeks	1	0.57
40 weeks	0.8	0.2

Prenatal and Infant Medication Exposures

Prenatal exposure to antidepressants, valproate, and beta-adrenergic agonists are each associated with increased risk of autism spectrum disorders (ASD).
Antidepressant use during pregnancy is associated with increased risk of autism spectrum disorders.
- STUDY SUMMARY
  selective serotonin reuptake inhibitor (SSRI) exposure during pregnancy may be associated with increased risk of autism spectrum disorders compared to no SSRI exposure
  SYSTEMATIC REVIEW: Reprod Toxicol 2016 Jul 26;65:170
  based on systematic review
  
  systematic review of 11 cohort and case-control studies evaluating SSRI exposure during pregnancy on risk of autism spectrum disorders (ASD)
  
  8 studies included in meta-analysis
  
  compared to no SSRI exposure during pregnancy, SSRI exposure during pregnancy associated with increased risk of ASD (odds ratio 1.45, 95% CI 1.15-1.82) in analysis of 7 studies
  
  no significant increased risk of ASD for SSRI exposure compared to other antidepressants in subgroup analysis of 3 studies
  
  among women with psychiatric disorder, no significant difference between risk of ASD comparing SSRI exposure during pregnancy vs. no exposure in subgroup analysis of 3 studies
  
  Reference - Reprod Toxicol 2016 Jul 26;65:170
- STUDY SUMMARY
  maternal use of antidepressants before, during, or after pregnancy associated with increased risk of autism spectrum disorders in offspring, but no significant association between first trimester use and risk of autism spectrum disorders in analysis of exposure-discordant siblings
  COHORT STUDY: JAMA 2017 Apr 18;317(15):1553
  based on retrospective population-based cohort study
  
  1,580,629 singleton live births in Sweden between 1996 and 2012 were evaluated according to maternal antidepressant use during pregnancy
  
  association between maternal antidepressant use and risk of autism spectrum disorders (ASD) was assessed based on
  maternal self-report of first trimester antidepressant use (whole sample)
  
  timing of maternal antidepressant use from medication dispensation records (2006-2012)
  
  association between maternal first trimester antidepressant use and risk of ASD was also evaluated in covariate-adjusted analysis comparing exposure-discordant siblings within families
  
  in multivariable analysis using maternal self-report of first trimester antidepressant use, maternal use of any antidepressant during first trimester associated with increased risk of ASD (adjusted hazard ratio [HR] 1.64, 95% CI 1.46-1.83)
  
  in cohort stratified by timing of maternal antidepressant use, increased risk of ASD associated with maternal antidepressant use
  before pregnancy (adjusted HR 1.4, 95% CI 1.02-1.93)
  
  during first trimester (adjusted HR 1.75, 95% CI 1.49-2.07)
  
  during second and/or third trimester (adjusted HR 2.18, 95% CI 0.98-4.85)
  
  after pregnancy (adjusted HR 1.45, 95% CI 1.03-2.04)
  
  in analysis of exposure-discordant siblings, no significant association between maternal self-report of first trimester antidepressant use and risk of ASD (adjusted HR 0.83, 95% CI 0.62-1.13)
  
  Reference - JAMA 2017 Apr 18;317(15):1553, editorial can be found in JAMA 2017 Apr 18;317(15):1533
- STUDY SUMMARY
  maternal use of antidepressants during pregnancy associated with increased risk of autism spectrum disorders in children compared to no antidepressant use in mothers with psychiatric disorders
  COHORT STUDY: BMJ 2017 Jul 19;358:j2811
  based on prospective population-based cohort study
  
  254,610 children aged 4-17 years living in Sweden in 2001-2011 were evaluated for autism spectrum disorders (ASD) according to maternal antidepressant use during pregnancy and presence of maternal psychiatric disorders
  3,342 had in utero exposure to antidepressants (81% had exposure to selective serotonin reuptake inhibitors)
  
  12,325 had no in utero exposure to antidepressants and had mother with psychiatric disorder
  
  comparing in utero exposure to antidepressants vs. no in utero exposure and mother with psychiatric disorder
  ASD 4.1% vs. 2.9% (adjusted odds ratio 1.45, 95% CI 1.13-1.85)
  
  ASD without intellectual disability 3.7% vs. 2.4% (adjusted odds ratio 1.57, 95% CI 1.21-2.04)
  
  no significant difference in risk of ASD with intellectual disability
  
  consistent results in propensity score-matched analyses including up to 6,426 children
  
  Reference - BMJ 2017 Jul 19;358:j2811, editorial can be found in BMJ 2017 Jul 19;358:j3388
- STUDY SUMMARY
  serotonergic antidepressant use during pregnancy may be associated with increased risk of autism spectrum disorders
  COHORT STUDY: JAMA 2017 Apr 18;317(15):1544
  based on retrospective cohort study
  
  35,906 singleton births in Canada between 2002 and 2010 were evaluated for autism spectrum disorders (ASD) according to maternal serotonergic antidepressant use during pregnancy
  
  serotonergic antidepressant exposure defined as ≥ 2 consecutive prescriptions for SSRIs or selective norepinephrine reuptake inhibitor filled between conception and delivery
  
  2,837 (7.9%) pregnancies were exposed to serotonergic antidepressants
  
  1.1% of children were diagnosed with autism spectrum disorders over mean follow-up 4.95 years
  
  incidence of ASD comparing exposed vs. unexposed pregnancies in
  overall analysis 4.51 per 1,000 person-years vs. 2.03 per 1,000 person-years (adjusted hazard ratio [HR] 1.59, 95% CI 1.17-2.17)
  
  mood or anxiety disorder-restricted cohort 4.75 per 1,000 person-years vs. 2.43 per 1,000 person years (adjusted HR 1.6, 95% CI 1.16-2.22)
  
  in additional analysis weighted by high-dimensional propensity score for likelihood of serotonergic antidepressant treatment
  serotonergic antidepressant exposure during pregnancy associated with
  nonsignificant increase in risk of ASD overall (HR 1.61, 95% CI 0.997-2.59)
  
  significant increase in risk of ASD in mood or anxiety disorder-restricted cohort (HR 1.62, 95% CI 1.04-2.52)
  
  for risk of ASD by trimester of exposure
  no significant increase in risk with first trimester exposure (HR 1.24, 95% CI 0.7-2.2)
  
  risk of ASD nonsignificantly increased with second or third trimester exposure (HR 1.64, 95% CI 0.99-2.71)
  
  consistent results in analysis limited to SSRI exposure during pregnancy
  
  Reference - JAMA 2017 Apr 18;317(15):1544
- STUDY SUMMARY
  maternal use of antidepressants, particularly SSRIs or ≥ 2 different antidepressant classes, during the second and/or third trimester appears associated with increased risk of autism spectrum disorders
  COHORT STUDY: JAMA Pediatr 2016 Feb;170(2):117
  based on population-based cohort study
  
