Modifiable risk factors for schizophrenia and autism — Shared risk factors impacting on brain development

Neurobiology of Disease 53 (2013) 3–9

Contents lists available at SciVerse ScienceDirect

Neurobiology of Disease journal homepage: www.elsevier.com/locate/ynbdi

Review

Modifiable risk factors for schizophrenia and autism — Shared risk factors impacting on brain development☆ Jess Hamlyn a, Michael Duhig b, c, John McGrath b, d, e, James Scott b, c, d,⁎ a

Gold Coast Hospital, 108 Nerang Street, Southport, QLD 4215, Australia Queensland Centre for Mental Health Research, The Park Centre for Mental Health, Wacol, QLD 4076, Australia The University of Queensland Centre for Clinical Research, Brisbane, QLD 4006, Australia d Discipline of Psychiatry, University of Queensland, St Lucia, QLD 4067, Australia e Queensland Brain Institute, University of Queensland, St Lucia, QLD 4067, Australia b c

a r t i c l e

i n f o

Article history: Received 11 July 2012 Revised 3 October 2012 Accepted 20 October 2012 Available online 2 November 2012 Keywords: Schizophrenia Autism Neurodevelopmental disorder Epidemiology Prevention

a b s t r a c t Schizophrenia and autism are two poorly understood clinical syndromes that differ in age of onset and clinical profile. However, recent genetic and epidemiological research suggests that these two neurodevelopmental disorders share certain risk factors. The aims of this review are to describe modifiable risk factors that have been identified in both disorders, and, where available, collate salient systematic reviews and meta-analyses that have examined shared risk factors. Based on searches of Medline, Embase and PsycINFO, inspection of review articles and expert opinion, we first compiled a set of candidate modifiable risk factors associated with autism. Where available, we next collated systematic-reviews (with or without meta-analyses) related to modifiable risk factors associated with both autism and schizophrenia. We identified three modifiable risk factors that have been examined in systematic reviews for both autism and schizophrenia. Advanced paternal age was reported as a risk factor for schizophrenia in a single meta-analysis and as a risk factor in two meta-analyses for autism. With respect to pregnancy and birth complications, for autism one meta-analysis identified maternal diabetes and bleeding during pregnancy as risks factors for autism whilst a meta-analysis of eight studies identified obstetric complications as a risk factor for schizophrenia. Migrant status was identified as a risk factor for both autism and schizophrenia. Two separate meta-analyses were identified for each disorder. Despite distinct clinical phenotypes, the evidence suggests that at least some non-genetic risk factors are shared between these two syndromes. In particular, exposure to drugs, nutritional excesses or deficiencies and infectious agents lend themselves to public health interventions. Studies are now needed to quantify any increase in risk of either autism or schizophrenia that is associated with these modifiable environmental factors. © 2012 Elsevier Inc. All rights reserved.

Contents Introduction . . . . . . . . . . . . . . . . Literature review . . . . . . . . . . . . . . Selected modifiable risk factors . . . . . . . Nutrition . . . . . . . . . . . . . . . Infection . . . . . . . . . . . . . . . Pregnancy and birth complications . . . Advanced paternal age . . . . . . . . . Migrant status . . . . . . . . . . . . . Clues to other modifiable shared risk factors Conclusion . . . . . . . . . . . . . . . . . Future directions . . . . . . . . . . . References . . . . . . . . . . . . . . . . .

. . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

4 4 4 4 4 5 5 6 7 8 8 8

☆ Author contributions: The study was designed by JM and JS. JH and MD undertook the literature search, and data extraction. All authors contributed to manuscript preparation. ⁎ Corresponding author at: Level 3 UQCCR, Metro North Mental Health, Royal Brisbane and Women's Hospital, Herston, QLD 4029, Australia. Fax: +61 7 3636 1166. E-mail address: [email protected] (J. Scott). Available online on ScienceDirect (www.sciencedirect.com). 0969-9961/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.nbd.2012.10.023

4

J. Hamlyn et al. / Neurobiology of Disease 53 (2013) 3–9

Introduction In recent decades epidemiology has provided important new clues to the neurodevelopmental origins of disorders such as schizophrenia and autism. Through access to larger and better characterised patient and general population samples, associations between potentially modifiable environmental risk factors and complex neurodevelopmental disorders can be examined. In addition to the publication of primary data, the field now has access to comprehensive systematic reviews (with or without subsequent meta-analysis of the data). As the epidemiological profiles of schizophrenia and autism have been enriched, there has been growing interest in the apparent overlap in risk factors between the two disorders. Clearly, there are differences between the two disorders; the age of onset and the impairment of language are examples of phenotypic disparity. Autistic disorders are apparent in early childhood, whereas schizophrenia usually manifests after onset of puberty. In autism, there is a severe and profound impairment in language, whereas schizophrenia is characterised by a wide array of neuropsychological deficits including language (Kalkstein et al., 2010). Importantly however, there is also phenotypic overlap between the two syndromes. Both are widely considered to be neurodevelopmental disorders — in other words, both are thought to be the end result of factors that disrupt early brain development (Bale et al., 2010). Both have an excess prevalence in males. For autism, the male:female risk ratio is 4.1:1 (Newschaffer et al., 2007), whilst for schizophrenia, the risk ratio is 1.4:1 (Aleman et al., 2003; McGrath et al., 2004b). Impaired theory of mind (the capacity to interpret one's own and other persons' mental states) is a core deficit in both syndromes (Brune, 2005) and there is some overlap between the disorders in terms of brain volumetrics (Cheung et al., 2010). The shared and distinct features of the two clinical syndromes have inspired researchers to generate ‘unified’ hypotheses that attempt to synthesise the body of evidence. For example, Crespi and Badcock have proposed that these two disorders exhibit diametric patterns for traits related to social brain development and suggest that differential maternal versus paternal genomic imprinting during brain development underlies the two syndromes (Crespi and Badcock, 2008). Genetic studies have suggested that certain types of copy number variants are shared between these two disorders. Attention has been drawn to a shared genetic diathesis underlying broadly defined neurodevelopmental disorders (including schizophrenia, autism-related disorders and attention deficit disorder) (Kendler, 2010; Mitchell, 2011). There is a growing recognition that, despite the clinical utility of syndromal labels such as schizophrenia and autism (e.g. with respect to prognosis and treatment options), from the perspective of aetiology and pathogenesis, scientific progress may be facilitated by incorporating a broader category of observation (Owen et al., 2011). If the same risk factor is associated with more than one disorder, does this make the association less plausible? Influential articles about causality in risk factor epidemiology have often emphasised the importance of specificity between an exposure and outcome. For example Bradford Hill suggested that if one particular risk factor was associated (specifically) with one particular adverse health outcome, this would add weight to the cumulative evidence that exposure was causally related to that outcome (Hill, 1965). The heritage of this proposal can be readily traced to infectious disease epidemiology and Koch's postulates. However, in recent decades it has become clear that some exposures can be confidently linked to many different outcomes (e.g. tobacco smoking with increased risk of several types of cancers and several types of cardiovascular events). Within the field of psychiatry, we are also familiar with the fact that some exposures can be linked to a wide range of mental health outcomes (e.g. childhood exposure to trauma) (Green et al., 2010; Kessler et al., 2010; Scott et al., 2010). In the spirit of the increased attention to shared genetic mechanisms, we wished to explore environmental risk factors that appear to be shared between autism and schizophrenia. In particular, we are

