Objective. This study determined the prevalence of autism for a defined community, Brick Township, New Jersey, using current diagnostic and epidemiologic methods.
Methods. The target population was children who were 3 to 10 years of age in 1998, who were residents of Brick Township at any point during that year, and who had an autism spectrum disorder. Autism spectrum disorder was defined as autistic disorder, pervasive developmental disorder-not otherwise specified (PDD-NOS), and Asperger disorder. The study used 4 sources for active case finding: special education records, records from local clinicians providing diagnosis or treatment for developmental or behavioral disabilities, lists of children from community parent groups, and families who volunteered for participation in the study in response to media attention. The autism diagnosis was verified (or ruled out) for 71% of the children through clinical assessment. The assessment included medical and developmental history, physical and neurologic evaluation, assessment of intellectual and behavioral functioning, and administration of the Autism Diagnostic Observation Schedule—Generic.
Results. The prevalence of all autism spectrum disorders combined was 6.7 cases per 1000 children. The prevalence for children whose condition met full diagnostic criteria for autistic disorder was 4.0 cases per 1000 children, and the prevalence for PDD-NOS and Asperger disorder was 2.7 cases per 1000 children. Characteristics of children with autism in this study were similar to those in previous studies of autism.
Conclusions. The prevalence of autism in Brick Township seems to be higher than that in other studies, particularly studies conducted in the United States, but within the range of a few recent studies in smaller populations that used more thorough case-finding methods.
Autism is a serious, lifelong developmental disability characterized by significant impairments in reciprocal social interactions and communication skills, as well as a restricted/repetitive pattern of interests and/or behaviors.1 Despite concerns about the number of children in the United States who might have autism, little information exists about the current prevalence of the disorder. Information about the rate of autism in this country must be gleaned from a few US population-based studies conducted 10 to 20 years ago,2–4studies conducted in other countries,5,,6 and data concerning services used by individuals with autism.7 This article reports results from a population-based prevalence study of autism in Brick Township, New Jersey. The study was undertaken because of concern from community members about a seemingly large number of children with autism residing in the town relative to the numbers expected and concern about potential links to environmental factors.8
Autism is a spectrum of neurobehavioral disorders that affects individuals from all ethnic and socioeconomic backgrounds.9 The diagnosis is based on the child's developmental and medical history as well as observations of his or her social, communicative, and play behaviors. Parents may suspect autism in a child as young as 12 to 15 months10,,11 but formal diagnosis is made most reliably in the preschool period or upon starting school.11,,12 Currently, there are no biological markers or standardized psychological measures available for making the diagnosis, although recent findings offer possibilities in this area.13–15 The Diagnostic and Statistical Manual, Fourth Edition (DSM-IV) published by the American Psychiatric Association1 provides diagnostic criteria for autism under the rubric of pervasive developmental disorders, which includes autistic disorder, Asperger disorder, and pervasive developmental disorder-not otherwise specified (PDD-NOS).FNa Collectively, these disorders are referred to as autism spectrum disorders (ASD). Distinctions between the specific diagnoses are based on the number and severity of the child's behaviors. Thus, clinicians make subjective decisions about the quality of some behaviors when applying theDSM-IV criteria, which can lead to inconsistencies in use of the diagnostic labels. All of these diagnostic characteristics of autism affect efforts to establish baseline prevalence rates.16
Population-based studies of autism prevalence in the United States were conducted in the 1980s or early 1990s and used earlier diagnostic criteria. Burd et al2 reported a prevalence in North Dakota of 0.33 cases per 1000 children for autistic disorder and estimated the prevalence at 1.1 cases per 1000 children for a more broadly defined category of autism.FNb A second study, conducted in Utah, reported a prevalence rate for autism of 0.4 cases per 1000 children/young adults.3 Finally, a study in Arkansas that used expanded birth defects surveillance to determine the prevalence of developmental disabilities in children under 4 years old also found a prevalence of 0.4 cases per 1000 children.4These rates of autism in the United States differ from most of the rates reported from more recent studies conducted in other countries. Reviews of these studies suggest that a conservative estimate of the prevalence of autistic disorder may be about 1 per 1000.5–6 However, a few of these studies have reported markedly higher rates ranging from 2.1 to 6.0 per 1000 children with autism.9,,17,18 In addition, the earlier reported US rates are not consistent with recent reports of individuals receiving services for autism7 or anecdotal reports from parents and clinicians. These discrepancies have led to speculation that the US studies underestimated the prevalence of autism or that the rate has increased over the past 2 decades.5,,6
The present study was conducted to determine the prevalence of ASD among children in a defined US population using current diagnostic and epidemiologic methods. This study used an active case finding method among all children receiving special education services and children examined by clinicians in the community who provide evaluation and treatment for children with developmental disabilities. In addition, we invited families to participate in clinical assessments to verify each child's diagnosis and obtain information about characteristics of this population of children with autism.
