Autism is a lifelong neurodevelopmental disorder
characterized by early onset of impairments in social interaction and
communication and unusual, stereotyped behaviors. Autism (i.e.,
autistic disorder) often is classified with two related, although less
severe, developmental disorders: Asperger disorder and pervasive
developmental disorder--not otherwise specified. These three constitute
the autism spectrum disorders (ASDs). Diagnosis of ASDs is based
exclusively on developmental pattern and behavioral observation (Box).
Two population-based studies conducted by CDC in selected U.S.
locations reported ASD prevalence of 3.4 and 6.7 per 1,000 children,
respectively (1,2). CDC also conducts two nationally
representative surveys, the National Health Interview Survey (NHIS) and
the National Survey of Children's Health (NSCH), in which parents are
asked whether their child ever received a diagnosis of autism. Because
of similarities in methodology used by the two surveys, CDC analyzed
2003--2004 data from NHIS and data from the first-ever NSCH (collected
during January 2003--July 2004) to 1) estimate the population-based
prevalence of parental report of diagnosed autism in the United States
and 2) assess parental reporting of child social, emotional, and
behavioral strengths and difficulties and special-health care needs
among children with and without reported autism. This report describes
the results of that analysis, which indicated that the prevalence of
parent-reported diagnosis of autism was 5.7 per 1,000 children in NHIS
and 5.5 per 1,000 children in NSCH. Prevalence estimates in the two
studies were similar across age, sex, and racial/ethnic populations.
The consistency in estimates between the two surveys suggests high
reliability for parental report of autism. These estimates suggest
that, as of 2003--2004, autism had been diagnosed in at least 300,000
U.S. children aged 4--17 years. In addition, parental reports of autism
were associated with reported social, emotional, and behavioral
symptoms and specialized needs. Thus, these surveys might be useful to
assess health, education, and social service needs of children with
NHIS is an ongoing, annual, in-person survey of the
civilian, noninstitutionalized U.S. population, based on a multistage
sampling of housing units. NHIS includes a child survey component in
which one child aged <17 years is selected randomly as the
interview subject from each applicable household. Responses to
questions about the child are obtained from parents or other
knowledgeable adults. For this report, the analysis was based on 24,673
children from the combined NHIS surveys of 2003 and 2004, the most
recent reporting years and a period comparable to that of the first
NSCH interviews. Response rates for the child survey component of NHIS
for 2003 and 2004 were 81.1% and 79.4%, respectively.
NSCH is a random-digit--dialed telephone survey
conducted by CDC as part of the State and Local Area Integrated
Telephone Survey sponsored by the Maternal and Child Health Bureau of
the Health Resources and Services Administration. NSCH was initiated in
2003 to estimate the prevalence of physical, emotional, and behavioral
child health indicators in combination with information on the family
context and neighborhood environment. The survey uses a multistage
sampling method based on identification of households with one or more
children aged <17 years and random selection of a sample
child as the interview subject. Parents or guardians of 102,353
children completed the interview during January 2003--July 2004. The
response rate was 55.5%. CDC plans to conduct this survey approximately
every 4 years.
In both surveys, autism was ascertained from the
question: "Has a doctor or health-care provider ever told you that
[child's name] has autism?" For both surveys, children who were aged
4--17 years at the time of the survey were selected; children with
missing data on autism (<0.2%) were excluded. The final samples
included 18,885 children from NHIS and 79,590 children from NSCH.
Population-based estimates of parent-reported autism
used weighted data to reflect the noninstitutionalized population of
children nationally. From NHIS, the consistency between parent-reported
autism and parental responses to the Strengths and Difficulties
Questionnaire (SDQ) was examined. The SDQ is a 25-item behavioral
screening instrument that includes both positive and negative
psychological attributes. The items are divided into five scales:
emotional symptoms, conduct problems, hyperactive behavior, peer
relationships, and prosocial behavior. For each scale, children are
categorized as having low, moderate, or high levels of difficulties on
the basis of established criteria for U.S. children (3). From
NSCH, the extent to which children reported to have autism also were
reported to have special health-care needs was examined (4).
Additionally, for children aged 4--5 years in the NSCH survey, a risk
score for developmental delay was computed from responses to a series
of questions from the Parent's Evaluation of Developmental Status
(PEDS) questionnaire (5). PEDS is designed to identify children
who have or are at risk for developmental problems generally, including
developmental problems associated with ASD.