interested in modifiable risk factors, as these lend themselves to public health intervention, and the primary prevention of these disorders. Literature review In July 2011 we searched electronic resources (PubMed; Embase and Medline) using the search terms “Autis* and risk factor* and environment” to identify systematic reviews of risk factors for autism. We also searched these same electronic databases and the Schizophrenia Research Forum (www.schizophreniaforum.org); and Schizophrenia Research Institute online Library (www.schizophreniaresearch.org.au/ library) using the search terms “schizo* and risk factor* and environment” to identify systematic reviews of risk factors for schizophrenia. From these searches, we identified three areas (pregnancy/birth complications, advanced paternal age and migrant status) where systematic reviews identified risk factors common to both disorders. These were examined in detail and results from these reviews were extracted and summarised in tabular form for each of the two disorders. Other risk factors shown to be associated with one disorder (nutrition and infection) are discussed. Selected modifiable risk factors Nutrition Schizophrenia research has examined prenatal nutritional deprivation as a candidate risk factor via the use of ‘natural experiment’ — the rates of schizophrenia have been examined in cohorts exposed in utero to catastrophic famines (the Dutch Hunger Winter and the famine in China associated with the Cultural Revolution). Individuals who were in utero during the Dutch famine showed an increased risk of schizophrenia and schizophrenia spectrum personality disorders (Susser and Lin, 1992). The risk-increasing effect of maternal starvation on risk of schizophrenia in the offspring has been replicated in those exposed prenatally to a catastrophic famine in China during the Cultural Revolution (St Clair et al., 2005). With respect to the association between risk of schizophrenia and specific maternal micronutrients, homocysteine (a marker of folate of metabolism) was found to be significantly elevated in the third trimester sera from mothers of individuals with schizophrenia (Brown et al., 2007). There is preliminary evidence to suggest that low vitamin D during early life may be associated with schizophrenia (McGrath et al., 2003, 2004a). Similarly, prenatal vitamin D deficiency has also been associated with an increased risk of schizophrenia in offspring (McGrath et al., 2010b). Commentators have suggested that low prenatal vitamin D may also explain epidemiological features of autism (Cannell, 2008; Grant and Soles, 2009). With respect to autism, prenatal nutrition has been proposed as a candidate risk factor. To date, we are unable to identify epidemiological studies examining maternal nutrition and risk of autism in the offspring. Infection Early research related to prenatal exposure to infection and risk of schizophrenia was based on ecological studies (examining the rate of schizophrenia in cohorts who were in utero during epidemics) (McGrath and Castle, 1995). Exposure to infection necessitating hospitalisation in the first or second trimester of pregnancy was associated with an increased risk of autism in offspring (Atladottir et al., 2010). However, access to biobanks has allowed these hypotheses to be tested with stronger, analytic methods. To date, there is evidence to suggest that the risk of schizophrenia is elevated in those with prenatal exposure to influenza (Brown et al., 2004), rubella (Brown et al., 2000), and Toxoplasmosis gondii (Brown et al., 2005; Mortensen et al., 2007). There is mixed evidence for herpes simplex virus type 2 (HSV2) (Brown et al., 2006; Buka et al., 2001). We are not aware of studies that have directly examined markers of prenatal infection

J. Hamlyn et al. / Neurobiology of Disease 53 (2013) 3–9

with respect to risk of autism. The association between infection and neurodevelopmental disorders has stimulated animal research examining prenatal immune activation and brain development (Meyer and Feldon, 2010). It is feasible that the deleterious impact to the maternal brain is mediated through the magnitude of the maternal immune response rather than the direct impact of specific infectious agent (Patterson, 2002, 2009, 2011).

