Accelerated head growth in early development of individuals with autism.
ABSTRACT Macrocephaly is one of the most consistent physical findings reported in autistic individuals. Previous studies attempted to determine if macrocephaly is associated with risk for autism. This study hypothesizes that an abnormal acceleration in head growth during early development, rather than macrocephaly, is associated with autism risk. To investigate this hypothesis, head circumference data were examined in 251 individuals from 82 multiplex (at least two individuals with autism) and 113 sporadic (no family history) families with autism. This examination included longitudinal measurements for 79 individuals. Nineteen percent of the original 251 individuals were found to have macrocephaly (head circumference >97%). Abnormal acceleration in head growth was defined as an increase of 25 or more percentile points in head circumference between two consecutive measurements. Thirty-five percent of individuals with multiple head circumference records had an abnormal increase in head circumference. Furthermore, autistic individuals with accelerated head growth in early childhood displayed higher levels of adaptive functioning and less social impairment. This study confirms the presence of abnormal acceleration in head growth during the first and second months of life in a subgroup of autistic individuals.
- SourceAvailable from: mit.eduActa paedopsychiatrica 02/1968; 35(4):100-36.
- [show abstract] [hide abstract]
ABSTRACT: To determine whether brain volume, as assessed on MRI scans, differs between individuals with autism and control subjects, and whether such differences are affected by age. Previous studies have found increased brain weight, head circumference, and MRI brain volume in children with autism. However, studies of brain size in adults with autism have yielded conflicting results. The authors hypothesize that enlargement of the brain may be a feature of brain development during early childhood in autism that normalizes with maturational processes. The authors measured total brain volumes from 1.5-mm coronal MRI scans in 67 non-mentally retarded children and adults with autism and 83 healthy community volunteers, ranging in age from 8 to 46 years. Head circumference was also measured. Groups did not differ on age, sex, verbal IQ, or socioeconomic status. Brain volumes were significantly larger for children with autism 12 years old and younger compared with normally developing children, when controlling for height. Brain volumes for individuals older than age 12 did not differ between the autism and control groups. Head circumference was increased in both younger and older groups of subjects with autism, suggesting that those subjects older than age 12 had increased brain volumes as children. Brain development in autism follows an abnormal pattern, with accelerated growth in early life that results in brain enlargement in childhood. Brain volume in adolescents and adults with autism is, however, normal, and appears to be due to a slight decrease in brain volume for these individuals at the same time that normal children are experiencing a slight increase.Neurology 08/2002; 59(2):175-83. · 8.25 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: To assess the prevalence of macrocephaly (head circumference > or = 1.88 standard deviations above normative data for age and sex or > 97th centile) in autism and other pervasive developmental disorders, 41 children with autism, and a comparison group of 21 children with tuberous sclerosis complex (TSC) or an unspecified seizure disorder were studied. Familiality of head circumference was also assessed from measurements of 133 first-degree relatives. Significantly higher rates of macrocephaly were found in probands with autism (12.2%) and their first-degree relatives (15.5%) when compared against a published normative sample. The incidence of macrocephaly in the comparison group of probands with TSC and seizure disorder (9.5%) and their first-degree relatives (8.3%) was higher than normative data as well, although the relation between macrocephaly and autism was more pronounced. Head circumference and extreme scores reflecting macrocephaly were moderately heritable in the present sample (H2 = 0.47). The increased prevalence of macrocephaly in relatives of children with autism compared with control children suggests that this characteristic may be a familial risk factor in the pathogenesis of autism.Developmental Medicine & Child Neurology 12/2000; 42(11):737-40. · 2.68 Impact Factor
Accelerated Head Growth in Early
Development of Individuals With Autism
Yulia A. Dementieva, PhD*†, Danica D. Vance*, Shannon L. Donnelly, BS*,
Leigh A. Elston, MS*, Chantelle M. Wolpert, MS, PA-C, MBA*, Sarah A. Ravan, MS‡,
G. Robert DeLong, MD§, Ruth K. Abramson, PhD‡, Harry H. Wright, MD‡and
Michael L. Cuccaro, PhD*?
