nature reviews neurology https://doi.org/10.

1038/s41582-022-00764-0

Perspective Check for updates

Cohort-guided insights into

gene–environment interactions
in autism spectrum disorders
W. Ian Lipkin1,2,3,4 , Michaeline Bresnahan2,5 & Ezra Susser2,5
Abstract Sections

Prospective birth cohorts offer unprecedented opportunities to Introduction

investigate the pathogenesis of complex disorders such as autism, Changes in reported

in which gene–environment interactions must be appreciated in a prevalence

temporal context. This Perspective article considers the history of Philanthropy and parental
advocacy
autism research, including missteps that reflected an incomplete
understanding of the epidemiology of autistic spectrum disorders, Prospective birth cohorts

the effects of advocacy and philanthropy on the trajectory of scientific Animal models
inquiry, and the current and future roles of prospective birth cohort Conclusions
research in illuminating the pathology of these and other complex
disorders wherein exposures during gestation might not manifest until
later in life.

1
Center for Infection and Immunity, Mailman School of Public Health, Columbia University, New York, NY, USA.
2
Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, NY, USA.
3
Department of Neurology, Vagelos College of Physicians and Surgeons, Columbia University, New York, NY, USA.
4
Department of Pathology and Cell Biology, Vagelos College of Physicians and Surgeons, Columbia University,
New York, NY, USA. 5New York State Psychiatric Institute, New York, NY, USA. e-mail: wil2001@cumc.columbia.edu

Nature Reviews Neurology | Volume 19 | February 2023 | 118–125 118

Perspective

Introduction that mimic or obscure bona fide changes in prevalence. One analysis
Autism spectrum disorders (ASDs) are characterized by impairments in attributed 33% of the increase in reported ASD prevalence among
social skills and communication and the presence of repetitive behav- children born in Denmark during the period 1980–1991 to changes in
iours. Affecting approximately 1 in 50 children, ASDs are four times diagnostic criteria alone, 42% to increased case ascertainment alone
more common in boys than in girls1. As there are no objective labora- and 60% to the combination of changes in diagnostic criteria and
tory tests for ASDs, diagnosis requires expert clinical assessment. This increased case ascertainment10. This analysis not only suggested that
situation impedes early identification of ASDs at a time when the brain such artefacts played an important role in the increased prevalence of
is highly plastic and amenable to positive reinforcement exercises that ASDs but also confirmed that a substantial part of the increase could
can mitigate their presentation2. The social and economic burdens of not be explained by them.
ASD for those with these disorders, their families and their communities
are lifelong and substantial. Accordingly, understanding the patho- Philanthropy and parental advocacy
physiology of ASD is an urgent unmet clinical need, as are developing It is difficult to overstate the importance of philanthropy and paren-
methods for early diagnosis and treatment and ways to support those tal advocacy in the course of autism research and treatment (Fig. 1).
with ASD as they navigate their life course. Figure 1 shows a timeline of An influential early advocate was Bernard Rimland, a clinical psycholo-
seminal events in the history of biomedical research in autism. gist and father of a son with autism, who rejected the psychogenic
In this Perspective article, we consider how cohort studies have hypothesis. Rimland founded the Autism Society of America, the
contributed to our understanding of the epidemiology and pathogen- Autism Research Institute and the Defeat Autism Now Foundation. He
esis of ASDs, place cohort research into the context of complementary also served as the technical adviser on the film Rain Man, in which Dustin
work conducted in animal models, and speculate on the promise and Hoffman’s portrayal of a man with Asperger syndrome brought atten-
potential challenges of cohort research for dissecting the causes of tion to ASDs. Cure Autism Now and the National Alliance for Autism
ASDs and developing interventional strategies for these and other Research were private foundations initiated by parents to support
chronic disorders. peer-reviewed research projects and physician training programmes.
Each foundation recruited strong scientific advisory boards and funded
Changes in reported prevalence investigators who would not otherwise have had the resources to begin
The public perception of autism and its pathogenesis has been strongly autism research. Cure Autism Now also focused on the collection of
affected by changes in its reported prevalence. In the 1960s and 1970s, clinical materials to enable genetic analyses (Autism Genetic Research
worldwide autism prevalence estimates ranged from 0.04 to 2.00 cases Exchange). These two organizations were subsequently assimilated
per 1,000 of the population3. These rates rose tenfold in the 1980s into the Autism Speaks Foundation established by the Chief Executive
and 1990s. A lack of clarity on the basis for these prevalence estimates Officer of NBC Universal, grandparent of an affected child, who lever-
prompted a search for environmental factors that had been introduced aged his considerable influence to promote funding for autism research
into paediatric care at around the time of this increase. Two came to and treatment programmes. Amongst those programmes is the Baby
the fore: vaccines and acetaminophen (paracetamol). Sibs Research Consortium, which was founded in 2003 to expedite
The measles, mumps and rubella (MMR) vaccine, which contains the discovery of insights into the diagnosis and biology of ASDs by
a mixture of three live attenuated viruses, was licensed in 1971. This conducting prospective studies of the siblings of children with autism
vaccine is administered in two doses, the first given at 9–15 months of (who have a 20-fold increased risk of themselves receiving a diagnosis of
age and the second between 15 months and 6 years of age. Killed bacte- ASD). The largest organization to date is the Simons Foundation Autism
rial vaccines, including diphtheria toxin, pertussis bacteria and tetanus Research Initiative (SFARI), which was established by mathematicians
toxin (DPT vaccine), that were first introduced in the 1950s also came Jim and Marilyn Simons to support basic and clinical research into the
to attention because they contained the preservative thimerosal (ethyl neuroscience of autism. This organization has provided more than US
mercury), which raised the concern that even the low levels of mercury $200 million to more than 150 investigators worldwide since 2007,
present in DPT vaccine doses (which were given at 2 months, 4 months, including funding to study the immunopathology of disease in the
6 months, 15–18 months and 4–6 years of age) might be neurotoxic. Norwegian Autism Birth Cohort (ABC).
In 1980, epidemiologists at the US Centers for Disease Control reported
a link between the use of salicylates for fever management in paediatric Prospective birth cohorts
patients and Reye syndrome, a fatal hepatic encephalopathy4, which The pathogenesis of ASD in most individuals is unknown. The first
led to the substitution of acetaminophen for salicylates. In each of description of autism as a rare disorder in 1943 suggested a role for
these instances, temporal associations between the introduction of a socially distant caregivers — the ‘refrigerator parent’ hypothesis11.
vaccine or treatment and the apparent increase in the prevalence Although not unchallenged, this model remained dominant until the
of autism (a condition that was only rarely diagnosed before the age of 1970s, when an increased risk of autism was reported in children with
2 years) led to popular concerns that there was a causal link. No causal congenital rubella12 and other researchers reported finding evidence of
relationships have been found, despite intensive investigation. None- heritability in a study of monozygotic and dizygotic twin pairs in which
theless, spurious links between ASDs and the MMR vaccine5 or the at least one twin had autism13. In this study, the concordance rate for
vaccine preservative thimerosal6 continue to have a profound effect autism was 36% in monozygotic twins, whereas none of the dizygotic
on public acceptance and uptake of MMR and DPT7,8 — as well as other twins were concordant for this disease13. For comparison purposes, a
vaccines, including polio and SARS-CoV-2 vaccines9. meta-analysis published in 2016 found heritability estimates for ASDs
Prevalence is very challenging to estimate accurately for complex ranging from 38% to 90%14. The reasons for the variation in heritability
syndromes such as autism, in which changes over time in diagnostic estimates remain unclear.
criteria and in access to specialists with the expertise required to diag- This substantial heritability has sometimes been misinterpreted to
nose the condition can result in large changes in case ascertainment imply that environmental contributions to ASDs are minimal. However,