The target population for this study was children aged 3 through 10 years whose parents resided in Brick Township, New Jersey, at any time during the 1998 calendar year. Brick Township, a primarily middle class community19 that had a population of about 76 00019 in 1998, is located about 50 miles north of Atlantic City in eastern New Jersey.
ASD was defined as autistic disorder, PDD-NOS, or Asperger disorder. Children with Rett disorder or childhood disintegrative disorder were excluded. Definition and diagnostic criteria for each of these disorders were based on the DSM-IV.1
In developing procedures to identify potential cases, we sought methods to capture children within the autism spectrum across a broad range of functioning levels. In addition, we made great effort to identify all potential data sources to make case ascertainment as complete as possible. To accomplish these goals, we reviewed existing educational and service provider records. Potential cases were identified through 4 sources: 1) school records, 2) records of private clinicians (eg, pediatric neurologists or developmental pediatricians), 3) lists of children with autism maintained by local parent groups, and 4) parents who self-referred their children directly to investigators after becoming aware of the study through media coverage.
The primary source for potential case children was special education records maintained by Brick Township Public Schools. A developmental psychologist (J.B.) independently reviewed files of all children who received special education services from Brick Township schools in 1998. This review included records of children with any type of special education classification (eg, speech/language disorder), children receiving private or out-of-district services (paid for by Brick Township), and children who were evaluated but determined not to qualify for services. Children were identified as potentially having autism if their special education classification or eligibility was autism or their record described behaviors consistent with 1 or more of the DSM-IV diagnostic criteria for one of the ASDs. Greater weight was given to behaviors more specific to individuals with autism (eg, poor eye contact, echolalia) than nonspecific behaviors (eg, language delay). In addition to the public school system, private schools in the area that provide education to children with developmental disabilities (including autism) were queried about children who might qualify.
To identify higher functioning children or those not receiving special education services, potential case children were sought by contacting local clinicians, private schools, and other sources. A list was developed in collaboration with town officials, parent groups, and phone listings that included 15 private schools, 4 psychiatric facilities that provide inpatient and/or outpatient services for children with autism and other psychiatric disorders, 3 child psychiatrists, 4 pediatric neurology practices, and 1 general pediatrician. Investigators contacted these potential sources to determine whether they provided services to any children from Brick Township. Several schools provided services for Brick Township children with autism, although these children already had been identified from the special education files maintained by the Brick Township schools. The psychiatric facilities all reported that in 1998 they served no children with autism from Brick Township. Of the clinicians, 3 pediatric neurologists were identified who provided diagnosis and treatment to children from Brick Township in 1998 and who allowed access to their records. The fourth pediatric neurologist reported having seen no children with an ASD from Brick Township. A review of the files from the general pediatrician indicated that we had already identified all children with autism in that practice. We were unable to obtain any information from the 3 child psychiatrists. Although it is unknown whether additional children would have been identified from the 3 child psychiatrists, only 1 of these clinicians was named in the records reviewed at the school and other sources. To the best of our knowledge, we contacted all potential sources within reasonable distance of Brick Township. However, a small number of children diagnosed at large medical centers outside the area (eg, Children's Hospital of Philadelphia) and who were not receiving services from the school system or local clinicians may not have been identified.