  145,456 singleton pregnancies and children from the Quebec Pregnancy/Children Cohort between 1998 and 2009 were included
  
  antidepressant exposure defined as having ≥ 1 prescription filled at any time during pregnancy or a prescription filled before pregnancy that overlapped with first day of gestation
  
  mean age of children at the end of follow-up 6.2 years
  
  ≥ 1 diagnosis of autism spectrum disorders (ASD) in 1,054 children (0.72%)
  
  exposure to antidepressant in utero in 4,724 infants (3.2%); exposure during
  first trimester in 4,200 (88.9%)
  
  second and/or third trimester in 2,532 (53.6%)
  
  compared to no exposure, increased risk of ASD associated with second and/or third trimester exposure to
  any antidepressants (adjusted hazard ratio [HR] 1.87, 95% CI 1.15-3.04)
  
  only SSRIs (adjusted HR 2.17, 95% CI 1.2-3.93)
  
  ≥ 2 antidepressant classes (adjusted HR 4.39, 95% CI 1.44-13.32)
  
  Reference - JAMA Pediatr 2016 Feb;170(2):117
- STUDY SUMMARY
  maternal antidepressant use for depression during pregnancy may be associated with increased risk of autism spectrum disorders without intellectual disability in offspring
  CASE-CONTROL STUDY: BMJ 2013 Apr 19;346:f2059
  based on case-control study
  
  4,429 children aged 0-17 years with autism spectrum disorders (ASD) (58.7% without intellectual disability) were compared to 43,277 matched controls without autism spectrum disorders
  
  1,679 children with ASD were compared to 16,845 of matched controls with available data on maternal antidepressant use (including SSRIs or non-SSRIs) during pregnancy
  maternal antidepressant use for depression associated with increased risk of ASD without intellectual disability (adjusted odds ratio [OR] 4.94, 95% CI 1.85-13.23)
  
  maternal antidepressant use for depression not associated with increased risk of ASD with intellectual disability (adjusted OR 1.81, 95% CI 0.39-8.56)
  
  in overall analysis, 1% of mothers and 0.4% of fathers of children with ASD had diagnosis of depressive disorder prior to birth of child
  maternal depression associated with increased risk of ASD without intellectual disability (adjusted OR 1.86, 95% CI 1.25-2.77)
  
  maternal depression not associated with increased risk of ASD with intellectual disability (adjusted OR 0.96, 95% CI 0.52-1.77)
  
  paternal depression not associated with risk of ASD with or without intellectual disability
  
  Reference - BMJ 2013 Apr 19;346:f2059

STUDY SUMMARY
in utero antibiotic exposure may be associated with slightly increased risks of autism spectrum disorder and language disorder in children in South Korea
COHORT STUDY: BMJ 2024 May 22;385:e076885
based on retrospective population-based cohort study

3,963,874 children born between 2009 and 2020 from national health insurance database in South Korea were assessed for exposure to antibiotic in utero or during first 6 months of life and followed until 2021

antibiotic exposure defined as record of prescription for antibiotic

propensity score for likelihood of receiving antibiotics was calculated for each child based on demographic and clinical factors
980,872 children who had in utero antibiotic exposure were propensity-matched to 980,872 children without in utero exposure to antibiotic

804,887 children who had exposure to antibiotic during first 6 months of life were propensity-matched to 804,887 children who did not have exposure to antibiotic during first 6 months of life

331,314 children had in utero exposure plus exposure during first 6 months of life

sibling analyses were performed in children who had > 1 sibling during study period to control for shared familial factors

median follow-up about 7 years

incidence of outcomes per 1,000 person-years comparing in utero antibiotic exposure vs. no in utero antibiotic exposure
autism spectrum disorder 1.25 vs. 1.1 (adjusted hazard ratio [HR] 1.14, 95% CI 1.11-1.17)

intellectual disorder 0.62 vs. 0.53 (adjusted HR 1.17, 95% CI 1.12-1.22)

language disorder 2.59 vs. 2.38 (adjusted HR 1.09, 95% CI 1.07-1.11)

epilepsy 1.12 vs. 1.03 (adjusted HR 1.08, 95% CI 1.05-1.12)

consistent results for all outcomes for exposure by trimester

in sibling cohort analysis, consistent results for exposure at any time during gestation for
autism spectrum disorder (adjusted HR 1.06, 95% CI 1.01-1.12)

language disorder (adjusted HR 1.05, 95% CI 1.02-1.09)

incidence of outcomes per 1,000 person-years comparing antibiotic exposure during first 6 months of life vs. no antibiotic exposure during first 6 months of life
autism spectrum disorder 1.25 vs. 1.2 (adjusted HR 1.04, 95% CI 1.01-1.08)

intellectual disorder 0.68 vs. 0.55 (adjusted HR 1.22, 95% CI 1.17-1.28)

language disorder 2.71 vs. 2.58 (adjusted HR 1.06, 95% CI 1.04-1.08)

epilepsy 1.01 vs. 0.8 (adjusted HR 1.27, 95% CI 1.23-1.32)

consistent results for
autism spectrum disorder for exposure at < 2 months of life

language disorder for exposure at < 2 months of life and 2-4 months of life

intellectual disorder and epilepsy for exposure at < 2 months of life, 2-4 months of life, and 4-6 months of life

in sibling cohort analysis, consistent results for exposure at any time up to 6 months of life for epilepsy (adjusted HR 1.13, 95% CI 1.09-1.18)