Nature Reviews Neurology | Volume 19 | February 2023 | 118–125 119

Perspective

Kanner describes 11 children Fig. 1 | Timeline of contributions to autism research. Although initially
with autism and emphasizes 1943 characterized as psychosis74–76, from the 1970s onwards autism was reframed as
emotionally cold parenting75 a developmental disorder with a heritable component15,77–81. Various exposures
Asperger’s paper on “autistic
psychopathy of childhood” were subsequently identified as contributing to the risk of autism spectrum
1944
is published in German76 disorder, the understanding of which continues to evolve5,16,17,19,20,25,82–89. Special
The Journal of Autism and projects initiated and supported by The National Alliance for Autism Research
Childhood Schizophrenia is Publication of Famine and Human
founded. Autism is framed within 1970 Development: The Dutch Hunger and Cure Autism Now include the Autism Tissue Program brain bank, the Autism
the paradigm of developmental Winter of 1944, investigating the Genetic Resource Exchange and the High-Risk Baby Siblings Consortium.
disorders rather than psychosis effects of gestational exposure to
famine on fetal growth and
1975
Folstein and Rutter’s paper postnatal development, mental
“Infantile autism: a genetic study and physical health78
of 21 twin pairs” examines these data are also consistent with genetic vulnerability to shared
monozygotic and dizygotic twins 1977 • A paper by Wing and Gould on
the prevalence of “severe
environmental exposures. This distinction can be illustrated by com-
and establishes a role for
genetics in autism pathogenesis15 impairments of social interaction” parison with phenylketonuria, in which virtually all affected individuals
1979 is published, establishing the have the genetic cause but the disease manifests only in response to
scope of the disorder79
• Chess’s paper describing the a specific (but ubiquitous) nutritional exposure that can, however, be
Wing raises awareness neuropsychiatric profiles of
of Asperger’s work and
eliminated from the diet. Thus, the heritability of phenylketonuria
1981 children exposed to rubella in
Asperger syndrome77 gestation during the rubella approaches 100% and the phenotype is ubiquitous without interven-
outbreak (1963–1965) highlights tion, and abrogating exposure to a specific environmental factor is key
First case reports of children with prenatal infectious exposures as
fragile X syndrome and autism81 1982
a potential risk factor for autism25
to its prevention. Therefore, the existence of seminal papers suggesting
contributions of genetic, nutritional, immunological and toxic environ-
• Asperger’s paper on “autistic
• Publication of Lord, Rutter mental factors to ASDs15–20 does not eliminate the social environment
psychopathy of childhood” is 1991 and Le Couteur’s paper, which
translated into English by Frith as having a critical role in neural development. Importantly, 50–70%
introduced what became the
• Results from randomized trial
of folic acid supplements before
gold standard research of children with ASD have comorbid attention deficit hyperactivity
1994 diagnostic interview, the ADI-R83
and during pregnancy and infant
• Paper describing cases of
disorder (ADHD)21, and Romanian infants confined to institutions
neural tube defects82 early in life who had only minimal contact with caregivers developed
autism in a cohort of individuals
with thalidomide embryopathy, behaviours consistent with ADHD22. Epigenetic analyses indicate that
A second parent-led organization, 1995 specifying a critical gestational
Cure Autism Now, is founded time-window exposure to the ADHD-like behaviours in these children might be associated with
thalidomide increasing the risk altered DNA methylation patterns versus those of children raised in
Publication of a rodent model 1996 of autism19
for valproate-induced autism84 family settings23. Of note, genetic and environmental models of ASD
• The National Alliance for pathogenesis are not mutually exclusive. Specific genetic defects,
Wakefield and colleagues Autism Research is founded, with
report in The Lancet that 1998 such as those that occur in patients with fragile X syndrome or are
the single goal of advancing
the MMR vaccine is biomedical research in autism by caused by gestational exposure to antiepileptic medications or tha-
associated with autism5 funding autism research lidomide, are sufficient to induce behaviours associated with autism.
1999
• Rutter and colleagues report In the majority of patients with ASDs, no links to a specific genetic or
Lord et al. introduce the
quasi-autistic traits in Romanian environmental factor can be identified. However, genetic vulnerability
orphans institutionalized before
Autism Diagnostic Observation 2000 adoption into UK families85
and environmental triggers almost certainly act together to cause ASD
Schedule (ADOS)86 in some individuals.
• The Autism Birth Cohort
(recruitment 1999–2008) is The observation that maternal use of anticonvulsants24 or thalido-
• Establishment of the
2003 established within the Norwegian
Simons Foundation Autism mide19 and rubella infection12,25 during pregnancy are all associated
Mother and Child Cohort Study,
Research Initiative
to study gene by environment by with an increased risk of autism suggested the presence of windows of
• Cure Autism Now merges
timing interactions and to
with Autism Speaks 2005 developmental vulnerability to ASDs prior to birth. Rigorous testing
identify markers for early
diagnosis of autism of this hypothesis requires the recruitment of population-based birth
Advanced paternal age
is linked to autism88 • Founding of the Baby Siblings
cohorts in which biological samples and environmental exposure
2006
Research Consortium, a data are collected at multiple time points during gestation. These
De novo germline mutations prospective study designed to
are established as significant early data then need to be combined with data on development, social
investigate the pathogenesis of
risk factors for autism16 2007 autism and to identify determinants and phenotypes collected during gestation and across
biomarkers for disease in childhood. Taken together, birth cohorts can (at least in theory) be used
The paper by Wakefield and siblings of children with autism
colleagues linking the MMR
2010 • Courchesne and colleagues to analyse omics data and environmental exposures, and determine
vaccine with autism is
retracted by The Lancet5
report association between which developmentally sensitive periods are most relevant. Some birth
early head growth and autism87
cohorts include almost all individuals with ASDs among population
The Integrative Psychiatric 2012 samples of >100,000 individuals and enable longitudinal follow-up
The Autism Birth Cohort
Research (iPSYCH) consortium is
reports that the use of folic of the children with ASDs. These cohorts can be used to analyse the
established as a population-
acid supplements during
based case-cohort sample (born relationships between genomic findings, environmental exposures
2013 pregnancy reduces the risk
between 1981 and 2005) with the
of autism in offspring17 and sensitive developmental periods. All existing cohorts, however,
aim of “unravelling the genetic
and environmental architecture have limitations. The ABC (nested within the Norwegian Mother, Father
of severe mental disorders”89 The Autism Birth Cohort
2022 and Child Cohort Study (MoBA)) is arguably the largest such cohort and
reports that evidence of
cytokine activation in maternal has most, although not all, of the features required for such analyses.
Research and cohort studies
mid-gestational plasma and in
Philanthropy and parental cord blood is linked to autism20 Yet it is still subject to selection bias; only 40% of the pregnant women
advocacy groups
approached chose to participate.