Macrocephaly is one of the most consistent physical
findings reported in autistic individuals. Previous stud-
ies attempted to determine if macrocephaly is associ-
ated with risk for autism. This study hypothesizes that
an abnormal acceleration in head growth during early
development, rather than macrocephaly, is associated
with autism risk. To investigate this hypothesis, head
circumference data were examined in 251 individuals
from 82 multiplex (at least two individuals with au-
tism) and 113 sporadic (no family history) families with
autism. This examination included longitudinal mea-
surements for 79 individuals. Nineteen percent of the
original 251 individuals were found to have macro-
cephaly (head circumference >97%). Abnormal accel-
eration in head growth was defined as an increase of 25
or more percentile points in head circumference be-
tween two consecutive measurements. Thirty-five per-
cent of individuals with multiple head circumference
records had an abnormal increase in head circumfer-
ence. Furthermore, autistic individuals with acceler-
ated head growth in early childhood displayed higher
levels of adaptive functioning and less social impair-
ment. This study confirms the presence of abnormal
acceleration in head growth during the first and second
months of life in a subgroup of autistic individuals.
© 2005 by Elsevier Inc. All rights reserved.
Dementieva YA, Vance DD, Donnelly SL, Elston LA,
Wolpert CM, Ravan SA, DeLong GR, Abramson RK,
Wright HH, Cuccaro ML. Accelerated head growth in
early development of individuals with autism. Pediatr
Macrocephaly (head circumference ?97th percentile) is
a consistent physical finding in individuals with autism.
Kanner first reported an association between autism and
increased head circumference in his initial description of
11 children with autism . Despite numerous subsequent
studies [2-11], there is limited understanding of the link
between autism and macrocephaly.
Macrocephaly in autism has been investigated via ret-
rospective [6,7,12], prospective , and familial studies
. In addition, several studies have compared head
circumference in normal or clinical control subjects (e.g.,
individuals with language disorders or tuberous sclerosis
complex) and individuals with autism [2,4,12]. The autism
groups consistently had a greater occurrence of macro-
cephaly than either normal or clinical control subjects,
with the percentage of macrocephaly in autism samples
ranging from 14% to 34%.
Despite the consistent findings of macrocephaly in
autism, sex and developmental change appear to differen-
tially influence macrocephaly patterns. For instance, the
rates of macrocephaly with respect to sex ranged from
twice as low to five times greater for males than females
[6,13]. Two patterns of head circumference growth were
observed in autism with respect to age: a linear increase in
the rate of macrocephaly across age groups  and an
increase in early childhood that normalizes with matura-
tion . Several studies examined head circumference in
early development of children with autism and determined
that macrocephaly is typically not present at or shortly
after birth [8,11]. Other investigators, using neuroimaging
and postmortem examination results, suggested that mac-
From the * Department of Medicine and Center for Human Genetics,
§Division of Pediatrics, Department of Medicine, and?Division of
Psychiatry, Duke University Medical Center, Durham, North Carolina;
†Division of Mathematics and Applied Science, Marshall University,
Huntington, West Virginia; and‡W.S. Hall Psychiatric Institute,
University of South Carolina, Columbia, South Carolina.
Communications should be addressed to:
Dr. Cuccaro; Center for Human Genetics; Duke University Medical
Center, Box 3445; Durham, NC 27710.
Received March 15, 2004; accepted August 2, 2004.
102PEDIATRIC NEUROLOGY Vol. 32 No. 2© 2005 by Elsevier Inc. All rights reserved.
doi:10.1016/j.pediatrneurol.2004.08.005●0887-8994/05/$—see front matter
rocephaly in individuals with autism was associated with
an enlarged brain [14-16] and that the brain volume is
increased in 2- to 5-year-old children with autism
[2,17,18]. Despite these varied research findings, all of the
studies concluded that macrocephaly is an important
clinical feature related to autism and should be investi-
Most recently, Courchesne et al.  suggested that
autism is characterized by a reduced head size at birth and
a sudden increase in head size between 1 to 2 months, as
well as between 6 to 14 months. The authors emphasized
that the sudden and excessive growth in their autism
sample was a potential early indicator of autism risk,
which well preceded any behavioral signs of the disorder.