Pregnancy and birth complications Two meta-analyses have examined the association between pregnancy and birth complications (PBCs) and risk of schizophrenia (Cannon et al., 2002; Geddes et al., 1999).Overall, there is robust evidence that PBCs have a significant but modest effect in increasing the risk of later schizophrenia. Based on prospective population-based studies, Cannon et al. (2002) reported that the following specific exposures were associated with increased risk of schizophrenia; antepartum haemorrhage, gestational diabetes, rhesus incompatibility, preeclampsia, low birth weight, congenital malformations, reduced head circumference, uterine atony, asphyxia, and emergency caesarean section. These complications were grouped into three broad categories as illustrated in Table 1. Complications pertaining to the increased risk of schizophrenia in offspring can occur throughout the entire course of pregnancy including childbirth. Gardener et al. (2009) summarised evidence related to pregnancy and birth complications and risk of autism in a comprehensive metaanalysis. Gestational diabetes was associated with a two-fold increased risk of autism in offspring and maternal bleeding during pregnancy was associated with an 81% increased risk. Any maternal medication use during pregnancy was associated with a 46% increased risk of autism. There was a positive association between psychiatric medication use during pregnancy and risk of autism (RR=1.68). Animal models are now being used to further understand the role that exposure to medications may have in increasing the risk of autism. One hypothesis is that

5

medications modulate the role of genes involved in the regulation of neuronal activity (Dufour-Rainfray et al., 2011).

Advanced paternal age Several studies have confirmed the association between advanced paternal age and increased risk of schizophrenia (El-Saadi et al., 2004; Malaspina et al., 2001; Sipos et al., 2004). This finding raises the possibility that accumulation of de novo mutations in paternal sperm with ageing contributes to the risk of schizophrenia. The evidence linking advanced paternal age and schizophrenia has influenced a range of aetiological theories of schizophrenia. For example, the persistence of schizophrenia in the population in spite of reduced fertility could be explained by the transgenerational accumulation of paternally-derived mutations. It is also feasible that epigenetic processes (e.g. chromatin folding, methylation of CpG bases, etc.) may be compromised in the sperm of older fathers, and that these mechanisms may contribute to the increased risk of schizophrenia in their offspring. A meta-analysis by Miller et al. (2011) reported that offspring of fathers aged 30 or older had a significant increased risk of schizophrenia compared to fathers aged 29 or younger. The greatest increased risk was found in fathers who were 50 years or older. There was also a small but significant increase in risk for offspring of fathers aged 25 or younger (compared to those aged 25–29) but this effect was absent in mothers. There is also strong evidence that increasing paternal age is associated with an increased risk of autism. A meta-analysis by Hultman et al. (2010) reported the risk of autism in offspring increased in line with chronological age for fathers aged 30 or older compared to those aged 29 or younger. The association remained significant after controlling for maternal age and other potential risk factors for autism and remained present in families with discordant siblings. As illustrated in Table 2, the greatest risk of autism in offspring was reported in fathers aged 50 or older compared to those 29 or younger. The increased risk of autism in increasing paternal age was also reported by Gardener et

Table 1 Review of obstetric complications, schizophrenia and autism. Author (year)

Number of studies

N

Key findings

Comment

Schizophrenia Cannon et al. (2002)

8

1952 (cases)

Three groups of complications were significantly associated with schizophrenia, however pooled estimates of effect sizes were generally less than two, they are:

• Small effect sizes with odds ratios of less than 2 were observed, which lends itself to controversy and is some way from indicating strong causality. • ‘Nested’ case–control design helps to improve odds ratio outcomes but the lack of independence of individual obstetric complications and possible interactive effects further erode statistical power. • Studies included for meta-analysis employed different combinations of exposures which hindered replication and comparison or pooling of results across studies. • Obstetric complications is a very broad term with 25% to 30% of all birth involving at least one therefore it is difficult to combine these exposures into a single risk factor. • Data regarding the prenatal period are often suboptimal, suggesting that less severe obstetric complications such as prenatal stress are potentially missed.

• Complications of pregnancy; bleeding, preeclampsia, diabetes and rhesus incompatibility. • Abnormal foetal growth and development; low birth weight, congenital malformations and small head circumference. • Complications of delivery; asphyxia, uterine atony, and emergency caesarean section.

Autism Gardener et al. (2009)

RR = relative risk.

40

1,086,271

• Maternal gestational diabetes was associated with a two-fold risk of autism. • Maternal gestational diabetes was also associated with a significant increase of maternal bleeding (81%). • Maternal medication was associated with a 46% increased risk of autism. • Psychiatric medication use during pregnancy reported a significant positive association with autism (RR=1.68).

• Most complications have only been examined in a single study preventing pooling of data and meta-analysis. • The studies were restricted to antenatal complications and did not report on birth complications as a risk factor. • 15 studies analysing medication examined the use of ‘general’ medication during pregnancy, only two related to psychiatric medication.

6

J. Hamlyn et al. / Neurobiology of Disease 53 (2013) 3–9

al. (2009) who showed an increased risk of autism in offspring of fathers aged 30 or older compared to 29 or younger. Relative risk estimates of autism in offspring increased as fathers aged; a 5-year increase in paternal age was associated with a 3.6% increase in risk. In addition, advanced maternal age has also been found to be associated with an increased risk of offspring with autism even after adjusting for paternal age. It is possible that this is explained by higher rates of pregnancy and obstetric complications in older mothers (Sandin et al., 2012). Curiously, the offspring of older mothers do not appear to be at increased risk of schizophrenia (Matheson et al., 2011). The offspring of older fathers have an increased risk of both schizophrenia and autism. In men, spermatogonia undergo cell division every sixteen days, resulting in approximately 200 divisions by the age of 20 years and 660 divisions by the age of 40 years (Drake et al., 1998). Each time the cell divides, the replication of the genome introduces the possibility of copy error mutations, such as point mutations, or larger copy number variants (e.g. deletions, amplifications). Animal experimental studies have recently confirmed that the offspring of older males have

an increased risk of de novo copy number variants (Flatscher-Bader et al., 2011), and these regions contain genes implicated in both schizophrenia and autism. Apart from mutations that change the DNA sequence, epigenetic mechanisms may also be involved (Perrin et al., 2007). For example, epigenetic changes are known to occur in the sperm at an elevated rate with increased age (Oakes et al., 2003, 2007). Migrant status There is now robust evidence showing that certain migrant groups in some countries have an increased risk of schizophrenia (Cantor-Graae and Selten, 2005; McGrath et al., 2004b). In summary, the studies show that both first and second generation migrants have an increased risk of developing schizophrenia, and that the effect is most pronounced in dark-skinned migrants (Cantor-Graae and Selten, 2005). The evidence with respect to migrant status and risk of schizophrenia is robust (Haglund and Kallen, 2011; Lauritsen et al., 2005). A recent large scale meta-analysis by Cantor-Graae and Selten (2005) reported that the

