Elsevier

Annals of Epidemiology

Volume 24, Issue 4, April 2014, Pages 260-266
Annals of Epidemiology

Original article
Population attributable fractions for three perinatal risk factors for autism spectrum disorders, 2002 and 2008 autism and developmental disabilities monitoring network

https://doi.org/10.1016/j.annepidem.2013.12.014Get rights and content

Abstract

Purpose

Numerous studies establish associations between adverse perinatal outcomes/complications and autism spectrum disorder (ASD). There has been little assessment of population attributable fractions (PAFs).

Methods

We estimated average ASD PAFs for preterm birth (PTB), small for gestational age (SGA), and Cesarean delivery (CD) in a U.S. population. Average PAF methodology accounts for risk factor co-occurrence. ASD cases were singleton non-Hispanic white, non-Hispanic black, and Hispanic children born in 1994 (n = 703) or 2000 (n = 1339) who resided in 48 U.S. counties included within eight Autism and Developmental Disabilities Monitoring Network sites. Cases were matched on birth year, sex, and maternal county of residence, race-ethnicity, age, and education to 20 controls from U.S. natality files.

Results

For the 1994 cohort, average PAFs were 4.2%, 0.9%, and 7.9% for PTB, SGA, and CD, respectively. The summary PAF was 13.0% (1.7%–19.5%). For the 2000 cohort, average PAFs were 2.0%, 3.1%, and 6.7% for PTB, SGA, and CD, respectively, with a summary PAF of 11.8% (7.5%–15.9%).

Conclusions

Three perinatal risk factors notably contribute to ASD risk in a U.S. population. Because each factor represents multiple etiologic pathways, PAF estimates are best interpreted as the proportion of ASD attributable to having a suboptimal perinatal environment resulting in PTB, SGA, and/or CD.

Introduction

Autism spectrum disorders (ASDs) are characterized by social and communication impairments and restricted, stereotyped patterns of behavior [1]. ASD prevalence in the U.S. [2], [3], [4] and other populations [5], [6], [7], [8], [9] has increased markedly and is currently estimated at 1%–2% [2], [3], [4], [5], [6], [7], [8], [9]. Numerous genetic factors are implicated in the etiology of ASDs [10] and twin studies suggest high heritability [11]; the composite evidence supports gene-environment interactions. Research on nongenetic risk factors is evolving. Numerous studies document associations between ASD and various adverse perinatal outcomes and complications [12]. Limited studies suggest associations between ASD and more specific maternal exposures such as infections [13], [14], medications [15], [16], [17], and environmental pollutants [18], [19].

For most ASD risk factors, there has been no assessment of population attributable fractions (PAFs). We estimated PAFs for three perinatal risk factors, preterm birth (PTB), small for gestational age (SGA), and Cesarean delivery (CD) among U.S. children included in the Autism and Developmental Disabilities Monitoring (ADDM) Network and compared PAFs from the most recent ADDM surveillance year with those from an earlier time. We chose factors that were both relatively common (≥10% population prevalence) and thus could substantively contribute to the population ASD burden and established as ASD risk factors through multiple studies in a range of populations [12]. Nonetheless, each factor represents a composite of multiple potential underlying etiologic mechanisms. Their PAF estimates are thus best interpreted as the proportion of ASD attributable to having a suboptimal perinatal environment resulting in PTB, SGA, and/or CD.

Section snippets

Study population

The ADDM Network is an ongoing ASD surveillance program among 8-year-old children residing in selected U.S. population–based sites. Thirteen sites participated in the 2002 ADDM surveillance year and 14 participated in the 2008 surveillance year.

For children meeting birth year and residence eligibility criteria, each ADDM site reviews special education records and medical records from providers who conduct developmental evaluations. Records with documentation of an ASD diagnosis or education

Results

The majority of ASD cases from both birth cohorts were male, NHW, and were from three sites: Arizona, Georgia, and Missouri (Table 1). For more than 40%, maternal age at birth was 30 years or older and for more than 50%, maternal education was greater than high school.

Overall, the frequencies of PTB, SGA, and CD among cases were 13%–14%, 11%–12%, and 27%–28%, respectively, for both cohorts (Table 2). The proportion of cases with any of the three was 44% for both cohorts. The proportion of cases

Discussion

The trend of increasing prevalence of diagnosed/identified ASDs has not yet abated [2], [3], [4]. We recently developed a mathematical model to quantify possible contributions of given risk factors to the trend and assessed several factors in depth including PTB, SGA, and CD [12]. We demonstrated that none contributed substantively to the recent ASD prevalence increases. Nonetheless, relative associations between ASD and each factor are well documented [12]. Thus, these factors might still

Acknowledgment

Disclaimer: The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease and Control and Prevention.

References (27)

  • T. Lazoff et al.

    Prevalence of pervasive developmental disorders among children at the English Montreal School Board

    Can J Psychiatry

    (2010)
  • M.T. Roelfsema et al.

    Are autism spectrum conditions more prevalent in an information-technology region? A school-based study of three regions in the Netherlands

    J Autism Dev Disord

    (2012)
  • T.S. Brugha et al.

    Epidemiology of autism spectrum disorders in adults in the community in England

    Arch Gen Psychiatry

    (2011)
  • Cited by (0)

    View full text