The current study is based on the hypothesis that
accelerated head growth is associated with autism risk and
that macrocephaly is a secondary manifestation of accel-
erated head growth. This hypothesis was tested by exam-
ining both macrocephaly and the rate of head circumfer-
ence growth during early childhood in participants from a
larger genetic epidemiology study of autism. In addition,
using standard measures of adaptive and behavioral func-
tioning, clinical correlates in the subset of individuals with
accelerated head growth were examined.
Materials and Methods
Two hundred fifty-one affected individuals from 195 autism families
ascertained for a genetic linkage research were included in this study.
Families were ascertained through clinical referrals and active recruit-
ment through lay organizations providing services to autism families.
The participants represent an independent sample of individuals from
families in our ongoing genetic studies. Informed consent was obtained
from all families. Ascertainment of families and evaluation of affected
individuals was conducted using research protocols approved by the
Institutional Review Board of the participating institutions.
Participants were from both sporadic families (1 individual affected in
a family and no family history of autism) and multiplex families (2 or
more individuals with autism in a family). Individuals were enrolled in
this study if they were between 3 and 21 years of age and had a clinical
diagnosis of autism. Medical records were reviewed to confirm the
medical and developmental history for each research participant. Indi-
viduals were excluded from the study if they had a disorder known to be
associated with autism (e.g., Rett syndrome, tuberous sclerosis complex,
fragile X syndrome, Smith-Lemli-Opitz syndrome, or a structural brain
abnormality) [20-23]. Additional exclusion factors included potentially
confounding factors such as significant prenatal or perinatal events,
meningitis, head trauma, or lead poisoning. These factors, although not
causally associated with autism, may significantly alter neurodevelop-
ment in early childhood, thereby complicating the clinical research
evaluation and possibly resulting in a phenotype similar to autism .
All participants met the criteria for autism set forth in the Diagnostic
and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV).
Clinical diagnoses were determined by experienced clinicians (M.C.,
H.H.W.) with careful review of available records and clinical evalua-
tions. The diagnosis of autism was supported by the Autism Diagnostic
Interview–Revised . The Autism Diagnostic Interview–Revised uses
a diagnostic algorithm for autism based on both DSM-IV criteria [26,27].
The Autism Diagnostic Interview–Revised is more comprehensive rela-
tive to screening or identification measures such as the Childhood Autism
Rating Scale . Research participants were assessed for their level of
adaptive functioning using the Vineland Adaptive Behavior Scales .
The Vineland Adaptive Behavior Scales are a standardized measure of
social, communication, and daily living skills and consists of a semi-
structured interview with a primary caregiver. Individuals with Vineland
Adaptive Behavior Scales–defined functioning of ?18 months (age-
equivalent) were excluded to ensure the validity of the Autism Diagnostic
Head circumference measurements were obtained through comprehen-
sive reviews of medical records and direct examinations by clinical staff
during field visits. Although we could not ensure the accuracy of
measurements recorded in medical charts, we attempted to correct for
potential errors by dropping measurements that were clearly inaccurate or
grossly inconsistent with previous measurements. Clinical staff (C.W.,
S.D., S.R.) were trained jointly during a collaborative autism research
meeting to ensure reliability. The corresponding percentiles were ob-
tained by using Percentile Data Files  for measurements between 0
and 36 months of age and by using head circumference data published by
Roche et al.  for measurements made from 36 to 216 months of age.