Reference - BMJ 2024 May 22;385:e076885
STUDY SUMMARY
prenatal valproate exposure associated with increased risk of autism spectrum disorders and autistic disorder in offspring
COHORT STUDY: JAMA 2013 Apr 24;309(16):1696
based on retrospective cohort study

655,615 children born between 1996 and 2006 were followed to mean age 8.8 years
autism spectrum disorders (ASD) (including autistic disorder, Asperger syndrome, atypical autism, and other or unspecified pervasive developmental disorders) in 5,437 children (0.83%)

autistic disorder in 2,067 children (0.32%)

prenatal exposure to valproate defined as ≥ 1 prescription from 1 month prior to conception until birth in 508 children (0.08%)

compared to no exposure, prenatal valproate exposure associated with increased risk of
ASD (adjusted hazard ratio [HR] 2.9, 95% CI 1.7-4.9)

autistic disorder (adjusted HR 5.2, 95% CI 2.7-10)

in analysis of 6,584 children born to women with epilepsy
prenatal valproate exposure not associated with increased risk of ASD compared to no exposure (adjusted HR 1.7, 95% CI 0.9-3.2)

prenatal valproate exposure associated with increased risk of childhood autism compared to no exposure (adjusted HR 2.9, 95% CI 1.4-6)

Reference - JAMA 2013 Apr 24;309(16):1696, editorial can be found in JAMA 2013 Apr 24;309(16):1730, commentary can be found in Arch Dis Child Educ Pract Ed 2014 Oct;99(5):198

STUDY SUMMARY
prenatal beta-adrenergic agonists exposure associated with increased risk of autism spectrum disorders
CASE-CONTROL STUDY: Pediatrics 2016 Feb;137(2):e20151316
based on population-based case-control study

5,200 children with autism spectrum disorders (ASD) who were born between 1997 and 2006 in Denmark were matched to 52,000 controls without ASD

included beta-adrenergic agonist drugs not reported

3.7% of children with ASD and 2.9% of controls were exposed to beta-adrenergic agonists in utero

1.6% of children with ASD and 1.3% of controls had mothers with diagnosis of asthma during pregnancy

in utero exposure to beta-adrenergic agonists associated with increased likelihood of developing ASD (adjusted odds ratio 1.3, 95% CI 1.1-1.5)

consistent results in analysis stratified by trimester of pregnancy

Reference - Pediatrics 2016 Feb;137(2):e20151316, editorial can be found in Pediatrics 2016 Feb;137(2):e20152703

Perinatal Environmental Exposures

Perinatal exposure to endocrine-disrupting chemicals (EDCs):
- There is inconsistent and limited evidence on the association between exposure to EDCs, which include environmentally persistent organic pollutants as well as certain nonpersistent chemicals, and risk for autism spectrum disorders (ASD).
- EDCs interfere with critical hormones in neurodevelopment, may interfere with immune system activity, and have been associated with a range of other neurodevelopmental endpoints.
- Exposure to EDCs is prevalent in developed countries.
- Reference - Annu Rev Public Health 2017 Mar 20;38:81
Perinatal exposure to air pollutants:
- Air pollutants are reported to be potentially associated with impaired neurodevelopment, resulting in lower cognitive test outcomes and an increased risk of development disorders such as ASD.
- Pollutants under investigation include ambient air pollution such as traffic-related emissions and household air pollution such as combustion fuel within the house or exposure to carbon monoxide from cooking with gas or solid fuel.
- Reference - World Health Organization (WHO) 2018 Oct 24 Air pollution and child health: prescribing clean air summary PDF, or in Spanish PDF, Arabic PDF, Chinese PDF, French PDF, or Russian PDF
Perinatal exposure to ambient air pollutants, particularly particulate matter, is associated with increased risk of ASD. However, evidence is mixed regarding the timing of exposure and specific pollutants.
- STUDY SUMMARY
  perinatal exposure to increasing amounts of ambient particulate matter < 10 micrometers and < 2.5 micrometers appears associated with increased likelihood of autism spectrum disorders
  SYSTEMATIC REVIEW: PLoS One 2016;11(9):e0161851
  based on systematic review with clinical heterogeneity
  
  systematic review of 23 studies (17 case-control, 4 population-based or school-based studies, and 2 cohort studies) evaluating perinatal exposure to air pollution and autism spectrum disorders (ASD)
  
  air pollutants included meta-analyses were particulate matter with diameter ≤ 10 micrometers (PM10) and with diameter ≤ 2.5 micrometers (PM2.5); most studies assigned exposure based on birth address
  
  clinical heterogeneity included assessment of air pollution exposure (many various, nonstandardized, and surrogate metrics were used, in addition to many classes of compounds), confounders adjusted for, timing of exposure or outcome measurement, and method for measuring autism outcome
  
  pollutants found associated with increased risk of ASD included
  PM10 (odds ratio [OR] 1.2, 95% CI 1-1.42 per 10-microgram/m³ increase) in analysis of 6 case-control studies with 153,437 children; consistent but attenuated effect in clustering analysis to reduce statistical heterogeneity (OR 1.07, 95% CI 1.06-1.08)
  
  PM2.5 (OR 1.88, 95% CI 1.11-3.2 per 10-microgram/m³ increase) in analysis of 3 case-control studies with 85,676 children; consistent results in clustering analysis to reduce statistical heterogeneity (OR 2.32, 95% CI 2.15-2.51)
  