The autism diagnosis was verified by clinical assessment. Families that had potential case children who had been identified from school records, through parent groups, or from self-referral were invited to participate in a comprehensive clinical assessment. Assessments were conducted in 2 parts by the 2 lead authors who have extensive experience in the evaluation and diagnosis of children with autism. Clinical assessment procedures are listed in Table 1.20–24 All instruments were administered according to standard procedures.20–24The clinical assessment included medical and developmental history, a physical examination, administration of the Autism Diagnostic Observation Schedule-Generic (ADOS-G),21 and evaluation of intellectual and behavioral functioning. Detailed medical and developmental histories were obtained through parental questionnaires and interviews by the clinicians. Information obtained through questionnaires included family history of developmental disabilities, prenatal conditions, birth history, early developmental milestones, medical events in the child's life, findings from previous evaluations, and behaviors and impairments associated with autism. In the first part of the assessment, the developmental pediatrician (A.M.) conducted a brief physical examination of the child, including anthropometric measurements, assessment of any dysmorphic features, and a basic neurologic examination. To facilitate diagnosis, the ADOS-G20–22 was administered by one of the authors (A.M.) and videotaped. The ADOS-G is a semistructured interview with children adapted for their language level. Specific social situations are designed so that the examiner directly observes social and communicative functioning of the child. Behaviors and skills are scored for presence and severity based on standardized coding. An algorithm derived from the behavioral scores assists in determining whether the child is functioning within the autism spectrum and provides thresholds for PDD-NOS and autistic disorder. An algorithm for diagnosis of Asperger disorder is not included in the ADOS-G. Researchers who use the ADOS-G receive specific training on administration and behavior scoring of the instrument and establish reliable and reproducible coding competence with the ADOS-G authors. This requirement ensures that the autism diagnostic labels used in this study are comparable with other studies using the ADOS-G. The ADOS-G is often used in conjunction with the Autism Diagnostic Interview-Revised (ADI-R).22 The ADI-R was not used in the present study because of a desire to limit the length of the evaluation for participating families and it was thought to be important to use a dignostic tool that directly assessed the children rather than one which relied on parent's perception or interpretation of behaviors. In the second part of the assessment, the developmental psychologist assessed intellectual and behavioral functioning through standardized developmental tests (Table 1).23–25 Final diagnosis and case status was determined by consensus of the 2 clinicians through consideration of all available developmental and medical information in conjunction with behaviors observed during both parts of the assessment, with particular emphasis on observation and scoring of the ADOS-G.
For children who were potential cases whose families declined participation in the clinical assessment, could not be contacted, or were identified solely through clinicians' offices, the diagnosis of ASD was evaluated through review of records and consensus of the 2 clinicians. For these children, school and/or medical records were examined and behavioral descriptions meeting the DSM-IVdiagnostic criteria for an ASD were noted according to a systematic coding scheme (available from the first author). For example, repeatedly lining up objects was always considered fulfilling the criteria concerning nonfunctional routines. The number and pattern of the noted behaviors were used to determine if the child's condition met the diagnostic criteria for one of the ASD and thereby met our case definition. Children whose condition met the case definition for an ASD were included as cases. All other children were considered noncases.
Estimation of Denominator for Prevalence Rates
The denominator for calculation of the prevalence rate was estimated by adjusting the 1990 census count for children aged 3 to 10 years (N = 7117) by a 25% inflation factor. This inflation factor was equivalent to the increase reported by Brick Township for the student population for grades K through 5 during the school years 1989 to 1990 and 1998 to 1999. Using this inflation factor, the estimated number of children aged 3 to 10 years in Brick Township in 1998 was 8896 (4364 males, 4532 females). Ninety-five percent confidence intervals (CIs) for these prevalence rates were calculated.