Macrocephaly was defined as a head circumference above the 97th
percentile (z score ?1.88). For individuals with multiple head circum-
ference measurements, macrocephaly was identified as at least one head
circumference measurement above the 97th percentile. Based on the
hypothesis that the head circumference of any individual child tends to
follow a particular percentile curve plus or minus 10 percentile points,
Lainhart et al. defined an abnormal rate of head growth as a change of 25
or more percentile points or as a z score difference of one or more
standard deviations .
Statistical analyses were performed using SAS software release 8.1
. P value of less than or equal to 0.05 was considered to be
statistically significant. Head circumference measurements were normal-
ized across sex and age by converting to the corresponding percentiles.
For group mean comparisons on clinical variables, a generalized esti-
mating equations approach (GENMOD procedure) was used. This
procedure is useful in that the model allows for the specification of the
within-group correlation structure (i.e., measurements taken on partici-
pants who share a common characteristic—in this case family). By using
this procedure, familial similarities in head circumference could be taken
into account. A nonparametric Wilcoxon test was used in case of small
sample size. Correlations between head circumference percentile and age
at examination were analyzed by taking into account the longitudinal
nature of the data (the fact that an individual can have head circumfer-
ence records in different age classes). When comparing the least square
means between different age classes, multiple comparison adjustments to
the P values were used.
Of the 251 autistic patients who participated, 183 (73%)
were males and 68 (27%) were females. This 2.7:1 male to
female ratio is not significantly different from the 3:1 ratio
observed in the general autism population . The
average age of patients when they entered the study was
8.15 years with a standard deviation of 4.43 years.
Demographics and family status (multiplex vs sporadic) of
these patients are presented in Table 1. Eighty-four percent
of participants were Caucasian. The sample was com-
prised of 82 multiplex families (42%) and 113 sporadic
For 251 patients in the study, 577 total head circumfer-
ence measurements were obtained with age at head mea-
103Dementieva et al: Accelerated Head Growth in Autism
surement ranging from birth to 18 years of age. The mean
head circumference percentile of the autism sample (64th
percentile) was significantly higher (z ? 8.0, P ? 0.0001)
than the 50th percentile mean of the general population.
Forty-seven patients (19%) were found to have macro-
cephaly (Macrocephalic group). This percentage is signif-
icantly higher (P ? 0.0001) than the expected rate (3.0%)
in the general population and similar to results reported in
previous studies [7,8]. Among these individuals with
macrocephaly, 31 (66%) were males and 16 (34%) were
females. A total of 204 individuals (81%) had a head
circumference below or equal to the 97th percentile
(Normocephalic group). Seven (3%) of 251 individuals
were microcephalic (head circumference ?3%). The Mac-
rocephalic and Normocephalic groups did not significantly
differ on age at Autism Diagnostic Interview–Revised
examination (Student’s t test, P ? 0.99). There was also
no significant difference between the two groups regard-
ing sex ratio (Fisher’s Exact Test, P ? 0.47). The groups
did not differ on any of the clinical variables from the
Autism Diagnostic Interview–Revised and the Vineland
Adaptive Behavior Scales.
Accelerated Head Growth
Seventy-nine (57 males, 22 females) individuals had at
least two head circumference measurements in a given
period of time (not necessarily starting at birth). Out of
these 79, 28 (35%) individuals had abnormal head growth
acceleration between two consecutive head circumference
measurements (not necessarily starting at birth; Acceler-
ated Head Growth group). Fifty-one individuals had nor-
mal head growth for consecutive measurements (not nec-
essarily starting at birth; Normal Head Growth group). The
number of females in the Accelerated Head Growth group
(6 girls vs 22 boys, 1:3.7 ratio) was lower than expected;
however the difference was not significant. There was no
significant difference between Accelerated Head Growth
and Normal Head Growth groups on age at Autism
Diagnostic Interview–Revised examination.