  Reference - PLoS One 2016;11(9):e0161851
- STUDY SUMMARY
  exposure to particulate matter < 2.5 micrometers and ozone during specific time windows during pregnancy may be associated with modestly increased risk of autism, and exposure to those pollutants, particulate matter < 10 micrometers, and nitrogen dioxide after birth appears to possibly increase risk
  SYSTEMATIC REVIEW: Environ Res 2016 Nov;151:763
  based on systematic review with clinical heterogeneity
  
  systematic review of 12 studies (7 cohort and 5 case-control studies) evaluating exposure to ambient air pollution and risk of autism spectrum disorders (ASD)
  
  pollutants evaluated included airborne particulate matter with diameter PM2.5 or PM10, nitrogen dioxide, and ozone
  
  included studies used varying measures of air pollution exposure
  
  PM2.5 per 10 microgram/m³ increases associated with
  no significant difference in risk of ASD with exposure during entire pregnancy (any point) in analysis of 5 studies (includes 1 collaborative study with 4 European cohorts), or second trimester in analysis of 4 studies
  
  increased risk of ASD with exposure
  in first-trimester (risk ratio [RR] 1.1, 95% CI 1.02-1.19) in analysis of 4 studies
  
  in third trimester (RR 1.33, 95% CI 1-1.77) in analysis of 4 studies
  
  after birth (RR 2.43, 95% CI 1.61-3.68) in analysis of 2 studies
  
  PM10 per 10 microgram/m³ increases associated with
  no significant difference in risk of ASD with exposure at any point during pregnancy in analysis of 3 studies (includes 1 collaborative study with 4 European cohorts) or during first trimester, second trimester, or third trimester in analysis of 3 studies
  
  nonsignificantly increased risk of ASD with exposure after birth (RR 1.33, 95% CI 0.86-2.05) in analysis of 2 studies
  
  nitrogen dioxide per 10 parts-per-billion increases associated with
  no significant difference in risk of ASD with exposure at any point during pregnancy in analysis of 3 studies (including 6 European cohorts in 1 study) or during first trimester, second trimester, or third trimester in analysis of 2 studies
  
  increased risk of ASD with exposure after birth (RR 2.72, 95% CI 1.04-7.07) in analysis of 2 studies
  
  ozone per 10 parts-per-billion increases associated with
  slightly increased risk of ASD with exposure at any point during pregnancy (RR 1.05, 95% CI 1.01-1.1), during second trimester (RR 1.02, 95% CI 1-1.04), and during third trimester (RR 1.03, 95% CI 1.02-1.05) in analysis of 2 studies
  
  no significant difference in risk of ASD with exposure during first trimester in analysis of 2 studies
  
  nonsignificantly increased risk of ASD with exposure after birth (RR 1.35, 95% CI 0.94-1.95) in analysis of 2 studies
  
  Reference - Environ Res 2016 Nov;151:763
- STUDY SUMMARY
  prenatal exposure to some pesticides may increase risk of autism spectrum disorders
  CASE-CONTROL STUDY: BMJ 2019 Mar 20;364:l962
  based on case-control study
  
  2,961 patients with autism spectrum disorders (ASD) were compared with 35,370 age, sex, and geographical region-matched controls
  
  among patients with ASD, 15% also had intellectual disability
  
  prenatal and infant exposure to pesticides was estimated as pounds of pesticides applied per acre per month within 2,000 meters from maternal residence
  
  factors associated with
  increased likelihood of ASD include prenatal exposure to
  glyphosate (adjusted odds ratio [OR] 1.16, 95% CI 1.06-1.27)
  
  chlorpyrifos (adjusted OR 1.13, 95% CI 1.05-1.23)
  
  diazinon (adjusted OR 1.11, 95% CI 1.01-1.21)
  
  malathion (adjusted OR 1.11, 95% CI 1.01-1.22)
  
  avermectin (adjusted OR 1.12, 95% CI 1.04-1.22)
  
  permethrin (adjusted OR 1.1, 95% CI 1.01-1.2)
  
  increased likelihood of ASD with intellectual disability include prenatal exposure to glyphosate, chlorpyrifos, diazinon, permethrin, methyl bromide, and myclobutanil
  
  Reference - BMJ 2019 Mar 20;364:l962

Factors Not Associated With Increased Risk

Vaccination

Vaccinations do not appear to be associated with an increased risk of autism spectrum disorders (ASD).
- STUDY SUMMARY
  measles, mumps, and rubella (MMR) vaccine not associated with increased risk of autism spectrum disorders
  COHORT STUDY: Ann Intern Med 2019 Apr 16;170(8):513
  Psychiatric_Disordersmeasles, mumps, and rubella (MMR) vaccine not associated with increased risk of autism spectrum disorders Ann Intern Med 2019 Mar 504/16/2019 01:22:06 PM
  based on population-based cohort study
  
  657,461 children born in Denmark from 1999 to 2010 were evaluated
  
  650,943 children were followed until end of study period contributing 5,025,754 person-years of follow-up data
  
  vaccination rates of MMR vaccine at 15 months (MMR1) and age 12 years (MMR2) obtained from Danish National Health Service Register
  no thimerosal-containing vaccines in Danish program during study period
  
  specific MMR vaccine used during study period contained following vaccine strains: Schwarz (measles, 2000-2007) or Ender Edmonton (measles, 2008-2013), Jeryl Lynn (mumps), and Wistar RA 27/3 (rubella)
  
  MMR1 vaccination rate 95.19%, median age at vaccination 1.34 years
  
  diagnosis of autism spectrum disorders (ASD) was obtained from the Danish Psychiatric Central Register using ICD-10 codes F84.0, F84.1, F84.5, F84.8, and F84.9
  
  children diagnosed with following disorders before first birthday were excluded from analysis: fragile X syndrome, tuberous sclerosis, Angelman syndrome, Down syndrome, DiGeorge syndrome, neurofibromatosis, Prader-Willi syndrome, and congenital rubella syndrome
  
  during follow-up
  6,517 children diagnosed with ASD (incidence rate 129.7 per 100,000 person-years)
  
  6,518 children were censored (335 had an autism-associated syndrome or condition, 628 died, 5,537 emigrated, and 18 disappeared from source registers for unknown reasons)
  
  mean age at first autism diagnosis 7.22 years (standard deviation 2.86)
  
  comparing MMR-vaccinated to MMR-unvaccinated children, no significant risk of ASD (adjusted hazard ratio 0.93 [95% CI 0.85-1.02])
  
  MMR vaccination not associated with increased ASD risk
  in 1-year risk periods after vaccination
  
  among other early childhood vaccinations subgroups
  
  among high-risk children
  
  among children with autistic siblings
  
  significant risk factors for ASD were older maternal and paternal age, smoking during pregnancy, preterm birth, poor 5-minute Apgar score, low birth weight, and large head circumference
  