Nature Reviews Neurology | Volume 19 | February 2023 | 118–125 120

Perspective

Table 1 | Types of cohort studies of autism spectrum disorder

Study exemplar and Representativeness of Phenotypic data Blood samplesb Exposure data Intergenerationalc
details participants (non-biological)

MARBLES (prospective) Over-represents strong Multiple direct intensive Maternal: pre-conception, Questionnaires, health No
in 463 baby siblings genetic transmission assessment first, second and third records, blood samples
of children with ASD, of Under-represents de novo trimester
whom 64 developed ASD germline mutations Child: cord blood, 3, 6, 12
Under-represents weak and 24 months postnatal
genetic transmission Younger sibling: cord
blood, 3, 6, 12, 24 and 36
months postnatal
MINERVA iCARE pure Only treated patients are Registry diagnosis only None Linkage to multiple Yes
registry (historical with identified in registries population registries
concurrent follow-up) in
5,766,794 participants
including 30,902 with ASD
iPSYCHa registry plus Only treated patients are Registry diagnosis only Heel stick (baby at birth) Heel stick, linkage to Yes
(historical with concurrent identified in registries multiple population
follow-up) in 1,472,762 registries
participants including
16,146 with ASD
ABC MoBAa population- Highly representative Registry diagnosis with Mother: prenatal and Blood samples, linkages Yes
based birth cohort with four complementary verification, systematic postnatal to multiple population
(prospective) in 114,552 methods of case chart review, intensive Father: prenatal registries, questionnaires,
participants including 750 ascertainment direct assessment, health records with
Child: birth, infancy,
with ASD multiple follow-up systematic review
childhood
questionnaires
ALSPAC hybrid, Highly representative for Extensive and multiple Repeated maternal Blood samples, Limited
(prospective) in 13,868 traits; multiple follow-up follow-up assessments blood samples during questionnaires
participants (not known assessments pregnancy
how many with ASD)
LBW exposure-based Representative children Extensive and follow-up Birth, infancy, childhood Biological samples, No
cohort (prospective) with ASD among preterm assessments questionnaires, record
in 1,105 participants infants review, neonatal cranial
including 14 with ASD ultrasonography, brain
scans
ABC MoBA, Autism Birth Cohort substudy of the Norwegian Mother and Child Cohort Study; ALSPAC, Avon Longitudinal Study of Children and Parents; ASD, autism spectrum disorder;
iCARE, International Collaboration for Autism Registry Epidemiology; LBW, low birth weight; MARBLES, Markers of Autism Risk in Babies – Learning Early Signs; MINERVA, Multigenerational
Familial and Environmental Risk for Autism. aFor ongoing studies, patient numbers reflect recruitment at the time the cited paper was published. bBiospecimen collections (maternal urine, hair,
saliva, placenta, vaginal secretions) not shown. cIncludes data from grandparents.