Total prevalence estimates of parent-reported diagnosis
of autism and estimates by sex, age, and race/ethnicity were similar
for the two surveys (Table
1). Prevalence was 5.7 per 1,000 children (95% confidence interval
[CI] = 4.5--7.2) in NHIS and 5.5 per 1,000 children (CI = 4.7--6.4) in
NSCH. In both surveys, prevalence was 3.7 times as high for males as
for females, peak prevalence was observed at ages 6--11 years, and
lower rates were observed among children of Hispanic ethnicity.
On the basis of NHIS data, children with parent-reported
autism were more likely than children without autism to have moderate
or high levels of emotional symptoms, conduct problems, hyperactivity,
peer problems, and total difficulties (Table
2). Approximately 83% of children reported with autism had moderate
or high levels of total difficulties compared with 15% of children
without autism. The most notable differences were for peer problems
(82.0% versus 15.9%) and hyperactivity (65.2% versus 11.9%).
Substantially fewer children reported with autism had a high level of
social skills (39.6%) compared with children without autism (82.3%).
On the basis of NSCH data, 93.8% of children with
parent-reported autism were classified as having special health-care
needs lasting or expected to last >12 months; 90.1% were
reported as needing more medical, mental health, or educational
services than usual for a child of the same age or needing treatment or
counseling for an emotional, developmental, or behavioral problem (Table
3). These percentages compared with 19.6% and 10.5%, respectively,
for children reported without autism.
Substantial differences also were observed among
children aged 4--5 years regarding their risk for developmental delay,
on the basis of the PEDS questions. Among those with reported autism,
92.6% were classified as at high risk for developmental delay, 6.4%
were classified as at moderate risk, and 1.1% as at low risk. Among
children without autism, 9.4% were classified as at high risk for
developmental delay, 17.1% were classified as at moderate risk, and
73.6% as at low risk or no risk (Table
Reported by: LA Schieve, PhD, C Rice, PhD, C
Boyle, PhD, Div of Birth Defects and Developmental Disabilities, SN
Visser, MS, Div of Human Development and Disability, National Center on
Birth Defects and Developmental Disabilities; SJ Blumberg, PhD, Div of
Health Interview Statistics, National Center for Health Statistics, CDC.
Because diagnosis of autism is made only by assessment
of developmental patterns and observation of behavioral symptoms,
establishing and tracking prevalence is difficult; thus, multiple
methods for case ascertainment can be useful. Parental report of autism
diagnosis has not been assessed previously. However, a study of
parental report of birth defects (6) documented high
specificity; sensitivity varied depending on the defect. Validation
studies of self-reported medical conditions among adults have reported
similar findings (7,8). Although the autism diagnosis question
from NHIS and NSCH has not been externally validated, the consistency
of results from these two independent surveys of the U.S. population
during the same approximate period suggests a degree of reliability of
these estimates. The associations between reported autism and 1)
parental rating of difficulties with SDQ items included in NHIS and 2)
reports of special health-care needs and risk for developmental delay
in NSCH suggest consistency between reported autism and expected
behavioral and emotional symptoms and health-care use patterns.
Moreover, the estimates of autism in this report and the male-to-female
prevalence ratios are consistent with estimates from recent
population-based studies of autism and ASD prevalence on the basis of
clinical examination and medical and educational record review (1,2).
Additionally the finding that parent-reported autism prevalence was
highest during early school age (i.e., 6--11 years) appears similar to
that of a 1996 study using medical and education record review in
metropolitan Atlanta in which peak prevalence was observed among
children aged 5--8 years (1).
The findings in this report of parental report of
diagnosed autism complement other CDC studies of the prevalence of
autism, such as population-based surveillance conducted in Atlanta and
other areas of the United States that compose the Autism and
Developmental Disabilities Monitoring (ADDM) Network. Estimates from
ADDM surveillance are on the basis of medical and education record
review of eligible children aged 8 years and provide more detail
regarding the prevalence of ASDs in selected U.S. populations. In
addition, ADDM surveillance provides information on the clinical
characteristics of children with ASDs and will provide data to track
ASD prevalence trends in these populations. NHIS and NSCH data
supplement ADDM data by providing national estimates of parent-reported
diagnoses of autism. Because these surveys contain information on
health-care use and family functioning, they might be helpful for
future analyses examining the impact of ASDs on children and their
families. Previous reports on parent-reported estimates of other
developmental disabilities from NHIS (9) and of
attention-deficit/hyperactivity disorder from NSCH (10)
have provided valuable data on the national prevalence and insight into
the impact of these disorders on children in the United States.