  Reference - Ann Intern Med 2019 Apr 16;170(8):513
- STUDY SUMMARY
  vaccines not associated with increased risk of autism spectrum disorders
  SYSTEMATIC REVIEW: Vaccine 2014 Jun 17;32(29):3623
  based on systematic review of retrospective cohort and case-control studies
  
  review of 5 cohort and 5 case-control studies assessing association between vaccination and autism spectrum disorders (ASD), including Asperger syndrome
  
  mean follow-up in cohort studies 8.6 years (range 5-11 years)
  
  vaccination not associated with increased risk of ASD
  odds ratio (OR) 0.89 (95% CI 0.92-1.04) in 5 cohort studies with 1,256,407 persons
  OR following mercury (Hg) exposure 1 (95% CI 0.93-1.07) in 2 cohort studies with 250,750 persons
  
  OR following thimerosal exposure 1 (95% CI 0.77-1.31) in 1 cohort study with 467,450 persons
  
  OR following MMR vaccine 0.84 (95% CI 0.7-1.01) in 2 cohort studies with 538,207 persons
  
  Reference - Vaccine 2014 Jun 17;32(29):3623, commentary can be found in Vaccine 2015 Oct 13;33(42):5497
- STUDY SUMMARY
  maternal influenza infection or vaccination during pregnancy may not be associated with increased risk of autism spectrum disorders in offspring
  COHORT STUDY: JAMA Pediatr 2017 Jan 2;171(1):e163609
  based on cohort study
  
  196,929 children born at gestational age ≥ 24 weeks to mothers with singleton pregnancies were evaluated for autism spectrum disorders (ASD) at age 2-15 years
  
  1,400 mothers (0.7%) had influenza infection and 45,231 mothers (23%) had influenza vaccination during pregnancy
  
  3,101 children (1.6%) were diagnosed with ASD overall
  
  no significant difference in risk of ASD in children comparing maternal exposure to influenza infection or vaccination during pregnancy vs. no maternal exposure
  
  Reference - JAMA Pediatr 2017 Jan 2;171(1):e163609
- no correlation found between autism spectrum disorders and use of MMR vaccine - original article retracted and subsequent studies found no association with MMR vaccination
  - 1998 article reporting association between autism spectrum disorders (ASD) and MMR vaccine RETRACTED
    - RETRACTION due to several incorrect elements (including false claims that children were consecutively referred and investigators approved by local ethics committee) (Lancet 2010 Feb 6;375(9713):445)
    - original article was study of 12 children and reported autism as an incidental behavioral disorder associated with MMR vaccine in 9 children (Lancet 1998 Feb 28;351(9103):637)
  - subsequent studies found no association between ASD and MMR vaccine
    - no association between ASD and MMR vaccine, based on systematic review of 12 articles (Arch Pediatr Adolesc Med 2003 Jul;157(7):628)
    - no association between ASD and MMR vaccination in retrospective cohort study of 95,727 children with older siblings with or without ASD (JAMA 2015 Apr 21;313(15):1534), correction can be found in JAMA 2016 Jan 12;315(2):204, editorial can be found in JAMA 2015 Apr 21;313(15):1518
    - no association between ASD and MMR vaccination in case-control study of 189 young adults with ASD and 224 controls (Vaccine 2012 Jun 13;30(28):4292)
    - no association between autism or ASD and MMR vaccination in retrospective study of 537,303 children (N Engl J Med 2002 Nov 7;347(19):1477), editorial can be found in N Engl J Med 2002 Nov 7;347(19):1474, commentary can be found in Evid Based Nurs 2003 Jul;6(3):89
    - no association between MMR vaccine and pervasive developmental disorders in case-control study with 1,294 cases and 4,469 controls, 78% vs. 82% had MMR vaccination before diagnosis (or matched age) (Lancet 2004 Sep 11;364(9438):963)
    - no association between MMR vaccine and ASD in community-based case-control study with 98 cases and 142 controls (Arch Dis Child 2008 Oct;93(10):832), correction can be found in Arch Dis Child 2008 Dec;93(12):1079, editorial can be found in Arch Dis Child 2008 Oct;93(10):905
    - no support for association of MMR vaccine with autism or variant autism in case-control study (BMJ 2002 Feb 16;324(7334):393), commentary can be found in Mol Psychiatry 2003 Feb;8(2):133
- STUDY SUMMARY
  thimerosal-containing vaccines not associated with autism spectrum disorders
  SYSTEMATIC REVIEW: Pediatrics 2004 Sep;114(3):793
  based on systematic review of mostly observational studies without meta-analysis
  
  systematic review of 12 studies (7 cohort studies, 3 ecological studies. 2 pharmacokinetic studies) examining association between thimerosal and autism spectrum disorders (ASD)
  
  4 population-based cohort studies in United States, Sweden, Denmark, and United Kingdom report no association between thimerosal and ASD
  
  2 pharmacokinetic studies report short half-life of ethylmercury making association unlikely between exposure and ASD
  
  Reference - Pediatrics 2004 Sep;114(3):793, commentary can be found in Pediatrics 2005 Jan;115(1):200

Other Factors Not Associated With Increased Risk

Creatine deficiency syndromes, assisted reproductive technology, and maternal smoking do not appear to be associated with autism spectrum disorders (ASD).
- STUDY SUMMARY
  creatine deficiency syndromes not associated with autism spectrum disorders
  CASE-CONTROL STUDY: Pediatrics 2016 Jan;137(1):e20152672
  based on case-control study
  
  443 children and adolescents aged 2-18 years with ASD (84% male) and 128 healthy controls aged 6 months to 18 years had random spot urine screening for creatine metabolites
  
  no cases of creatine deficiency syndromes found in patients with autism spectrum disorders (ASD)
  
  no significant differences in creatinine, creatine, guanidinoacetate, or arginine levels comparing patients with ASD vs. controls
  
  Reference - Pediatrics 2016 Jan;137(1):e20152672
- STUDY SUMMARY
  assisted reproductive technology does not appear associated with diagnosis age of autism spectrum disorders or symptom severity
  COHORT STUDY: J Autism Dev Disord 2015 Sep;45(9):2991
  based on cohort study
  
  30,483 children with autism spectrum disorders (ASD) were evaluated for conception with assisted reproductive technology (ART) or natural conception (non-ART)
  