Table 2 Review of advanced paternal age, schizophrenia and autism. Author (year) Schizophrenia Miller et al. (2011)

Autism Hultman et al. (2010)

Number of studies

N

Key findings

12 (6 cohort, 6 case–control) 24,465 (cases) • Significant increase in risk of schizophrenia in offspring of older fathers (≥30) vs those aged 25–29. • Relative risk (RR) in oldest fathers (≥50) was 1.66 (p = .01). • Significant increase in risk for younger fathers (≤25) (RR = 1.08, p = .01) but not mothers (RR = 1.04, p = .28). • PARP was 10% for paternal age ≥ 30 and 5% for paternal age b 25.

12 (6 cohort, 6 case–control) 860 (cases)

Gardener et al. (2009) 40

1,086,271

PAR% = population attributable risk percentage, RR = relative risk.

Comment • Captures all published cohort studies (and one unpublished) of the association between schizophrenia and advanced paternal age. • Able to use the same age class across all studies which enabled a consistent reference group, estimation of 5-year age groups, incorporation of data from several studies not included in previous meta analyses, calculation of PAR% for paternal age and to test the association in younger fathers. • All except one study (Malaspina et al., 2001) were restricted to diagnosis of schizophrenia only. • Effect sizes were calculated as crude risks by paternal age. • Unable to control for potential confounds such as socioeconomic status.

• Combination of birth cohort, family study and meta-analysis allows for greater confidence in observed relationship between advanced paternal age and autism. • Youngest member of cohort was followed up to the age of 10 ensuring coverage during the age range most likely to yield an autism diagnosis. • The use of statistical models allowed adjustment for correlations amongst siblings, thus addressing the potential confound of delayed fatherhood as a reflection of paternal psychopathology. • Information relating to clinical features such as severity of mental retardation and language level was unavailable. • Only four studies included in meta-analysis of • Increased paternal age was reported as paternal age. a significant risk factor for autism with a • Both advanced maternal and paternal ages 5-year increase in paternal age associated are associated with increased risk of autism, with a of 3.6% increase in risk (p = .004). however maternal and paternal ages are • Offspring of fathers aged 30 to 39 years had strongly correlated. an increased RR of autism of 1.24 compared to those aged ≤29. • Most studies did not adjust for the age of the other parent. • Offspring of fathers aged ≥40 years had an increased RR of autism of 1.44 compared to • Of the 4 studies that did adjust for age of the other parent, three found a significant those aged 25–29. association between advanced paternal age and autism. Only one found an association between maternal age and autism suggesting paternal age is the more important risk factor. • Meta analysis reported advancing paternal age was associated with increased risk of autism across studies. • Compared to fathers aged ≤29 years the random pooled estimates of RR for autism in offspring were 1.22 in fathers aged 30–39 years (CI=1.05–1.42), 1.78 in fathers aged 40–49 years (CI=1.52–2.07) and 2.46 for fathers aged 50 and older (CI=2.20–2.76).

J. Hamlyn et al. / Neurobiology of Disease 53 (2013) 3–9

relative risk of developing schizophrenia among first generation migrants was 2.7. Surprisingly, this figure almost doubled for second generation migrants who reported a relative risk of 4.5. Similarly, McGrath et al. (2004b) reported an increased incidence of schizophrenia in migrants compared to native born populations. As described in Table 3, this systematic review found that the median migrant to native-born ratio was 4.6. The association between migrant status and autism has been investigated. A recent systematic review by Dealberto (2011) reported an increase in relative risk of offspring for those who were of maternal foreign birth. The increase was observed across Europe, North America and Australia. Relative risks also varied depending on ethnicity, these figures are illustrated in Table 3. Offspring of dark skinned migrants were at particularly high risk of autism. Furthermore, an earlier meta-analysis by Gardener et al. (2009) also reported an increased

7

risk of autism for children who had mothers born abroad. These findings were particularly strong in Nordic countries, where maternal immigration was associated with a large (58%) increased risk of autism. It is unclear if this strong finding in these Northern countries is a result of the latitude or the quality of the data available in these jurisdictions. Clues to other modifiable shared risk factors Finally, we would like to mention several additional risk factors that were represented in schizophrenia research but not autism, and vice versa. There is robust evidence linking an increased risk of schizophrenia in those who are born and grow up in urban, more densely populated settings. For example, population-based studies from Holland (Marcelis et al., 1998) and Denmark (Mortensen et al., 1999), both found a relative risk of developing schizophrenia

Table 3 Review of migrant status, schizophrenia and autism. Author (year) Schizophrenia Cantor-Graae and Selten (2005)

McGrath et al. (2004a, 2004b)

Autism Dealberto (2011)

Gardener et al. (2009)

Number of studies 18

161

7

40

N

Key findings

Comment

3683

• Mean weighted RR of developing schizophrenia among 1st generation migrants was 2.7 (95% CI = 2.3–3.2). • 2nd generation RR of developing schizophrenia was 4.5 (95% CI = 1.5–13.1). • Combined 1st and 2nd RR of developing schizophrenia RR was 2.9 (95% CI = 2.5–3.4). • Migrants from developing countries yielded significantly greater effect sizes compared to developed countries: RR=3.3 (95% CI=2.8–3.9). • Majority black populations also yielded significantly greater effect sizes compared to white or neither black or white populations: RR=4.8 (95% CI=3.7–6.2).