The next set of analyses yielded significant (P ? 0.05)
differences between the Accelerated Head Growth and Nor-
mal Head Growth groups on Vineland Adaptive Behavior
Scales socialization, daily living skills, and adaptive behavior
composite standard scores. The Accelerated Head Growth
group had significantly higher mean scores on each of these
measures of adaptive functioning (Table 2). The Accelerated
Head Growth and Normal Head Growth group did not differ
significantly on Autism Diagnostic Interview–Revised scores
although Accelerated Head Growth means were suggestive
of less impairment (data not shown).
Thirty-seven (10 females, 27 males) of 79 individuals
with at least two head measurements had serial head
circumference measurements starting at birth. Twenty-
four (65%) of these individuals manifested an abnormal
acceleration in head growth (Accelerated Head Growth
Starting at Birth group). Thirteen (35%) had consecutive
measurements with no large increase in head circumfer-
ence percentile (Normal Head Growth Starting at Birth
group). The number of females in the Accelerated Head
Growth Starting at Birth group (5 girls vs 19 boys, 1:3.8
ratio) was lower than expected; however, the difference
was not significant. There was no significant difference
between Accelerated Head Growth Starting at Birth and
Normal Head Growth Starting at Birth groups on age at
Autism Diagnostic Interview–Revised examination.
The comparisons of Accelerated Head Growth Starting
at Birth and Normal Head Growth Starting at Birth groups
on Vineland Adaptive Behavior Scales scores
Comparison of accelerated head growth and normal head growth groups
P Value Mean S.D.MeanS.D.
Communication Standard score
Socialization standard score
Daily living skills standard score
Adaptive Behavior Composite
P values are adjusted for familial similarities.
VABS ? Vineland Adaptive Behavior Scale
? Accelerated head growth
? Normal head growth
Table 1. Demographics
104 PEDIATRIC NEUROLOGY Vol. 32 No. 2
on Vineland Adaptive Behavior Scales and Autism Diag-
nostic Interview–Revised scores revealed significant dif-
ferences on the Vineland Adaptive Behavior Scales com-
munication (P ? 0.05) and adaptive behavior composite
standard scores (P ? 0.03) using the exact Wilcoxon test.
Individuals with abnormal acceleration in head growth
starting at birth had higher mean scores on these two
indices (Vineland Adaptive Behavior Scales communica-
tion: 59.3 [S.D. ? 35.2] vs 69.75 [S.D. ? 19.5]; Vineland
Adaptive Behavior Scales adaptive behavior composite:
52.7 [S.D. ? 24.5] vs 67.6 [S.D. ? 19.4]).
Association Between Head Circumference Percentile
and Age at the Time of Head Measurement
To investigate the age at abnormal acceleration in head
growth in children with autism, we first considered the
association between head circumference percentiles and
ages at the time of head measurement for all 245 individ-
uals in the study. Head circumference measurements of
autistic individuals were first grouped at 1-year intervals
based on age. For every individual with two or more
measurements in the specific age interval, his/her average
percentile for this period of time was computed before the
A significant (F18, 145? 4.06, P ? 0.0001) association
between head circumference percentile and age classes
was documented. Further analysis indicated that the sig-
nificant (adjusted P value ? 0.005) difference in least
square means occurred between the first (0 ? age in
months ? 1) and the second (1 ? age in months ? 12) age
classes. There was no significant relation between any of
the remaining age classes.
After the patients were grouped into 6-month time
intervals on the basis of the age at head measurement, it
became evident that the significant (adjusted P ? 0.02)
relationship in least square means was between the first (0
? age in months ? 1) and the second (1 ? age in months
? 6) age classes. Again, no relationship was noticed
between any of the remaining age classes.
We also looked at the linear regression between aver-
aged head circumference percentile (over all individuals in
class) and time classes based on the age at head measure-
ment. Our null hypothesis was that there would be no
significant relationship between age at head measurement
and mean head circumference percentile for the normal
population. Linear regression analysis for all time classes
(from birth to 18 years of age) did not yield a significant
correlation between mean head circumference percentile
and time (F1, 53? 0.56, P ? 0.46). However, when the
analysis was limited to 36 months of age (Fig 1), a
significant (F1, 35? 4.91, P ? 0.03) correlation between
mean head circumference percentile and the age class for
autistic individuals was observed.