  1.7% of children conceived with ART (530 children)
  
  comparing ART-conceived children vs. non-ART conceived children, no significant differences in risk after adjusting for sociodemographic factors for
  early (< 4 years) diagnosis age of ASD
  
  severity of ASD symptoms at diagnosis (severe deficits in communication and social functioning)
  
  co-occurring intellectual disability (IQ < 70)
  
  Reference - J Autism Dev Disord 2015 Sep;45(9):2991
- STUDY SUMMARY
  maternal smoking during pregnancy does not appear associated with increased risk of autism spectrum disorders in children
  SYSTEMATIC REVIEW: J Autism Dev Disord 2015 Jun;45(6):1689
  based on systematic review of observational studies
  
  systematic review of 15 studies (11 case-control, 1 nested case-control, and 3 cohort studies) evaluating maternal prenatal smoking and increased risk of autism spectrum disorders (ASD) in offspring
  
  maternal smoking during pregnancy not significantly associated with increased risk of ASD in children
  
  Reference - J Autism Dev Disord 2015 Jun;45(6):1689

Factors Associated With Decreased Risk

Maternal nutritional supplementation:
- STUDY SUMMARY
  prenatal folic acid might be associated with decreased risk of autism spectrum disorders
  SYSTEMATIC REVIEW: Acta Paediatr 2019 Apr;108(4):600
  based on systematic review of mostly observational studies
  
  systematic review of 16 studies (2 randomized trials, 3 case-control, 11 population-based cohort) with 756,365 children aged 11 months to 15 years from 10 countries examining folic acid and neurodevelopment (7 studies) and autism spectrum disorders (ASD) (9 studies) were evaluated
  
  comparing any folic acid prenatal intake vs. no folic acid prenatal intake
  supplementation associated with decreased risk of ASD, pooled odds ratio 0.58 (96% CI 0.46-0.75), results limited by significant heterogeneity and potential publication bias
  
  supplementation not associated with mental performance or motor development
  
  Reference - Acta Paediatr 2019 Apr;108(4):600
- STUDY SUMMARY
  maternal multivitamin supplementation during pregnancy associated with reduced risk of autism spectrum disorders with intellectual disability in offspring
  COHORT STUDY: BMJ 2017 Oct 4;359:j4273
  based on prospective population-based cohort study
  
  273,107 children born between 1996 and 2007 in Sweden were evaluated at age 4-15 years
  23% had in utero exposure to multivitamins with or without iron and/or folic acid (multivitamin supplementation)
  
  43.3% had in utero exposure to iron and/or folic acid alone
  
  33.7% had no in utero exposure to nutritional supplements
  
  overall, 2% were diagnosed with autism spectrum disorders (ASD) and 0.43% were diagnosed with ASD with intellectual disability
  
  comparing in utero exposure to multivitamin supplementation vs. no in utero exposure to nutritional supplements
  ASD with intellectual disability in 0.25% vs. 0.47% (adjusted odds ratio 0.69, 95% CI 0.57-0.84)
  
  ASD without intellectual disability in 1.4% vs. 1.8% (not significant)
  
  no significant difference in risk of ASD with intellectual disability comparing in utero exposure to iron and/or folic acid alone vs. no in utero exposure to nutritional supplements
  
  consistent results in propensity score-matched analysis
  
  Reference - BMJ 2017 Oct 4;359:j4273
- STUDY SUMMARY
  prenatal folic acid supplementation around time of conception associated with decreased risk of autistic disorder in children
  COHORT STUDY: JAMA 2013 Feb 13;309(6):570
  based on prospective cohort study
  
  85,176 children born in 2002-2008 were followed for mean 6.4 years (age range 3.3-10.2 years)
  all children were from single-gestation pregnancies and born at ≥ 32 gestational weeks with birth weight ≥ 2.5 kg
  
  72% of mothers reported folic acid use from 4 weeks before conception to 8 gestational weeks but daily intake not specified
  
  0.13% of children were diagnosed with autistic disorder, 0.07% with Asperger syndrome, and 0.12% with pervasive developmental disorder not otherwise specified
  
  maternal folic acid supplementation associated with decreased risk of autistic disorder in children compared to no folic acid supplementation (adjusted odds ratio 0.61, 95% CI 0.41-0.9)
  
  no association between maternal folic acid supplementation and
  Asperger syndrome in adjusted analysis of 30,117 children
  
  pervasive developmental disorder not otherwise specified in adjusted analysis of 59,152 children
  
  Reference - JAMA 2013 Feb 13;309(6):570, editorial can be found in JAMA 2013 Feb 13;309(6):611

Associated Conditions

Developmental, medical, or psychiatric conditions are reported in > 70% of patients with autism spectrum disorders (ASD).
- Associated developmental conditions include:
  - Intellectual disability; reported in about 45%
  - Language disorders
  - Attention deficit hyperactivity disorder (ADHD) in children or adults; reported in 30%-40%
  - Tic disorders; reported in 14%-38% (including Tourette syndrome, reported in about 6.5%)
  - Motor abnormalities including motor delay, hypotonia, catatonia, impairments in coordination and movement preparation, praxis, and impairments in gait and balance
- Associated medical conditions include:
  - Epilepsy (in children or adults); reported in 8%-30%
  - Gastrointestinal problems such as chronic constipation (in adults or children), chronic diarrhea, abdominal pain, gastroesophageal reflux, and other conditions
  - Immune dysregulation; reported in < 40%
  - Sleep disorders; reported in 50%-80%
  - Tuberous sclerosis complex; reported in 25%-50%
  - Genetic syndromes (syndromic autism) are reported in about 15% of patients with autism spectrum disorders (ASD). In patients with:
    - Fragile X syndrome, ASD is reported in 21%-50%
    - Rett syndrome, ASD-like features are reported in most
    - Tuberous sclerosis, ASD is reported in 24%-60%
    - Down syndrome, ASD is reported in 5%-39%
    - Phenylketonuria (PKU), ASD is reported in 5%-20%
    - Childhood Autism Risks from Genetics and the Environment (CHARGE) syndrome, ASD is reported in 15%-50%
    - Angelman syndrome, ASD is reported in 50%-81%
    - Timothy syndrome, ASD is reported in 60%-70%
    - Joubert syndrome, ASD is reported in about 40%
    - Duchenne muscular dystrophy, ASD is reported in about 20%
- Associated psychiatric conditions include:
  - Anxiety disorders; reported in 42%-56%
    - Social anxiety disorder; reported 13%-29%
    - Generalized anxiety disorder; reported in 13%-22%
    - Obsessive compulsive disorder; reported in 7%-24%
  - Depression
  - Psychotic disorders; reported in 12%-17% (mostly in adults)
  - Substance use disorders; reported in ≤ 16%
  - Oppositional defiant disorder; reported in 16%-28%
  - Eating disorders; reported in 4%-5%
- Associated personality disorders, particularly in high-functioning adults, include:
  - Paranoid personality disorder; reported in < 20%
  - Schizoid personality disorder; reported in 21%-26%
  - Schizotypal personality disorder; reported in 2%-13%
  - Obsessive-compulsive personality disorder; reported in 19%-32%
  - Avoidant personality disorder; reported in 13%-25%
- Associated behavioral disorders include:
  - Self-injurious behaviors; reported in 85%
  - Pica; reported in about 36%
  - Suicidal ideation or attempts; reported in 11%-14%
- References - J Psychopharmacol 2018 Jan;32(1):3, Neurol Sci 2018 Jul;39(7):1279, Annu Rev Public Health 2017 Mar 20;38:81, Lancet 2014 Mar 8;383(9920):896, Pediatr Neurol 2015 Jan;52(1):25)