176,056 (cases)

• Overall migrant groups reported increased incidence of schizophrenia compared to native born populations. • Migrant to native-born rate ratio median (10%–90% quantiles) was 4.6 (1.0–12.8). • Difference in harmonic means between migrants and native-born people was significant for all persons. (pb .001), males (pb .001) and females (p=.01).

• Notion that migrants preferentially receive schizophrenia diagnoses because of cultural misunderstands and/or language difficulties. • Predominance of studies conducted in Europe partly due to feasibility with incident data. • Must consider the differential paths to treatment as most studies of migrants are based on rates of treated cases, as treated cases are based on higher social class, minority groups may be underrepresented in samples. • Assignment to categories in two analyses (4 and 5) is based on birth place, levels of economic development and skin colour of the majority of population do not necessarily reflect an individual's income or physical appearance. • 24 migrant studies identified in systematic review, all of which were from the United Kingdom/Europe.

7169 (cases)

• Four studies found an increase of RR of Autism for those who were of a maternal foreign birth ranging from .4 for Californian children from Mexican mothers to 2.9 for Swedish children from mothers outside Europe and North America (ARR of 0.6 and 3.0 respectively). • The only Australian study found that children born in New South Wales from mothers outside of Australia had a RR of 1.5 (ARR=1.4) with highest risk mothers from East Asia. Three studies reported varying rates of RR of autism in African American children born to African American mothers ranging from .5 (U.S. Georgia and Atlanta, without MR) to 2.0 (same study with MR) (ARR was 1.5 and 3.6 respectively).

1,086,271

• Maternal birth abroad was marginally associated with a 28% increased risk of autism (p = .06). • Three studies using only Nordic countries found maternal birth abroad was significantly associated with a 58% increased risk of autism.

• This study attempts to overcome inherent methodological issues of the Autism and immigrant status/ethnic origin relationship by using only large samples and employing multivariate analyses to control for potential confounds. • Clear definitions of immigrant status and ethnic origin were used but selected studies on immigrant status reported varying effects between continent of birth and RR of autism, this may also be evident for ethnic origin. • Use of ethnic origin as an indicator for skin pigmentation. • Case ascertainment may be biased as access to educational, medical and social services may differ according to immigrant status, ethnic origin or social class. It is impossible to know if these biases underestimated or overestimated association rates. • Autism severity could be biased due to differential referral as children of ethnic backgrounds or immigrant communities may only be referred when symptoms are severe. • Only five studies included for meta-analysis. • Only studies from Nordic countries found a significant association. • Definition of ‘abroad’ varied between studies.

PAR% = population attributable risk percentage, RR = relative risk, ARR = adjusted relative risk, U.S. = United States, MR = mental retardation.

8

J. Hamlyn et al. / Neurobiology of Disease 53 (2013) 3–9

when born in the city vs. being born in the country of about 2.4. The evidence suggests that urbanicity of place of birth is a proxy marker for a yet-to-be-identified risk-modifying variable operating at or before birth. However, because most people who are born in a city are also brought up there, it is difficult to disentangle pre- and perinatal effects from those operating later in childhood. With respect to autism, some epidemiological studies have identified an increased risk of autism in urban setting (Chen et al., 2008; Lauritsen et al., 2005; Rosenberg et al., 2009), however it is thought that differential access to care and diagnosis may explain at least part of this gradient. Autism research has also examined the role of food additives, thimerosal and/or mercury exposure in vaccines, prenatal ultrasound (Grether et al., 2010), and a range of environmental toxins (Dufour-Rainfray et al., 2011). However, the evidence for these exposures increasing the risk of autism has not been convincing. Conclusion In keeping with the evidence of shared genetic factors associated with both schizophrenia and autism, we found evidence from systematic reviews and meta-analyses suggesting that these two disorders also share non-genetic risk factors. Mindful that not all areas of the two disorders had sufficient primary data to justify systematic reviews and meta-analysis, it is striking to find several candidate exposures that are associated with an increased risk of both schizophrenia and autism. Advanced paternal age has been identified as a risk factor for both schizophrenia and autism in the offspring. It is plausible that this arises from a combination of both genetic mutations arising from increased risk of de novo copy number variants (Flatscher-Bader et al., 2011) and epigenetic changes in the sperm of older males (Oakes et al., 2003, 2007). Obstetric complications are plausibly associated with an increased risk of both schizophrenia and autism. The prenatal period is critical for the orderly cascade of brain development. It is feasible that whilst some types of pregnancy and birth complications are shared between the two disorders, perhaps the timing of the insult is more critical with respect to the subsequent phenotype. Interactions with genetic susceptibility factors may also play a role, and future studies will hopefully be able to explore gene by environment interactions. It is curious that migrant status is associated with both schizophrenia and autism. Some researchers have suggested that belonging to an obviously different ethnic minority group (e.g. dark skin) may contribute to psychosocial stress and ‘social defeat’ (Cantor-Graae, 2007; Selten and Cantor-Graae, 2005). However, it is not clear how these stressors could impact on infants who usually experience their social milieu from the immediate family rather than the wider community. The impact of low vitamin D may warrant further exploration (Currenti, 2010; Dealberto, 2011; Eyles, 2010; Fernell et al., 2010; Grant and Soles, 2009; Humble et al., 2010; Kinney et al., 2010; McGrath et al., 2010a). Future directions There is evidence of shared aetiological mechanisms underpinning other neuropsychiatric disorders. For example, bidirectional evidence is emerging between epilepsy and affective disorders which is thought to be mediated by dysregulation of the hypothalamic–pituitary axis (Kanner, 2011). We predict that with advances in neuroimaging, genetics and epidemiology, there will be increasing awareness of the overlap that exists between discrete neuropsychiatric disorders. Although this may pose challenges to the current diagnostic systems, it will provide opportunities to better understand the aetiopathogenesis of these neuropsychiatric syndromes. Epidemiological research has proven itself a good source for generating candidate exposures, but left to its own resources this field is prone to cycles of uninformative replications. One