Results are presented in 2 sections. In the first section, findings are described for all children identified with ASD. In the second section, we briefly present medical and developmental findings for children who participated in the clinical assessments.
Results From All Children
Seventy-five children were identified as possibly having ASD through the case-ascertainment process. Most (83%) were identified at >1 source (eg, school and clinicians' offices). Of the 75 possible case-children, 53 (71%) participated in the clinical assessment, and 22 were evaluated by clinical review of diagnostic information included in school and/or clinician records. Sixty of the 75 potential case-children met the DSM-IV criteria for one of the ASDs on the clinical assessment or record review. Fifteen of the identified children did not meet criteria for any ASD and were excluded from additional analysis. These children had a number of other developmental disorders, such as attention-deficit/hyperactivity disorder, mental retardation, or a speech disorder, that may have been responsible for their behaviors. Thirty-six (60%) of the 60 children with ASD met the full criteria for autistic disorder (30 by clinical assessment and 6 by record review).
The overall rate for autism spectrum disorders was 6.7 cases per 1000 children aged 3 through 10 years (95% CI: 5.1–8.7; Table 2). Forty-four (73%) of the 60 children were boys, yielding a ratio of 2.7 males for each female for the entire autistic spectrum. The prevalence rate of autistic disorder was 4.0 cases per 1000 children aged 3 through 10 years (95% CI: 2.8–5.6). The male-to-female ratio for children with autistic disorder was 2.2 males for each female. For children whose condition did not meet the criteria for autistic disorder but were functioning within the autism spectrum (ie, PDD-NOS/Asperger disorder), the prevalence rate was 2.7 cases per 1000 children (95% CI: 1.7–4.0) with a male-to-female ratio of 3.7:1. The racial/ethnic distribution of children with ASD (89% white non-Hispanic, 4% Hispanic, 4% other races, and 3% unknown) was comparable with that of the Brick Township general population.
Age-specific rates were calculated for preschool (3- to 5-year-old) and school-aged (6- to 10-year-old) children (Table 2). CIs for the 2 age groups overlapped, indicating that the prevalence rates for the 2 age-groups were not different. However, rates tended to be lower among the school-aged children for both autistic disorder and all ASD combined.
Specific medical conditions found in other studies to be associated with autism were noted in 5 (8%) of the 60 case-children, including fragile X syndrome (2 children), seizure disorder (2 children), and a genetic translocation (1 child). Four families had >1 child with ASD. For 3 families, both children met the age criteria for inclusion in the study. In addition, 6 children were reported to have 1 or more siblings with a developmental disability other than autism: attention-deficit/hyperactivity disorder (2 children), speech/language disorders (2 children), and Down syndrome (1 child); 1 child had 1 sibling with cerebral palsy and another sibling with attention-deficit/hyperactivity disorder. One child with ASD had been adopted, and information about biological siblings was not available.
Educational placement and special education eligibility classification were obtained from school records. Four children with ASD did not receive any special education services. Of the remaining 56 children with ASD, 28 (50%) received services under the autism classification, 7 (12%) under traumatic brain injury, 2 (4%) under language/communication impairment, 2 (4%) under learning disability, 3 (5%) under multiple disabilities, and 14 (25%) in special needs preschool (ie, noncategorical placement). For children with autistic disorder, 1 child was not in special education; 23 (63%) were classified under the autism classification, 1 (3%) under language impairment, 1 under multiple disabilities (3%), 1 under traumatic brain injury (3%), and 9 (25%) were in special needs preschool.