Because autism is defined behaviorally, autism usually
is not diagnosed before age 4 years. Later identification of children
with autism might suggest an underestimate of autism prevalence among
younger age groups. Similarly, although autism is considered a chronic
condition, to what extent the "ever diagnosed" cases described in this
report reflect parental report of current levels of clinical
symptomatology rather than past diagnoses is unclear; thus, the peak
prevalence estimates for children aged 6--11 years might be reflective
of peak ages for diagnosis and treatment. Although Hispanic children
had lower rates of reported autism in the two surveys, whether the
lower rates resulted from etiologic differences or differential
cultural factors related to autism symptom recognition and access to
services for diagnosis and treatment could not be determined. Thus,
differences observed between age or race/ethnicity subgroups should not
be used to infer potential etiologic associations. Nonetheless, such
differences point to the need to consider potential underdiagnosis in
certain populations. CDC has recognized the need to improve early
detection of autism and has begun a public education campaign with
national partner groups (Learn the Signs. Act Early) to educate parents
and professionals about early warning signs of autism and other
developmental disorders and to encourage developmental screening and
The findings in this report are subject to at least four
limitations. First, the NSCH response rate was lower than the NHIS
response rate; however, the comparability of the prevalence estimates
suggests a differential nonresponse bias did not occur. Second,
parental report of autism is dependent on access to appropriate health
or educational services for diagnosis and communication of that
diagnosis to the parent. Third, because the survey asked only about
autism, how parents of children with diagnoses of other, less severe,
ASD disorders (i.e., pervasive developmental disorder--not otherwise
specified or Asperger disorder) might have responded is unclear.
However, because prevalence estimates from NHIS and NSCH are greater
than the estimate for ASDs (3.4 per 1,000 population) reported in a
previous study (1) and the estimate for autism (4.0 per 1,000
population [compared with 6.7 per 1,000 for ASDs]) reported in another (2),
the findings in this report might indicate that children with the other
two ASDs were reported by their parents as having autism. Finally, the
findings in this report represent cross-sectional analyses of NHIS and
NSCH data from interviews conducted during the same approximate period
and do not assess trends in the rate of autism.
Results from these two national surveys of parental
report of diagnosed autism suggest that, as of 2003--2004, autism had
been diagnosed in at least 300,000 U.S. children aged 4--17 years.
Parents who reported that their children had autism also reported these
children experienced moderate or high levels of social, emotional, and
behavioral difficulties and needed special health-care and educational
services. These population-based surveys might be useful to assess the
specialized health and educational needs of families and children with
disabilities such as autism.
- Yeargin-Allsopp M, Rice C, Karapurkar T, Doernberg N,
Boyle C, Murphy C. Prevalence of autism in a US metropolitan area. JAMA
- Bertrand J, Mars A, Boyle C, Bove F, Yeargin-Allsopp
M, Decoufle P. Prevalence of autism in a United States population: the
Brick Township, New Jersey, investigation. Pediatrics
- Bourdon KH, Goodman R, Rae DS, Simpson G, Koretz DS.
The strengths and difficulties questionnaire: U.S. normative data and
psychometric properties. J Am Acad Child Adolesc Psychiatry 2005;44:
- Bethell CD, Read D, Stein RE, Blumberg SJ, Wells SN,
Newacheck PN. Identifying children with special health care needs:
development and evaluation of a short screening instrument. Ambul
- Glascoe FP. Collaborating with parents: using
parents' evaluation of developmental status to detect and address
developmental and behavioral problems in children. Nashville, TN:
Ellsworth & Vandermeer Press, Ltd; 1998.
- Rasmussen SA, Mulinare J, Khoury MJ, Maloney EK.
Evaluation of birth defect histories obtained through maternal
interviews. Am J Hum Genet 1990;46:478--85.
- Bayliss EA, Ellis JL, Steiner JF. Subjective
assessments of comorbidity correlate with quality of life health
outcomes: initial validation of a comorbidity assessment instrument.
Health Qual Life Outcomes 2005;3:51.
- Martin LM, Leff M, Calonge N, Garrett C, Nelson DE.
Validation of self-reported chronic conditions and health services in a
managed care population. Am J Prev Med 2000;18:215--18.
- Boyle CA, Decoufle P, Yeargin-Allsopp M. Prevalence
and health impact of developmental disabilities in US children. Pediatr
Prevalence of diagnosis and medication treatment for
attention-deficit/hyperactivity disorder---United States, 2003. MMWR
Use of trade names and commercial sources
is for identification only and does not imply endorsement by the U.S.
Department of Health and Human Services.
References to non-CDC
sites on the Internet are provided as a service to MMWR readers
and do not constitute or imply endorsement of these organizations or
their programs by CDC or the U.S. Department of Health and Human
Services. CDC is not responsible for the content of pages found at
these sites. URL addresses listed in MMWR were current as of
the date of publication.