strategy to optimise discovery is to build strong links between epidemiology and neuroscience. This is especially important for exploring the biological plausibility of candidate risk factors derived from ecological studies (McGrath, 2007). Cross-disciplinary projects between epidemiology and neuroscience may help us understand the shared pathways underpinning both schizophrenia and autism. The ultimate goal must be to find ways to reduce exposure to modifiable risk factors thus reducing the risk of these disorders. In particular, exposure to drugs, nutritional excesses or deficiencies and infectious agents lend themselves public health interventions. Furthermore, elucidating the mechanisms that underpin the deviations in the neurodevelopmental trajectory may allow interventions in infants who through exposure to identified factors are at increased risk of developing schizophrenia or autism.

References Aleman, A., Kahn, R.S., Selten, J.P., 2003. Sex differences in the risk of schizophrenia: evidence from meta-analysis. Arch. Gen. Psychiatry 60 (6), 565–571. Atladottir, H.O., Thorsen, P., Ostergaard, L., Schendel, D.E., Lemcke, S., Abdallah, M., et al., 2010. Maternal infection requiring hospitalization during pregnancy and autism spectrum disorders. J. Autism Dev. Disord. 40 (12), 1423–1430. Bale, T.L., Baram, T.Z., Brown, A.S., Goldstein, J.M., Insel, T.R., McCarthy, M.M., et al., 2010. Early life programming and neurodevelopmental disorders. Biol. Psychiatry 68 (4), 314–319. Brown, A.S., Cohen, P., Greenwald, S., Susser, E., 2000. Nonaffective psychosis after prenatal exposure to rubella. Am. J. Psychiatry 157 (3), 438–443. Brown, A.S., Begg, M.D., Gravenstein, S., Schaefer, C.A., Wyatt, R.J., Bresnahan, M., et al., 2004. Serologic evidence of prenatal influenza in the etiology of schizophrenia. Arch. Gen. Psychiatry 61 (8), 774–780. Brown, A.S., Schaefer, C.A., Quesenberry Jr., C.P., Liu, L., Babulas, V.P., Susser, E.S., 2005. Maternal exposure to toxoplasmosis and risk of schizophrenia in adult offspring. Am. J. Psychiatry 162 (4), 767–773. Brown, A.S., Schaefer, C.A., Quesenberry Jr., C.P., Shen, L., Susser, E.S., 2006. No evidence of relation between maternal exposure to herpes simplex virus type 2 and risk of schizophrenia? Am. J. Psychiatry 163 (12), 2178–2180. Brown, G.W., Craig, T.K., Harris, T.O., Handley, R.V., Harvey, A.L., 2007. Validity of retrospective measures of early maltreatment and depressive episodes using the Childhood Experience of Care and Abuse (CECA) instrument - A life-course study of adult chronic depression - 2. J. Affect. Disord. 103 (1–3), 217–224. http://dx.doi.org/10.1016/ j.jad.2007.06.003 (Comment Research Support, Non-U.S. Gov't). Brune, M., 2005. “Theory of mind” in schizophrenia: a review of the literature. Schizophr. Bull. 31 (1), 21–42. Buka, S.L., Tsuang, M.T., Torrey, E.F., Klebanoff, M.A., Bernstein, D., Yolken, R.H., 2001. Maternal infections and subsequent psychosis among offspring. Arch. Gen. Psychiatry 58, 1032–1037. Cannell, J.J., 2008. Autism and vitamin D. Med. Hypotheses 70 (4), 750–759. Cannon, M., Jones, P.B., Murray, R.M., 2002. Obstetric complications and schizophrenia: historical and meta-analytic review. Am. J. Psychiatry 159 (7), 1080–1092. Cantor-Graae, E., 2007. The contribution of social factors to the development of schizophrenia: a review of recent findings. Can. J. Psychiatry 52 (5), 277–286. Cantor-Graae, E., Selten, J.P., 2005. Schizophrenia and migration: a meta-analysis and review. Am. J. Psychiatry 162 (1), 12–24. Chen, C.Y., Liu, C.Y., Su, W.C., Huang, S.L., Lin, K.M., 2008. Urbanicity-related variation in help-seeking and services utilization among preschool-age children with autism in Taiwan. J. Autism Dev. Disord. 38 (3), 489–497. Cheung, C., Yu, K., Fung, G., Leung, M., Wong, C., Li, Q., et al., 2010. Autistic disorders and schizophrenia: related or remote? An anatomical likelihood estimation. PLoS One 5 (8), e12233. Crespi, B., Badcock, C., 2008. Psychosis and autism as diametrical disorders of the social brain. Behav. Brain Sci. 31 (3), 241–261 (discussion 61–320). Currenti, S.A., 2010. Understanding and determining the etiology of autism. Cell. Mol. Neurobiol. 30 (2), 161–171. Dealberto, M.J., 2011. Prevalence of autism according to maternal immigrant status and ethnic origin. Acta Psychiatr. Scand. 123 (5), 339–348. Drake, J.W., Charlesworth, B., Charlesworth, D., Crow, J.F., 1998. Rates of spontaneous mutation. Genetics 148 (4), 1667–1686. Dufour-Rainfray, D., Vourc'h, P., Tourlet, S., Guilloteau, D., Chalon, S., Andres, C.R., 2011. Fetal exposure to teratogens: evidence of genes involved in autism. Neurosci. Biobehav. Rev. 35 (5), 1254–1265. El-Saadi, O., Pedersen, C.B., McNeil, T.F., Saha, S., Welham, J., O'Callaghan, E., et al., 2004. Paternal and maternal age as risk factors for psychosis: findings from Denmark, Sweden and Australia. Schizophr. Res. 67 (2–3), 227–236. Eyles, D.W., 2010. Vitamin D and autism: does skin colour modify risk? Acta Paediatr. 99 (5), 645–647. Fernell, E., Barnevik-Olsson, M., Bagenholm, G., Gillberg, C., Gustafsson, S., Saaf, M., 2010. Serum levels of 25-hydroxyvitamin D in mothers of Swedish and of Somali origin who have children with and without autism. Acta Paediatr. 99 (5), 743–747. Flatscher-Bader, T., Foldi, C.J., Chong, S., Whitelaw, E., Moser, R.J., Burne, T.H.J., et al., 2011. Increased de novo copy number variants in the offspring of older males. Transl. Psychiatry 1 (e34).