Of the 60 children with ASD, maternal residence at the time of the child's birth was obtained from school records or other sources (eg, birth certificates) for 56 (93%) children. Of these 56 children with known birth residence, 36 (64%) were born in Brick Township, and 20 (36%) lived outside of Brick Township at the time of their birth.
Results From Clinical Assessment
Results presented in this section were obtained from the 43 case-children and their families who participated in the clinical assessments.
Four (9%) children were macrocephalic, which was defined as a head circumference above the 97th percentile for their age. Microcephaly (small head circumference) was not observed in any of the children. Dysmorphic features were observed in 13 (30%) children. The most commonly observed features involved abnormalities of formation or orientation of 1 or both ears (7 children), followed by abnormalities of the skin (4 children), nose (2 children), hands (2 children), eyes (1 child), and palate (1 child). A clinical geneticist reviewed photographs and videotapes of the children to assess facial and other features of all 43 children and concluded that none had a major recognizable syndrome.
Loss of Skills
Parents were asked during the clinical assessment whether their child had experienced loss of acquired skills before the diagnosis of autism with the question “Did your child experience any loss of previously acquired skills?” Parents who answered “yes” to this question, were asked to describe the what skills the child lost and at what age they noticed the loss of skills. Ten (24%) children, all with autistic disorder, were reported by their parents to have lost emerging skills. Typically, the parents described that the child stopped using single words and/or stopped social smiling. Parents reported the approximate age at which they first noticed the absence of skills as 12 months for 4 children, 13 months for 1 child, 15 months for 2 children, and 18 months for 3 children.
Intellectual and Behavioral Functioning
Four children (all with autistic disorder) could not complete intellectual testing because of limited language skills and/or lack of cooperation and are not included in these findings. The mean IQ score for all 39 children with ASD was 71.2 (SD: 19.5) and ranged from 34 to 118 (Table 3). Twenty (51%) of the children achieved scores in the normal range of functioning. The remaining nineteen children (49%) achieved scores in the range of mental retardation (ie, IQ score <70). For the 26 children whose condition met criteria for autistic disorder specifically, the mean IQ was 67.2 (SD: 18.8) and ranged from 34 to 118. Eleven (42%) of these children scored in the normal range and the remaining 15 (58%) scored in the range of mental retardation. (If the 4 children who were untestable are assumed to have IQs below 70, the proportion of children with autistic disorder who have mental retardation increases to 63%.) The 13 children with PDD-NOS/Asperger achieved higher IQs, with a mean of 81.5 (SD: 18.2) with 4 of these children scoring in the range of mental retardation.
Nonverbal abilities generally exceed verbal abilities for the group of children in this study with autistic disorder (Table 3). Pairwiset tests, using the difference between verbal and nonverbal cluster scores as the dependent variable, confirmed this pattern for the children with autistic disorder (t = 2.2,P < .05), but not for the remaining children within the autism spectrum (t = .40, P = .69). However, at the individual level, half (50%) of the children for whom appropriate scores were obtained achieved higher nonverbal than verbal cluster scores. Of the children with autistic disorder, 14 (54%) demonstrated higher nonverbal than verbal skills and 4 (30%) of the remaining children demonstrated this pattern.
Mean composite and domain scores for the Vineland Adaptive Behavior Scales are presented in Table 4. For all 42 children with available scores, the Adaptive Behavior Composite ranged from 27 to 105. The majority (58%) of children scored at the Low Adaptive Level, with 7 children (16%), including 2 children with autistic disorder, achieving scores at the Adequate Adaptive Level (ie, age-appropriate functioning). The Childhood Autism Rating Scale (CARS) provides information about the severity of autistic features. The mean total score on this measure for all children with ASD was 34.9 (SD: 10.2), which is in the mildly to moderately autistic range. The mean for those children with autistic disorder was 39.4 (SD: 8.2), which is well within the severely autistic range. Fifteen children, including 4 with autistic disorder, obtained ratings below the autistic threshold. Eighteen children (17 with autistic disorder) obtained ratings in the severely autistic range. The remaining 10 children (9 with autistic disorder) obtained ratings in the mildly to moderately autistic range.