Etiology and Pathogenesis

Causes

Although autism spectrum disorders (ASD) are heritable, most causes are unknown. (Annu Rev Public Health 2017 Mar 20;38:81)
Approximately 15% of cases are reported to be linked to a known genetic cause via monogenic syndromes such as fragile X syndrome , tuberous sclerosis, and Timothy syndrome (Annu Rev Public Health 2017 Mar 20;38:81)
Infections during pregnancy, such as congenital rubella syndrome, are associated with an increased risk of ASD; however, there is little evidence for an association with more widely experienced infectious illnesses such as influenza (Cerebrum 2017 Jan;2017).
There is limited evidence that some medications during pregnancy such as valproic acid and serotonin reuptake inhibitors may increase risk of ASD (Cerebrum 2017 Jan;2017).

Pathogenesis

The specific mechanisms underlying clinical symptoms of autism spectrum disorder (ASD) are unclear, but likely involves brain changes and abnormalities and may also involve altered metabolism, gut, and immune function.
Numerous abnormalities in brain structure, connectivity, and function have been associated with ASD:
- Functional magnetic resonance imaging (MRI) and positron emission tomography (PET) scans show brain areas that are hypoactive during social perception and cognition tasks in patients with ASD, including:¹
  - Medial prefrontal cortex
  - Superior temporal sulcus
  - Temporoparietal junction
  - Amygdala
  - Fusiform gyrus
- Consistent evidence suggests that ASD may be associated with hyperserotonemia or a reduction in gamma-aminobutyric-acid (GABA) synthetic enzymes and receptors.
- Other possible brain abnormalities in patients with ASD have mostly inconsistent evidence, but include:
  - Dysfunction in executive functioning in frontal, parietal, and striatal areas
  - Dysfunction in the mirror system
  - Enhanced recruitment of temporal, occipital, and parietal cortices during visual tasks
  - Reduced recruitment of frontal cortices during visual tasks
  - Lower number of neurons in the amygdala, fusiform gyrus, and cerebellum in postmortem analyses (usually in older children)
  - Possible atypical connectivity patterns such as
    - Decreased frontoposterior connectivity
    - Increased parietal-occipital connectivity
    - Decreased long-range and increased short-range connectivity
  - In children aged < 2 years, increases in whole brain volume and amygdala volume have been reported.
- References - Annu Rev Public Health 2017 Mar 20;38:81, Lancet 2014 Mar 8;383(9920):896, Hum Brain Mapp 2012 Jul;33(7):1553
Evidence of brain changes reported in patients with ASD:
- STUDY SUMMARY
  autism spectrum disorder associated with decreased gray matter volume in amygdala-hippocampus complex and precuneus/medial parietal cortex
  SYSTEMATIC REVIEW: Arch Gen Psychiatry 2011 Apr;68(4):409
  based on systematic review of observational studies
  
  systematic review of 20 observational studies comparing gray matter volume (with whole-brain voxel-based morphometry of neuroimages) in patients with autism spectrum disorder to persons with typical development
  
  autism spectrum disorder associated with
  decreased gray matter volume in amygdala-hippocampus complex and precuneus/medial parietal cortex
  
  increased gray matter volume in left middle frontal gyrus
  
  autism spectrum disorder not associated with global gray matter volume in analysis of 12 studies
  
  Reference - Arch Gen Psychiatry 2011 Apr;68(4):409
- STUDY SUMMARY
  autism spectrum disorder associated with decreased area of corpus callosum
  SYSTEMATIC REVIEW: Biol Psychiatry 2009 Nov 15;66(10):935
  based on systematic review of case-control studies
  
  review of 10 studies with 253 patients with ASD and 250 controls
  
  ASD associated with moderate reduction in area of corpus callosum in analysis of 10 studies
  
  consistent results for subdivision of corpus callosum
  
  Reference - Biol Psychiatry 2009 Nov 15;66(10):935
- STUDY SUMMARY
  autism spectrum disorder associated with increased dendritic spine density and decreased pruning of dendritic spines
  CASE-CONTROL STUDY: Neuron 2014 Sep 3;83(5):1131
  based on case-control study
  
  10 persons ≤ 19 years old with ASD and 10 age-matched persons with typical development had postmortem examination of fixed and frozen temporal lobes (Brodmann Area 21) for dendritic spine morphology in cortical layer V pyramidal neurons
  
  comparing persons with ASD vs. persons with typical development
  mean dendritic spine density 11.32 spines/10 micrometers dendritic length vs. 8.81 spines/10 micrometers dendritic length (p < 0.017)
  
  decrease in dendritic spine density from ages 2-9 years to ages 13-19 years 16% vs. 45% (p < 0.001)
  
  estimated change in dendritic spine density per year -0.19 spines/10 micrometers vs. -0.4 spines/10 micrometers (p = 0.007) in linear regression analysis
  
  additional molecular analyses suggest that ASD is associated with high mammalian target of rapamycin (mTOR) activity and subsequent reduced macroautophagy, which has been suggested to lead to impaired synaptic pruning
  