Although population-based birth cohorts are potentially the most cohorts32. A study conducted in California had previously suggested
informative, other cohort designs (for example, those focused on that increased maternal folate intake overall was associated with a
children resulting from subsequent pregnancies in mothers of a child reduced risk of ASD in offspring33. However, the Norwegian study was
with ASD) can provide complementary data. Table 1 presents one distinctive in several ways. Perhaps the most important is that, unlike
exemplar for each of six types of cohort study, and notes the typical many countries (including the USA and UK), Norway does not fortify
features of each design as used in ASD research26–30. Given the variety of foods with folic acid. Accordingly, the folate levels of women who did
characteristics among cohort studies that share the same basic design and did not take supplements during gestation were markedly dif-
and the large number of cohort studies now established, ‘typical’ as ferent. A related strength of the ABC study is that supplementation
used here does not imply uniformity. Presentation of all current ASD usually involved small doses, such as 400 μg of folic acid included in a
cohorts and in-depth analysis of the differences between them are multivitamin pill. This factor is relevant because some studies suggest
beyond the scope of this paper. Accordingly, we focus on how these that high levels of synthetic folic acid (as opposed to natural folate
six cohort designs offer complementary data with respect to three in foods) might have adverse effects34,35. Further analysis of data from
gestational exposures linked to ASD: maternal folate intake, maternal a study of baby siblings of children with an ASD, which examined the
immune response and preterm birth. effect of folic acid supplementation on the risk of ASD recurrence in
high-risk families, showed that supplementation in the first month
Folate and folic acid of pregnancy halved the risk of autism36. Complementary evidence
The ABC study was the first to show specifically that maternal folic has also been provided by registry studies that demonstrates an
acid supplementation is associated with a reduced risk of language increased risk of ASD associated with short interpregnancy intervals37,38.
delay31 and ASD in offspring17. Although these initial findings were based One potential explanation for this finding is that folate depleted dur-
on early data in a subset of individuals, the same associations have ing pregnancy might not be fully restored by the start of a subsequent
now been found in subsequent larger studies, in both this and other pregnancy.

Nature Reviews Neurology | Volume 19 | February 2023 | 118–125 121

Perspective

Folate is fundamental to brain development. For example, folate Preterm birth

is a critical donor of the methyl groups used in DNA and histone Although infant survival following preterm birth has improved expo-
methylation. Although early gestation and/or periconception have nentially over the past 40 years, prematurity is not benign. Indeed,
been identified as sensitive periods with regard to the effects of folic linked registry studies provided the first evidence of an association
acid supplementation, the epigenetic activity of folate might extend between preterm birth and ASD53–55. The association is particularly
the period of folate-dependent vulnerability to ASD. Animal studies strong in brain imaging studies conducted during infancy, which reveal
suggest that maternal early life exposure to either low or high levels evidence of structural abnormalities, including ventricular enlarge-
of folate, including the mother’s own gestational exposure (which is ment and brain volume overgrowth56,57. The addition of genetic analyses
when her oocytes are developing) as well as paternal lifetime exposure and fetal imaging to studies in pregnant women at high risk of prema-
to folic acid deficiency and supplementation, all pose a theoretical ture delivery might ultimately help to untangle the factors underlying
risk to offspring by altering DNA and histone methylation patterns39,40. these associations. An example of such a study that measures autistic
The findings of several birth cohort studies that the characteristics of traits is the population-based prospective Generation R cohort study58,
a child’s maternal grandmother (such as smoking during pregnancy) which aims to identify early environmental and genetic causes of nor-
are linked to the child’s risk of autistic traits and autism41–43 might be mal and abnormal growth, development and health in participants
tied to this intergenerational mechanism. monitored from fetal life until young adulthood.

Immune dysregulation Animal models

Activation of dysregulated maternal immune responses during preg- Animal models cannot fully reproduce the complex spectrum of behav-
nancy has been implicated in the pathogenesis of ASD and other ioural and social deficits that occur in people with ASDs. Nonetheless,
neurodevelopmental disorders. We set aside the question of whether features such as repetitive movements and impaired social interac-
the observed immune dysregulation is due to infection, medication, tion can be used to investigate the plausible possibility that genetic,
fever, autoimmunity or genetic or other factors. Instead, we consider microbial, immune, toxic and nutritional factors contribute to the risk
the evidence that certain kinds of immune dysregulation could be of ASD, and to explore mechanisms of ASD pathogenesis and potential
involved in the pathogenesis of ASD and/or be an early indicator of interventions. The majority of animal models use genetically modified
heightened risk of ASD. mice, in which overexpression or knockdown of specific genes has been
Robust evidence for immune dysregulation in mothers midway used to assess their contributions and those of associated pathways to
through their pregnancy and in neonatal cord blood has been devel- development and behaviour. Some animal models focus on the envi-
oped in the ABC substudy of the MoBA cohort in Norway, which was ronmental exposures implicated in ASD by epidemiological research.
designed to interrogate biological processes occurring at multiple We highlight some examples of these animal models; however, readers
points in pregnancy. Immune profiling of blood samples collected will find more details in recent reviews59,60.
from mothers at 17–21 weeks of pregnancy and cord blood samples Gestational exposure to thalidomide and valproic acid are both
collected at birth found differences consistent with systemic inflam- associated with an increased risk of autism19,61. Accordingly, rodent
mation in a wide range of analytes in samples from both boys and girls models were established to dissect the developmental consequences
with ASDs20. The number of elevated pro-inflammatory cytokines, of exposure to these agents. In studies of pregnant rats, exposure
chemokines and adhesion molecules associated with ASD and the mean to thalidomide resulted in auditory hypersensitivity in offspring62.
effect sizes of these elevations were larger in girls than in boys. These Auditory hypersensitivity is common in individuals with autism, in
effects of systemic inflammation, and the difference in effect sizes whom this feature has been linked to electrophysiological abnormali-
between girls and boys, were most pronounced in mid-gestation20. ties in the primary auditory cortex63. Gestational exposure of rats to
Previous research examining immune signatures associated with ASD valproic acid resulted in reduced social interaction and exploratory
was limited: only a few studies had examined maternal immune mol- activity reminiscent of features described in autism, accompanied
ecules in mid-gestation44–46 and a handful had examined immune by reduced numbers of myelinated axons and aberrant myelin sheath
molecules in dried blood spots47–51. Most of these studies examined only ultrastructure in the corpus callosum64. These histopathological find-
a small number of analytes, and although most had controlled for sex, ings are consistent with corpus callosum abnormalities reported in
sex-specific differences were not detailed44–51. individuals with autism and with conceptual models of the disorder
The strength of these findings in the ABC study prompted further that focus on impaired brain connectivity65.
scrutiny of the biomarker potential of immune signatures in maternal Rodent and primate models of gestational exposure to bacteria
mid-gestation and cord blood samples. Five predictive models were and viruses and their gene products, as well as cytokines and antibod-
built for girls and boys in relation to each sample type, and tested ies, have also been established66–69. Influenza virus infection in pregnant
in 80% of the ABC population. Model validation was conducted in mice resulted in offspring with deficient prepulse inhibition of their
test sets composed of the remaining 20%20. The predictive models acoustic startle response, as well as deficits in exploratory behaviours
were able to distinguish children with ASD from control children and social interaction. These findings did not reflect a direct effect of
without ASD among both boys and girls in the two sample types: exposure to this infectious virus because they could be reproduced
model average areas under the receiver operating characteristic curve by exposure to the viral mimic, synthetic double-stranded RNA poly­
were 0.848 in mid-gestation and 0.846 in cord blood for boys, and inosinic-polycytidylic acid (poly I:C), or to the pro-inflammatory
0.965 in mid-gestation and 0.917 in cord blood for girls20. Ultimately cytokine IL-6 (refs. 66,70). Other researchers found additional support
this work might result in the development of early biomarker(s) for for an indirect, immune-mediated mechanism of behavioural deficits in
ASD52. As yet, other cohorts have produced only limited complemen- the offspring of pregnant mice exposed to bacterial superantigens71 and
tary data, but we expect that additional data will be reported in the mitigation of the behavioural deficits in animals gestationally exposed
near future. to anti-inflammatory drug treatment following immune stimulation