In this study, we found a prevalence rate for ASD of 6.7 cases per 1000 children, and 4.0 cases per 1000 children for the subset of children who met full criteria for autistic disorder. These rates are substantially higher than the 0.3 to 0.4 cases per 1000 obtained from previous US studies.2–4 Differences in case-finding methods, changes in the diagnostic criteria for autism, and increased awareness may have influenced the number of children identified in this study.9,,26 Furthermore, previous studies were conducted when the spectrum of autism was less well understood, possibly concentrating on classic cases of severely affected individuals. How these factors influence prevalence rates in this study or other studies is not well understood.
Two comprehensive reviews of autism prevalence studies indicate that the findings in Brick Township were at the upper end of the range of prevalence rates found in other studies.5,,6,13 Both of these reviews found that most studies conducted since 1989 reported prevalence rates for autism of about one case per 1000 children.c However, 3 recent studies, conducted in Europe, found rates of 3.1 to 6.0 cases of autism per 1000 children,13,,17,18 close to those found in Brick Township. Like the Brick Township study, these studies were conducted in relatively small populations that allowed for more thorough case finding efforts. Working in small communities, such as Brick Township, also may allow for greater community contribution through increased awareness of case-finding efforts, resulting in higher rates. However, it should be noted that this community was selected for this study because of a possible high rate of autism and a study of a larger region or even surrounding communities could yield very different prevalence rates for autism.
Although this was not a study of migration patterns of children with autism, or other disabilities, we did note that about one third (36%) of the children lived outside Brick Township at birth. It would be helpful to be able to put this rate of moving into the town within a context of the rate of children in general or even the rate of children with other disabilities who moved into Brick Township after birth. Such information could be important for epidemiologic studies, because it would tell if the prevalence of autism or other disabilities for particular communities was the result of differential migration, such as when families move to find particular educational or medical services. Unfortunately, such comparison data were not available for this study. We present the rate of children with autism who moved into Brick Township so that prevalence studies of other communities can be compared in the future.
The identification of several well-defined characteristics at expected rates in the children with autism in Brick Township support the validity of the study's methodology and results. A greater number of males than females has been one of the most consistent findings in the autism literature, with ratios ranging from 2:1 to 4:1.9,,12 In Brick Township, the ratio was within this range, with 2.2 males per female with autistic disorder and 3.7 males per female for ASD. Also, the proportion of children with an associated medical condition in this study (8%) was within the range of other studies and similar to the median proportion of 6% described by Fombonne in his recent review of autism prevalence studies.5 A few studies have reported 14% to 20% of children with autism seen in clinic settings have macrocephaly,27–31 and about one fifth of patients had dysmorphic features.32–34 In this study, a subset of children with autism (9%) had macrocephaly and 30% had dysmorphic features. Difference in the rates of macrocephaly and dysmorphia in this study, as compared with previous studies, may reflect the fact that this study was population-based, whereas previous studies examined children referred to medical facilities and clinics. However, shape and orientation of ears was the most common dysmorphic feature, supporting findings from developmental biology studies of autism.35
The results of behavioral and cognitive measures also were consistent with previous literature on overall functioning of individuals with autism. Approximately half the children had mental retardation with most children, especially in whom autistic disorder was diagnosed, demonstrating higher nonverbal than verbal skills.36 The pattern of adaptive skills on the Vineland Adaptive Behavior Scales for this group of children was similar to the pattern reported by Carter et al37 describing children with autism below the age of 10 years. The full range of severity of autistic features displayed by children also was observed in this study, as measured by the CARS. Just less than a quarter of the children (23%) scored in the range of severely autistic, similar to the population of children with autism used to develop the CARS.25 However, 35% of children who scored in the nonautistic range on the CARS, is less than the 46% of the population of children with autism used to develop the CARS.25 Finally, 24% of children participating in the clinical assessment experienced a loss of emerging skills by parental report. Although this phenomenon is not well understood, estimates from a variety of studies range from 10% to over 50% of children with autism experience such loss of skills.12,,38
One finding in this study did differ from other epidemiologic studies of autism. In Brick Township, 50% more children were determined to have autistic disorder than other ASD (ie, PDD-NOS or Asperger disorder). A recent review of previous studies of this issue report relative proportions in the opposite direction (ie, 2 children with PDD-NOS or Asperger disorder for every one child with autistic disorder).5 Methodologic factors in the present study may explain this finding. Our methods may not have identified all the higher functioning children with PDD-NOS or Asperger disorder, especially if they had not come to the attention of a specialist. In addition, the ADOS-G has limited ability to discriminate autistic disorder from PDD-NOS.20 That is, autistic disorder may have been more likely to be diagnosed using the ADOS-G than if another diagnostic instrument had been used in the clinical assessment. Thus, the present study may have overestimated the prevalence of autistic disorder, compared with the other autism spectrum disorders, in Brick Township.