  Reference - Neuron 2014 Sep 3;83(5):1131, commentary can be found in Neuron 2014 Sep 3;83(5):994
- STUDY SUMMARY
  hyperconnectivity within activity-defined brain networks appears to occur in children but not adolescents or adults with autism spectrum disorders, and hypoconnectivity between networks appears to occur in children and adolescents but not adults with the disorder
  CASE-CONTROL STUDY: Neuroimage Clin 2015;7:732
  based on case-control study
  
  72 persons with autism spectrum disorder and 72 age- and IQ-matched persons with typical development had resting state functional magnetic resonance imaging (fMRI)
  
  functional networks estimated with independent component analysis of blood oxygen level dependent (BOLD) activity from all persons
  within network differences estimated by comparing components between groups (stratified by age)
  
  between network differences estimated by comparing correlations between components between groups (stratified by age)
  
  children ≤ 11 years old with ASD associated with
  functional hyperconnectivity within networks involving
  right frontal pole
  
  bilateral insula and subcortical areas involving thalamus, hippocampus, and amygdala
  
  functional hypoconnectivity between parts of default mode network (consisting of posterior cingulate cortex, medial temporal lobes, and medial prefrontal cortex)
  
  adolescents aged 12-18 years with ASD associated with
  functional hypoconnectivity between parts of
  default mode network
  
  network involving insula and subcortical areas
  
  adults > 18 years old with ASD not associated with any functional connectivity differences within or between networks
  
  Reference - Neuroimage Clin 2015;7:732
- STUDY SUMMARY
  focal disruptions in prefrontal and temporal cortical laminar structure appears to be more common in children and adolescents with audism spectrum disorders
  CASE SERIES: N Engl J Med 2014 Mar 27;370(13):1209
  based on case series
  
  postmortem neocortical samples from 11 children aged 2-15 years with autism and 11 children without autism were analyzed
  
  focal patches of abnormal laminar structure and disorganization of neurons (but not glia) in prefrontal and temporal cortex found in 10 samples from children with autism and 1 sample from child without autism
  
  Reference - N Engl J Med 2014 Mar 27;370(13):1209
Altered metabolism, gut, and immune function may be associated with ASD via maternal immune function in response to environmental factors while pregnant.
- Increased inflammation has been reported in the central nervous structures in postmortem analyses of patients with ASD, including increased microglia activation and increased levels of interferon-gamma, interleukin (IL)-1 beta, IL-6, IL-12p40, tumor necrosis factor-alpha, and chemokine-C-C motif ligand-2 in the brain and cerebral spinal fluid.
- Indications of altered systemic immune function include:
  - Increased serum levels of IL-1 beta, IL-6, IL-8 and IL-12p40, macrophage inhibitory factor, platelet derived growth factor, C-C motif ligand 2 (CCL2), CCL5, complement proteins
  - Decreased serum levels of anti-inflammatory cytokine, transforming growth factor, immunoglobulin antibody G (IgG), and IgM
- Based on these findings, certain behavioral characteristics of ASD are thought to be due, in part, to the effects of
  - Altered cytokine levels on neuronal cell proliferation, neuron death, and synaptic pruning
  - Altered microglia on phagocytosis of neurons
- Reference - Brain Behav Immun 2012 Mar;26(3):383

Next Section >

Tables

All (2)

ASD Screening Tools for Children and Adolescents

View In Context

DynaMed Levels of Evidence

Quickly find and determine the quality of the evidence.

DynaMed provides easy-to-interpret Level of Evidence labels so users can quickly find and determine the quality of the best available evidence. Evidence may be labeled in one of three levels:

1Level 1 (likely reliable) Evidence

Representing research results addressing clinical outcomes and meeting an extensive set of quality criteria which minimizes bias.

There are two types of conclusions which can earn a Level 1 label: levels of evidence for conclusions derived from individual studies and levels of evidence for conclusions regarding a body of evidence.

2Level 2 (mid-level) Evidence

Representing research results addressing clinical outcomes, and using some method of scientific investigation, but not meeting the quality criteria to achieve Level 1 evidence labeling.

3Level 3 (lacking direct) Evidence

Representing reports that are not based on scientific analysis of clinical outcomes. Examples include case series, case reports, expert opinion, and conclusions extrapolated indirectly from scientific studies.

Grades of Recommendation

Guideline producers are now frequently using classification approaches for their evidence and recommendations, and these classifications are recognized and requested by guideline users. When summarizing guideline recommendations for DynaMed users, the DynaMed Editors are using the guideline-specific classifications and providing guideline classification approach when this is done.

Download the full version of Levels of Evidence

Gestational Age	Prevalence
Gestational Age	Male Persons (%)	Female Persons (%)
26 weeks	2.6	0
30 weeks	2	0.43
34 weeks	1	0.57
40 weeks	0.8	0.2

Condition

Autism Spectrum Disorders

Background Information

Description

Also Called

Epidemiology

Incidence/Prevalence

Details

Details

Details

Details

Risk Factors

Family History

Details

Details

Details

Genetic Risk Factors

Advanced Parental Age

Details

Details

Obstetric Factors

Details

Details

Details

Details

Details

Details

Details

Details

Details

Details

Details

Gestational Age and Weight at Birth

Details

Details

Details

Details

Prenatal and Infant Medication Exposures

Details

Details

Details

Details

Details

Details

Details

Details

Details

Perinatal Environmental Exposures

Details

Details

Details

Factors Not Associated With Increased Risk

Vaccination

Details

Details

Details

Details

Other Factors Not Associated With Increased Risk

Details

Details

Details

Factors Associated With Decreased Risk

Details

Details

Details

Associated Conditions

Etiology and Pathogenesis

Causes

Pathogenesis

Details

Details

Details

Details

Details

Tables

Related Topics

DynaMed Levels of Evidence

Quickly find and determine the quality of the evidence.

1Level 1 (likely reliable) Evidence

2Level 2 (mid-level) Evidence