Nature Reviews Neurology | Volume 19 | February 2023 | 118–125 122

Perspective

with poly I:C68. Animal models have also been used to examine the role of case ascertainment and sampling, but also because genetic and
of the gastrointestinal microbiome in ASD pathogenesis. Intestinal den- environmental factors can vary. For example, in the ABC, the strength
dritic cells from pregnant mice colonized with segmented filamentous of the mitigating effect of folic acid supplements on the risk of autism
bacteria secrete cytokines that promote the differentiation of intestinal was considerably increased with respect to some other studies because
T helper 17 cells, which results in inflammation in their offspring as well foods are not supplemented with folate in Norway. As another example,
as cortical and behavioural abnormalities72. cytokine levels in plasma samples stored at ultra-low temperatures
Germ-free mice colonized with microbiota from human donors differ from those in whole blood collected on filter paper and stored
with autism show autistic behaviours and have reduced levels of two at room temperature. Findings that remain consistent across cohorts
GABA agonists, 5-aminovaleric acid and taurine, in their brain and despite differences in sampling methods are considered particularly
faeces. The autistic behaviours of these mice normalized when these robust; however, failures to replicate findings across different cohorts
two GABA agonists were provided through oral supplementation73. must be considered in this context. Although DNA, serum and plasma
The extent to which the results of these individual studies can be rep- samples are readily collected and their storage at −20 °C is adequate
licated and the fidelity of animal models in reproducing the cardinal for genetic analyses or serology, the materials required for transcrip-
features of autism are unclear. Nonetheless, in concert, they confirm tomics, metabolomics, and proteomics are more labile. In the ABC, we
that gestational exposures to toxins and inflammation can result in stored plasma and RNA samples at −80 °C and learned to our chagrin
neurodevelopmental damage that culminates in social deficits and/or that this was insufficient to maintain their quality for use in some
neurodiversity (the term neurodiversity acknowledges that not all assays. Accordingly, at least a portion of each sample should be stored
differences are deficits). at −140 °C for future use, including with platforms that do not yet exist.
We also did not collect faecal or vaginal samples in the ABC that could
Conclusions have been used in microbiome studies. However, sample biobanking is
For millennia, philosophers and theologians have debated the rela- resource-intensive and expensive. Therefore, despite our best efforts
tive importance of nature and nurture as determinants of biology and to future-proof data collection, regret for opportunities missed can
behaviour. With the advent of prospective birth cohorts, we have an be anticipated as laboratory methods continually evolve and require
opportunity to move from abstract discussions to rigorous dissec- different sample types or management.
tion of the interactions of genetic and environmental factors in health Ethical challenges will almost certainly emerge as population-
and disease. Several important points that must be considered in the based cohort research reveals links between health outcomes and
design of future prospective cohorts are highlighted in the following genetic as well as environmental factors. In some instances, ASD
paragraphs. biomarkers might be found that enable early intervention to prevent
The effect of an environmental factor could vary according to or ameliorate the condition. In other situations, the information found
when (for example, in which trimester of gestation) and to whom might only predict an individual’s risk of ASD. Leadership teams should
(for instance, mother, father or grandparent) it is introduced; thus, our consider whether and how such sensitive data will be released to cohort
ability to identify its consequences depends, at minimum, on whether participants, and should include ethicists who can appropriately
we have data that bracket critical intervals. Several windows exist dur- manage its communication.
ing development wherein the absence of a nutrient (such as folate), Finally, the future of cohort research will almost certainly entail
lack of social interaction or the presence of toxins (such as alcohol, the use of machine learning methods for integrating genetic, epige-
heavy metals, anticonvulsant drugs and by-products of infection) netic and multiomics datasets across populations in meta-analyses.
can profoundly influence brain structure and function. The effect In concert, these approaches aim to generate important insights into
of those same factors might be qualitatively or quantitatively differ- the pathology of a wide range of acute and chronic diseases and might
ent at a different time. For example, lead intoxication in children can lead to biomarkers for early diagnosis of individuals at risk, as well
result in cognitive dysfunction, whereas in adults it is more commonly as new methods for mitigating risk. In this new era of ASD research,
associated with peripheral neuropathy. investigators will have an opportunity to use birth cohorts to embrace
The value of a specific cohort as a resource is determined by the the complexity of interactions of genetic and environmental factors
size and composition of its population, the onset and frequency of col- in a temporal context.
lection of data and samples, the types of data and samples collected,
the duration of follow-up and the commitment of cohort members to Published online: 16 January 2023
continued engagement with the project. In the Norwegian ABC, the References
parental questionnaire completion rate was 95% at the 18-month time 1. Lai, M. C., Lombardo, M. V. & Baron-Cohen, S. Autism. Lancet 383, 896–910 (2014).
point but decreased to 61% at the 36-month time point74. Fortunately, 2. Orinstein, A. J. et al. Intervention for optimal outcome in children and adolescents with
a history of autism. J. Dev. Behav. Pediatr. 35, 247–256 (2014).
the National Patient Registry (NPR) allowed us to identify additional 3. Fombonne, E. Editorial: The rising prevalence of autism. J. Child. Psychol. Psychiatry 59,
individuals with ASDs who were not found through questionnaire 717–720 (2018).
responses, either because parents did not respond or diagnosis was 4. Starko, K. M., Ray, C. G., Dominguez, L. B., Stromberg, W. L. & Woodall, D. F. Reye’s
syndrome and salicylate use. Pediatrics 66, 859–864 (1980).
delayed. Without the NPR we would not have been able to use the ges- 5. Wakefield, A. J. et al. Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and
tational questionnaires and patient samples that were required to find pervasive developmental disorder in children. Lancet 351, 637–641 (1998); retraction
links between outcomes and exposures. A thoughtful experimental 375, 445 (2010).
6. Geier, D. A. & Geier, M. R. A comparative evaluation of the effects of MMR immunization
design should limit participant fatigue and maximize retention by and mercury doses from thimerosal-containing childhood vaccines on the population
reducing the frequency and complexity of data and sample collection. prevalence of autism. Med. Sci. Monit. 10, PI33–PI39 (2004).
The validity and generalizability of findings from individual 7. Hornig, M. et al. Lack of association between measles virus vaccine and autism with
enteropathy: a case-control study. PLoS ONE 3, e3140 (2008).
cohorts must be tested against those from other cohorts. Such com- 8. Gerber, J. S. & Offit, P. A. Vaccines and autism: a tale of shifting hypotheses. Clin. Infect.
parisons can be challenging, not only because of differences in methods Dis. 48, 456–461 (2009).