Strengths and Limitations
This study has several strengths. First, a multiple-source approach maximized identification of potential cases. Second, records for all children receiving special education services were independently reviewed rather than records of only children receiving special education services under the autism classification. This procedure was critically important because 50% of the children in Brick Township who had ASD (47% of those with autistic disorder) and receiving special education services, did not have autism as their eligibility classification. A third strength of this study was verification of the autism diagnosis using the ADOS-G, allowing for consistent application of the diagnostic labels and increased comparability with other recent studies. Finally, systematic collection of developmental, medical, and behavioral functioning data through the clinical assessments increased our understanding of this population of children with autism.
As mentioned earlier, the major limitation of this study was an inability to ascertain higher functioning individuals who were not in any special education class in public schools or had not been seen by participating clinicians. Consequently, because of these case-finding limitations, the results from Brick Township must be considered a minimal prevalence for autism. Categorical distinctions between autistic disorder and the other ASD were probably limited because the ADOS-G has been found to over estimate autistic disorder relative to PDD-NOS.20 Also, because clinical assessments could not be conducted for 17 children and the diagnosis had to be based on records alone, the reliability and validity of the diagnosis for those children is limited. Discrimination between PDD-NOS and autistic disorder also may have been influenced for these cases given that over 56% of the children who participated in the clinical assessment were determined to have autistic disorder in comparison to only 27% of the children assessed by record review only. Finally, the prevalence rates for autism obtained in this study must be generalized with caution since the community was selected for study because of a suspicion of increased numbers of children with the disorder. Studies of larger populations, such as one that included surrounding communities, may yield different findings.
Although autism has been considered a relatively rare developmental disability, it seems that the true prevalence of autism may be higher than previously appreciated in the United States. Although the rates obtained in Brick Township were higher than those of most studies, the rates were within the range of a few recent studies that also used more thorough case-finding methods. Examination of the characteristics of the children with autism included in this study indicated that these characteristics seem to be similar to those in other studies of children with autism. Studies of developmental disabilities, like this one for autism in Brick Township, provide a foundation for epidemiologic studies that facilitate identification of risk factors and causes of autism. Additional prevalence studies from a number of large and diverse populations will allow for monitoring of autism trends, provide an overall context for the prevalence of autism in individuals communities, and help communities and service agencies provide resources and treatment for individuals with autism and their families.
We thank the children and families for their participation in this study. Brick Township Public Schools, Ocean County Health Department, and Brick Township administration provided valuable help and support in conducting this study. Dr Paul Fernhoff (Centers for Disease Control and Prevention and Emory University) was the clinical geneticist. Nancy Doernberg, Courtney Alison, Fiona Steele, and Kim McKee abstracted school and medical records. Jon Biao assisted with identification of local clinicians and other data sources.