J. Hamlyn et al. / Neurobiology of Disease 53 (2013) 3–9 Gardener, H., Spiegelman, D., Buka, S.L., 2009. Prenatal risk factors for autism: comprehensive meta-analysis. Br. J. Psychiatry 195 (1), 7–14. Geddes, J.R., Verdoux, H., Takei, N., Lawrie, S.M., Bovet, P., Eagles, J.M., et al., 1999. Schizophrenia and complications of pregnancy and labor: an individual patient data meta-analysis. Schizophr. Bull. 25 (3), 413–423. Grant, W.B., Soles, C.M., 2009. Epidemiologic evidence supporting the role of maternal vitamin D deficiency as a risk factor for the development of infantile autism. Dermato-Endocrinology 1 (4), 223–228. Green, J.G., McLaughlin, K.A., Berglund, P.A., Gruber, M.J., Sampson, N.A., Zaslavsky, A.M., et al., 2010. Childhood adversities and adult psychiatric disorders in the national comorbidity survey replication I: associations with first onset of DSM-IV disorders. Arch. Gen. Psychiatry 67 (2), 113–123. Grether, J.K., Li, S.X., Yoshida, C.K., Croen, L.A., 2010. Antenatal ultrasound and risk of autism spectrum disorders. J. Autism Dev. Disord. 40 (2), 238–245. Haglund, N.G., Kallen, K.B., 2011. Risk factors for autism and Asperger syndrome. Perinatal factors and migration. Autism 15 (2), 163–183. Hill, A.B., 1965. The environment and disease: association or causation? Proc. R. Soc. Med. 58, 295–300. Hultman, C.M., Sandin, S., Levine, S.Z., Lichtenstein, P., Reichenberg, A., 2010. Advancing paternal age and risk of autism: new evidence from a population-based study and a meta-analysis of epidemiological studies. Mol. Psychiatry 16 (12), 1203–1212. Humble, M.B., Gustafsson, S., Bejerot, S., 2010. Low serum levels of 25-hydroxyvitamin D (25-OHD) among psychiatric out-patients in Sweden: relations with season, age, ethnic origin and psychiatric diagnosis. J. Steroid Biochem. Mol. Biol. Kalkstein, S., Hurford, I., Gur, R.C., 2010. Neurocognition in schizophrenia. Curr. Top. Behav. Neurosci. 4, 373–390. Kanner, A.M., 2011. Depression and epilepsy: a bidirectional relation? Epilepsia 52 (Suppl. 1), 21–27. Kendler, K.S., 2010. Advances in our understanding of genetic risk factors for autism spectrum disorders. Am. J. Psychiatry 167 (11), 1291–1293. Kessler, R.C., McLaughlin, K.A., Green, J.G., Gruber, M.J., Sampson, N.A., Zaslavsky, A.M., et al., 2010. Childhood adversities and adult psychopathology in the WHO World Mental Health Surveys. Br. J. Psychiatry 197 (5), 378–385. Kinney, D.K., Barch, D.H., Chayka, B., Napoleon, S., Munir, K.M., 2010. Environmental risk factors for autism: do they help cause de novo genetic mutations that contribute to the disorder? Med. Hypotheses 74 (1), 102–106. Lauritsen, M.B., Pedersen, C.B., Mortensen, P.B., 2005. Effects of familial risk factors and place of birth on the risk of autism: a nationwide register-based study. J. Child Psychol. Psychiatry 46 (9), 963–971. Malaspina, D., Harlap, S., Fennig, S., Heiman, D., Nahon, D., Feldman, D., et al., 2001. Advancing paternal age and the risk of schizophrenia. Arch. Gen. Psychiatry 58 (4), 361–367. Marcelis, M., Navarro-Mateu, F., Murray, R., Selten, J.-P., van Os, J., 1998. Urbanization and psychosis: a study of 1942–1978 birth cohorts in The Netherlands. Psychol. Med. 28, 871–879. Matheson, S.L., Shepherd, A.M., Laurens, K.R., Carr, V.J., 2011. A systematic meta-review grading the evidence for non-genetic risk factors and putative antecedents of schizophrenia. Schizophr. Res. 133 (1–3), 133–142. McGrath, J.J., 2007. The surprisingly rich contours of schizophrenia epidemiology. Arch. Gen. Psychiatry 64 (1), 14–16. McGrath, J., Castle, D., 1995. Does influenza cause schizophrenia? A five year review. Aust. N. Z. J. Psychiatry 29 (1), 23–31. McGrath, J., Eyles, D., Mowry, B., Yolken, R., Buka, S.L., 2003. Low maternal vitamin D as a risk factor for schizophrenia: a pilot study using banked sera. Schizophr. Res. 63, 73–78. McGrath, J., Saha, S., Welham, J., El Saadi, O., MacCauley, C., Chant, D., 2004a. A systematic review of the incidence of schizophrenia: the distribution of rates and the influence of sex, urbanicity, migrant status and methodology. BMC Med. 2 (1), 13.