Nature Reviews Neurology | Volume 19 | February 2023 | 118–125 123

Perspective

9. Velasquez-Manoff, M. The anti-vaccine movement’s new frontier. New York Times 41. Golding, J. et al. Ancestral smoking and developmental outcomes: a review of
https://www.nytimes.com/2022/05/25/magazine/anti-vaccine-movement.html (2022). publications from a population birth cohort. Biol. Reprod. 105, 625–631 (2021).
10. Hansen, S. N., Schendel, D. E. & Parner, E. T. Explaining the increase in the prevalence 42. Golding, J., Steer, C. & Pembrey, M. Parental and grandparental ages in the autistic
of autism spectrum disorders: the proportion attributable to changes in reporting spectrum disorders: a birth cohort study. PLoS ONE 5, e9939 (2010).
practices. JAMA Pediatr. 169, 56–62 (2015). 43. Golding, J. et al. Grand-maternal smoking in pregnancy and grandchild’s autistic traits
11. Kanner, L. Problems of nosology and psychodynamics of early infantile autism. and diagnosed autism. Sci. Rep. 7, 46179 (2017).
Am. J. Orthopsychiatry 19, 416–426 (1949). 44. Goines, P. E. et al. Increased midgestational IFN-γ, IL-4 and IL-5 in women bearing a child
12. Chess, S. Autism in children with congenital rubella. J. Autism Child. Schizophr. 1, 33–47 with autism: a case-control study. Mol. Autism 2, 13 (2011).
(1971). 45. Jones, K. L. et al. Autism with intellectual disability is associated with increased levels
13. Folstein, S. & Rutter, M. Infantile autism: a genetic study of 21 twin pairs. J. Child. Psychol. of maternal cytokines and chemokines during gestation. Mol. Psychiatry 22, 273–279 (2017).
Psychiatry 18, 297–321 (1977). 46. Casey, S. et al. Maternal mid-gestation cytokine dysregulation in mothers of
14. Tick, B., Bolton, P., Happe, F., Rutter, M. & Rijsdijk, F. Heritability of autism spectrum children with autism spectrum disorder. J. Autism Dev. Disord. https://doi.org/10.1007/
disorders: a meta-analysis of twin studies. J. Child. Psychol. Psychiatry 57, 585–595 s10803-021-05271-7 (2021).
(2016). 47. Krakowiak, P. et al. Neonatal cytokine profiles associated with autism spectrum disorder.
15. Folstein, S. & Rutter, M. Genetic influences and infantile autism. Nature 265, 726–728 (1977). Biol. Psychiatry 81, 442–451 (2017).
16. Sebat, J. et al. Strong association of de novo copy number mutations with autism. 48. Abdallah, M. W. et al. Amniotic fluid chemokines and autism spectrum disorders:
Science 316, 445–449 (2007). an exploratory study utilizing a Danish historic birth cohort. Brain Behav. Immun. 26,
17. Suren, P. et al. Association between maternal use of folic acid supplements and risk 170–176 (2012).
of autism spectrum disorders in children. JAMA 309, 570–577 (2013). 49. Heuer, L. S. et al. An exploratory examination of neonatal cytokines and chemokines as
18. Zhou, X. et al. Integrating de novo and inherited variants in 42,607 autism cases identifies predictors of autism risk: the early markers for autism study. Biol. Psychiatry 86, 255–264
mutations in new moderate-risk genes. Nat. Genet. 54, 1305–1319 (2022). (2019).
19. Stromland, K., Nordin, V., Miller, M., Akerstrom, B. & Gillberg, C. Autism in thalidomide 50. Zerbo, O. et al. Neonatal cytokines and chemokines and risk of autism spectrum disorder:
embryopathy: a population study. Dev. Med. Child. Neurol. 36, 351–356 (1994). the early markers for autism (EMA) study: a case-control study. J. Neuroinflammation 11,
20. Che, X. et al. Maternal mid-gestational and child cord blood immune signatures are 113 (2014).
strongly associated with offspring risk of ASD. Mol. Psychiatry https://doi.org/10.1038/ 51. Abdallah, M. W. et al. Neonatal levels of cytokines and risk of autism spectrum disorders:
s41380-021-01415-4 (2022). an exploratory register-based historic birth cohort study utilizing the Danish newborn
21. Hours, C., Recasens, C. & Baleyte, J. M. ASD and ADHD comorbidity: what are we talking screening biobank. J. Neuroimmunol. 252, 75–82 (2012).
about? Front. Psychiatry 13, 837424 (2022). 52. Walsh, P., Elsabbagh, M., Bolton, P. & Singh, I. In search of biomarkers for autism:
22. Kennedy, M. et al. Early severe institutional deprivation is associated with a persistent scientific, social and ethical challenges. Nat. Rev. Neurosci. 12, 603–612 (2011).
variant of adult attention-deficit/hyperactivity disorder: clinical presentation, 53. Hultman, C. M., Sparen, P. & Cnattingius, S. Perinatal risk factors for infantile autism.
developmental continuities and life circumstances in the English and Romanian Epidemiology 13, 417–423 (2002).
Adoptees study. J. Child. Psychol. Psychiatry 57, 1113–1125 (2016). 54. Eaton, W. W., Mortensen, P. B., Thomsen, P. H. & Frydenberg, M. Obstetric complications
23. Kumsta, R. et al. Severe psychosocial deprivation in early childhood is associated with and risk for severe psychopathology in childhood. J. Autism Dev. Disord. 31, 279–285 (2001).
increased DNA methylation across a region spanning the transcription start site of 55. Schendel, D. & Bhasin, T. K. Birth weight and gestational age characteristics of children
CYP2E1. Transl. Psychiatry 6, e830 (2016). with autism, including a comparison with other developmental disabilities. Pediatrics
24. Moore, S. J. et al. A clinical study of 57 children with fetal anticonvulsant syndromes. 121, 1155–1164 (2008).
J. Med. Genet. 37, 489–497 (2000). 56. Movsas, T. Z. et al. Autism spectrum disorder is associated with ventricular enlargement
25. Chess, S. Follow-up report on autism in congenital rubella. J. Autism Child. Schizophr. 7, in a low birth weight population. J. Pediatr. 163, 73–78 (2013).
69–81 (1977). 57. Hazlett, H. C. et al. Early brain development in infants at high risk for autism spectrum