- Received February 1, 2001.
- Accepted June 13, 2001.
- Address correspondence to Jacquelyn Bertrand, PhD, Centers for Disease Control and Prevention, 4770 Buford Hwy, MS-F49, Atlanta, GA 30341. E-mail:
Portions of this study were presented in a CDC report to the community of Brick Township, New Jersey.
↵FNa Although children diagnosed with Rett disorder or childhood disintegrative disorder may have features of autism and are included under the classification of pervasive developmental disorders by the DSM-IV,1these disorders generally are not included in most prevalence studies nor in the term ASD.
FNc These studies used a wide range of methods and different case-ascertainment procedures, diagnostic criteria, and case definitions. So comparisons across studies and conclusions about base rates of ASD or even autistic disorder are problematic.
↵FNb This study used theDSM-III criteria, which used the terms infantile autism (corresponding to autistic disorder) and childhood onset PDD or atypical PDD. In current terminology, these 3 categories combined would correspond to most of ASD, although some individuals with Asperger disorder may not be included in earlier case definitions.
- DSM-IV =
- Diagnostic and Statistical Manual, Fourth Edition •
- PDD-NOS =
- pervasive developmental disorder-not otherwise specified •
- ASD =
- autism spectrum disorders •
- ADOS-G =
- Autism Diagnostic Observation Schedule-Generic •
- ADI-R =
- Autism Diagnostic Review-Revised •
- CI =
- confidence interval •
- CARS =
- Childhood Autism Rating Scale
- ↵American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th Edition. Washington, DC: American Psychiatric Association; 1994
- ↵California Department of Developmental Services. Changes in the population of persons with autism and pervasive developmental disorders in California's Developmental Services System: 1987 through 1998.A Report to the Legislature, Department of Developmental Services. Sacramento, CA: California Department of Developmental Services; 1999
- ↵Agency for Toxic Substances and Disease Registry. ATSDR Public Health Assessment: Brick Township Investigation. Atlanta, GA: Agency for Toxic Substances and Disease Registry; 2000
- Howlin P,
- Moore A
- Filipek PA,
- Accardo PJ,
- Baranek GT,
- et al.
- ↵Nelson KB, Grether JK, Dambrosia JM, et al. Neuropeptides and Neurotrophins in Neonatal Blood of Children With Autism, Mental Retardation, or Cerebral Palsy [abstract]. San Diego, CA: American Academy of Neurology; 2000
- Boyle CA,
- Bertrand J,
- Yeargin-Allsopp M
- Arvidsson T,
- Danielsson D,
- Forsberg P,
- Gillberg C,
- Johansson M
- ↵US Bureau of the Census. Report on Economic Indicators Data. Washington, DC: US Department of Commerce; 1998
- ↵Lord C. The Autism Diagnostic Observation Schedule-Generic (ADOS-G). Paper presented at: NIH State of the Science in Autism: Screening and Diagnosis Working Conference; June 15–17, 1998; Bethesda, MD
- ↵Lord C, Rutter M, DiLavore P, Risi S. Autism Diagnostic Observation Schedule: Manual. Los Angeles, CA: Western Psychological Services; 1999
- ↵Elliott CD. Differential Ability Scales. New York, NY: The Psychological Corporation; 1990
- ↵Sparrow SS, Balla DA, Cicchetti DV. Vineland Adaptive Behavior Scales. Survey Edition. Circle Pines, MN: American Guidance Service; 1984
- ↵Schopler E, Reichler RJ, Renner BR. The Childhood Autism Rating Scale (CARS). Los Angeles, CA: Western Psychological Services; 1986
- Hillman RE,
- Kanafani N,
- Takahashi TN,
- Miles JH
- Davidovitch M,
- Patterson B,
- Gartside P
- Walker HA
- Tuchman RF
- Copyright © 2001 American Academy of Pediatrics