9

McGrath, J., Saari, K., Hakko, H., Jokelainen, J., Jones, P., Järvelin, M.-R., et al., 2004b. Vitamin D supplementation during the first year of life and risk of schizophrenia: a Finnish birth-cohort study. Schizophr. Res. 67 (2,3), 237–245. McGrath, J.J., Burne, T.H., Feron, F., Mackay-Sim, A., Eyles, D.W., 2010a. Developmental vitamin D deficiency and risk of schizophrenia: a 10-year update. Schizophr. Bull. 36 (6), 1073–1078. McGrath, J.J., Eyles, D.W., Pedersen, C.B., Anderson, C., Ko, P., Burne, T.H., et al., 2010b. Neonatal vitamin D status and risk of schizophrenia: a population-based case–control study. Arch. Gen. Psychiatry 67 (9), 889–894. Meyer, U., Feldon, J., 2010. Epidemiology-driven neurodevelopmental animal models of schizophrenia. Prog. Neurobiol. 90 (3), 285–326. Miller, B., Messias, E., Miettunen, J., Alaraisanen, A., Jarvelin, M.R., Koponen, H., et al., 2011. Meta-analysis of paternal age and schizophrenia risk in male versus female offspring. Schizophr. Bull. 37 (5), 1039–1047. Mitchell, K.J., 2011. The genetics of neurodevelopmental disease. Curr. Opin. Neurobiol. 21 (1), 197–203. Mortensen, P.B., Pedersen, C.B., Westergaard, T., Wohlfahrt, J., Ewald, H., Mors, O., et al., 1999. Effects of family history and place and season of birth on the risk of schizophrenia. N. Engl. J. Med. 340, 603–608. Mortensen, P.B., Norgaard-Pedersen, B., Waltoft, B.L., Sorensen, T.L., Hougaard, D., Yolken, R.H., 2007. Early infections of Toxoplasma gondii and the later development of schizophrenia. Schizophr. Bull. 33 (3), 741–744. Newschaffer, C.J., Croen, L.A., Daniels, J., Giarelli, E., Grether, J.K., Levy, S.E., et al., 2007. The epidemiology of autism spectrum disorders. Annu. Rev. Public Health 28, 235–258. Oakes, C.C., Smiraglia, D.J., Plass, C., Trasler, J.M., Robaire, B., 2003. Aging results in hypermethylation of ribosomal DNA in sperm and liver of male rats. Proc. Natl. Acad. Sci. U S A 100 (4), 1775–1780. http://dx.doi.org/10.1124/jpet.107.121699. Oakes, C.C., Kelly, T.L., Robaire, B., Trasler, J.M., 2007. Adverse effects of 5-aza-2'deoxycytidine on spermatogenesis include reduced sperm function and selective inhibition of de novo DNA methylation. J. Pharmacol. Exp. Ther. 322 (3), 1171–1180. http://dx.doi.org/10.1124/jpet.107.121699. Owen, M.J., O'Donovan, M.C., Thapar, A., Craddock, N., 2011. Neurodevelopmental hypothesis of schizophrenia. Br. J. Psychiatry 198 (3), 173–175. Patterson, P.H., 2002. Maternal infection: window on neuroimmune interactions in fetal brain development and mental illness. Curr. Opin. Neurobiol. 12 (1), 115–118. Patterson, P.H., 2009. Immune involvement in schizophrenia and autism: etiology, pathology and animal models. Behav. Brain Res. 204 (2), 313–321. Patterson, P.H., 2011. Maternal infection and immune involvement in autism. Trends Mol. Med. 17 (7), 389–394. Perrin, M.C., Brown, A.S., Malaspina, D., 2007. Aberrant epigenetic regulation could explain the relationship of paternal age to schizophrenia. Schizophr. Bull. 33 (6), 1270–1273. Rosenberg, R.E., Daniels, A.M., Law, J.K., Law, P.A., Kaufmann, W.E., 2009. Trends in autism spectrum disorder diagnoses: 1994–2007. J. Autism Dev. Disord. 39 (8), 1099–1111. Sandin, S., Hultman, C.M., Kolevzon, A., Gross, R., MacCabe, J.H., Reichenberg, A., 2012. Advancing maternal age is associated with increasing risk for autism: a review and meta-analysis. J. Am. Acad. Child Adolesc. Psychiatry 51 (5), 477–486 (e1). Scott, J., Varghese, D., McGrath, J., 2010. As the twig is bent, the tree inclines: adult mental health consequences of childhood adversity. Arch. Gen. Psychiatry 67 (2), 111–112. Selten, J.P., Cantor-Graae, E., 2005. Social defeat: risk factor for schizophrenia? Br. J. Psychiatry 187, 101–102. Sipos, A., Rasmussen, F., Harrison, G., Tynelius, P., Lewis, G., Leon, D.A., et al., 2004. Paternal age and schizophrenia: a population based cohort study. BMJ 329 (7474), 1070. St Clair, D., Xu, M., Wang, P., Yu, Y., Fang, Y., Zhang, F., et al., 2005. Rates of adult schizophrenia following prenatal exposure to the Chinese famine of 1959–1961. JAMA 294 (5), 557–562. Susser, E.S., Lin, S.P., 1992. Schizophrenia after prenatal exposure to the Dutch hunger winter of 1944–1945. Arch. Gen. Psychiatry 49, 938–988.