26. Hertz-Picciotto, I. et al. A prospective study of environmental exposures and early disorder. Nature 542, 348–351 (2017).
biomarkers in autism spectrum disorder: design, protocols, and preliminary data from 58. Jaddoe, V. W. et al. The Generation R study: design and cohort profile. Eur. J. Epidemiol.
the MARBLES study. Env. Health Perspect. 126, 117004 (2018). 21, 475–484 (2006).
27. Pedersen, C. B. et al. The iPSYCH2012 case-cohort sample: new directions for unravelling 59. Kazdoba, T. M. et al. Translational mouse models of autism: advancing toward
genetic and environmental architectures of severe mental disorders. Mol. Psychiatry 23, pharmacological therapeutics. Curr. Top. Behav. Neurosci. 28, 1–52 (2016).
6–14 (2018). 60. Ergaz, Z., Weinstein-Fudim, L. & Ornoy, A. Genetic and non-genetic animal models for
28. Schendel, D. E. et al. The International Collaboration for Autism Registry Epidemiology autism spectrum disorders (ASD). Reprod. Toxicol. 64, 116–140 (2016).
(iCARE): multinational registry-based investigations of autism risk factors and trends. 61. Christensen, J. et al. Prenatal valproate exposure and risk of autism spectrum disorders
J. Autism Dev. Disord. 43, 2650–2663 (2013). and childhood autism. JAMA 309, 1696–1703 (2013).
29. Boyd, A. et al. Cohort profile: the ‘children of the 90s’ – the index offspring of the 62. Tsugiyama, L. E., Ida-Eto, M., Ohkawara, T., Noro, Y. & Narita, M. Altered neuronal activity
Avon Longitudinal Study of Parents and Children. Int. J. Epidemiol. 42, 111–127 (2013). in the auditory brainstem following sound stimulation in thalidomide-induced autism
30. Pinto-Martin, J. et al. The central New Jersey neonatal brain haemorrhage study: design model rats. Congenit. Anom. 60, 82–86 (2020).
of the study and reliability of ultrasound diagnosis. Paediatr. Perinat. Epidemiol. 6, 63. Matsuzaki, J. et al. Differential responses of primary auditory cortex in autistic spectrum
273–284 (1992). disorder with auditory hypersensitivity. Neuroreport 23, 113–118 (2012).
31. Roth, C. et al. Folic acid supplements in pregnancy and severe language delay in children. 64. Uccelli, N. A. et al. Neurobiological substrates underlying corpus callosum
JAMA 306, 1566–1573 (2011). hypoconnectivity and brain metabolic patterns in the valproic acid rat model of autism
32. Liu, X., Zou, M., Sun, C., Wu, L. & Chen, W. X. Prenatal folic acid supplements and spectrum disorder. J. Neurochem. 159, 128–144 (2021).
offspring’s autism spectrum disorder: a meta-analysis and meta-regression. J. Autism 65. Frith, C. Is autism a disconnection disorder? Lancet Neurol. 3, 577 (2004).
Dev. Disord. 52, 522–539 (2022). 66. Shi, L., Fatemi, S. H., Sidwell, R. W. & Patterson, P. H. Maternal influenza infection causes
33. Schmidt, R. J. et al. Maternal periconceptional folic acid intake and risk of autism marked behavioral and pharmacological changes in the offspring. J. Neurosci. 23,
spectrum disorders and developmental delay in the CHARGE (Childhood Autism Risks 297–302 (2003).
from Genetics and Environment) case-control study. Am. J. Clin. Nutr. 96, 80–89 (2012). 67. Bauman, M. D. et al. Activation of the maternal immune system during pregnancy alters
34. Maruvada, P. et al. Knowledge gaps in understanding the metabolic and clinical behavioral development of rhesus monkey offspring. Biol. Psychiatry 75, 332–341 (2014).
effects of excess folates/folic acid: a summary, and perspectives, from an NIH workshop. 68. De Miranda, J. et al. Induction of Toll-like receptor 3-mediated immunity during gestation
Am. J. Clin. Nutr. 112, 1390–1403 (2020). inhibits cortical neurogenesis and causes behavioral disturbances. mBio 1, e00176-10 (2010).
35. Naderi, N. & House, J. D. Recent developments in folate nutrition. Adv. Food Nutr. Res. 83, 69. Bauman, M. D. et al. Maternal antibodies from mothers of children with autism alter brain
195–213 (2018). growth and social behavior development in the rhesus monkey. Transl. Psychiatry 3, e278
36. Schmidt, R. J., Iosif, A. M., Guerrero Angel, E. & Ozonoff, S. Association of maternal (2013).
prenatal vitamin use with risk for autism spectrum disorder recurrence in young siblings. 70. Hsiao, E. Y. & Patterson, P. H. Activation of the maternal immune system induces
JAMA Psychiatry 76, 391–398 (2019). endocrine changes in the placenta via IL-6. Brain Behav. Immun. 25, 604–615 (2011).
37. Cheslack-Postava, K. et al. Increased risk of autism spectrum disorders at short and 71. Glass, R., Norton, S., Fox, N. & Kusnecov, A. W. Maternal immune activation with
long interpregnancy intervals in Finland. J. Am. Acad. Child. Adolesc. Psychiatry 53, staphylococcal enterotoxin A produces unique behavioral changes in C57BL/6 mouse
1074–1081.e4 (2014). offspring. Brain Behav. Immun. 75, 12–25 (2019).
38. Gunnes, N. et al. Interpregnancy interval and risk of autistic disorder. Epidemiology 24, 72. Kim, S. et al. Maternal gut bacteria promote neurodevelopmental abnormalities in mouse
906–912 (2013). offspring. Nature 549, 528–532 (2017).
39. Ly, L. et al. Impact of mothers’ early life exposure to low or high folate on progeny 73. Sharon, G. et al. Human gut microbiota from autism spectrum disorder promote
outcome and DNA methylation patterns. Environ. Epigenet. 6, dvaa018 (2020). behavioral symptoms in mice. Cell 177, 1600–1618.e17 (2019).
40. Ly, L. et al. Intergenerational impact of paternal lifetime exposures to both folic acid 74. Stoltenberg, C. et al. The Autism Birth Cohort: a paradigm for gene-environment-timing
deficiency and supplementation on reproductive outcomes and imprinted gene research. Mol. Psychiatry 15, 676–680 (2010).
methylation. Mol. Hum. Reprod. 23, 461–477 (2017). 75. Kanner, L. Autistic disturbances of affective contact. Nerv. Child. 2, 217 (1943).

Nature Reviews Neurology | Volume 19 | February 2023 | 118–125 124

Perspective

76. Asperger, H. Die “Autistischen Psychopathen” im Kindesalter. Arch. Psychiatr. Nervenkr. Author contributions
117, 76–136 (1944). All authors contributed to researching data for the article, writing the manuscript, discussions
77. Wing, L. Asperger’s syndrome: a clinical account. Psychol. Med. 11, 115–129 (1981). of its content and review or editing the manuscript before submission.
78. Stein, Z., Susser, M., Saenger, G. & Marolla, F. Famine and Human Development:
The Dutch Hunger Winter of 1944–1945 (Oxford Univ. Press, 1975).
79. Wing, L. & Gould, J. Severe impairments of social interaction and associated abnormalities
Competing interests
The authors declare no competing interests.
in children: epidemiology and classification. J. Autism Dev. Disord. 9, 11–29 (1979).
80. Schopler, E., Rutter, M. & Chess, S. Editorial: Change of journal scope and title. J. Autism
Dev. Disord. 9, 1–10 (1979). Additional information
81. Meryash, D. L., Szymanski, L. S. & Gerald, P. S. Infantile autism associated with the Correspondence should be addressed to W. Ian Lipkin.
fragile-X syndrome. J. Autism Dev. Disord. 12, 295–301 (1982).
82. MRC Vitamin Study Research Group. Prevention of neural tube defects: results of the Peer review information Nature Reviews Neurology thanks Catherine Lord and the other,
Medical Research Council Vitamin Study. Lancet 338, 131–137 (1991). anonymous, reviewers for their contribution to the peer review of this work.
83. Lord, C., Rutter, M. & Le Couteur, A. Autism diagnostic interview-revised: a revised
version of a diagnostic interview for caregivers of individuals with possible pervasive Reprints and permissions information is available at www.nature.com/reprints.
developmental disorders. J. Autism Dev. Disord. 24, 659–685 (1994).
84. Rodier, P. M., Ingram, J. L., Tisdale, B., Nelson, S. & Romano, J. Embryological origin for Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
autism: developmental anomalies of the cranial nerve motor nuclei. J. Comp. Neurol. published maps and institutional affiliations.
370, 247–261 (1996).
85. Rutter, M. et al. Quasi-autistic patterns following severe early global privation. English and
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this
Romanian Adoptees (ERA) Study Team. J. Child. Psychol. Psychiatry 40, 537–549 (1999).
article under a publishing agreement with the author(s) or other rightsholder(s); author self-
86. Lord, C. et al. The autism diagnostic observation schedule-generic: a standard measure
archiving of the accepted manuscript version of this article is solely governed by the terms
of social and communication deficits associated with the spectrum of autism. J. Autism
of such publishing agreement and applicable law.
Dev. Disord. 30, 205–223 (2000).
87. Courchesne, E., Carper, R. & Akshoomoff, N. Evidence of brain overgrowth in the first year
of life in autism. JAMA 290, 337–344 (2003).
88. Reichenberg, A. et al. Advancing paternal age and autism. Arch. Gen. Psychiatry 63, Related links
1026–1032 (2006). SFARI: https://www.sfari.org
89. Sadik, A. et al. Parental inflammatory bowel disease and autism in children. Nat. Med. 28,
1406–1411 (2022). © Springer Nature Limited 2023

Nature Reviews Neurology | Volume 19 | February 2023 | 118–125 125