Rates of Head Growth During the First Year of Life
To further assess the relationship between accelerated
head growth and age, longitudinal head circumference
data from individuals with repeated head circumference
records during discrete time intervals between birth and 1
year of age were examined (Table 3).
Forty-two children (31 males, 11 females) had a head
circumference measurement at the first time period (0 ?
Figure 1. Linear regression result for correlation between mean percentile and age for all individuals. Classes of age are defined by 1-month intervals
from 0 to 36 months. Mean percentile is computed for all individuals in the age class. Two classes, 10 ? age ? 11 and 13 ? age ? 14 (months), had
only one representative. Both individuals were females with head circumference percentile ?20th.
Time classes and individuals with repeated head
0 ? age ? 1
1 ? age ? 2
2 ? age ? 6
6 ? age ? 12
105Dementieva et al: Accelerated Head Growth in Autism
age in months ? 1, Period 1). The mean head circumfer-
ence percentile for these individuals was equal to the 48th
percentile, with a standard deviation of 29. Two (5%)
patients manifested macrocephaly. These data are consis-
tent with the hypothesis that at or shortly after birth
autistic children have head circumference percentile dis-
tribution similar to the general population, and usually do
not have macrocephaly. Note that for both macrocephalic
cases, the families had a history of macrocephalic individ-
uals and in each case at least one parent had a head
circumference measurement above the 97th percentile.
Among the 42 individuals with a measurement in Period
1, 17 (12 males, 5 females) had a head circumference
measurement in the second time period (1 ? age in
months ? 2, Period 2). The mean head circumference
percentile for these individuals was equal to the 75th
percentile (S.D. ? 20), significantly higher than the 50th
percentile mean of the general population. Two (12%) of
these children, including one from Period 1, had a head
circumference percentile above the 97th percentile. Be-
tween the first and second observations, 11 (65%) had
head growth of more than 25 percentile points, suggesting
abnormal acceleration of head growth during this time
period. Among these 11 patients with high acceleration of
head growth, only two were females, suggesting that this
pattern of accelerated head growth may occur more often
All 17 individuals with head circumference record in
Period 2 had a measurement in the third time period (2 ?
age in months ? 6, Period 3). Three (18%) of them
satisfied the definition for macrocephaly, including the
two individuals from Period 2. The mean head circumfer-
ence percentile for these individuals was equal to the 77th
percentile (S.D. ? 25.7), significantly higher than the 50th
percentile mean of the general population. Only one (6%)
individual had abnormal acceleration of head growth
during this time period.
Fifteen individuals (11 males, 4 females) with measure-
ments in the three previous time classes had measurements
between 6 and 12 months of age (Period 4). The mean
head circumference percentile for these individuals was
equal to the 75th percentile (S.D. ? 24), significantly
higher than the 50th percentile mean of the general
population. Three individuals had macrocephaly, includ-
ing the one individual from the previous time periods. The
head circumference percentiles for these 15 individuals for
all four period of times are presented in Table 4.
Autism is a behaviorally defined disorder with diagno-
sis dependent on history and observation of behavioral,
social, and communication impairments, making early
diagnosis problematic. Although the onset of autism oc-
curs before 3 years of age, the average age of diagnosis is
usually later and can stretch to 6 years of age . Recent
studies suggest that macrocephaly in some individuals
may be a risk factor for autism. If this hypothesis is true,
the physical finding of macrocephaly in autism could be
an important early indicator with etiologic and prognostic
Our findings suggest that in autism, macrocephaly
appears to be of secondary import relative to abnormal
acceleration rate in head growth. Abnormal acceleration in
head growth at the earliest stages of postnatal development
15 autistic individuals
Changes in head circumference percentile between four periods of time in
Mean Head Circumference Percentile
Class 2 Class 1Class 3 Class 4
Class 1: birth to 1 month of age. Class 2: 1 to 2 months of age. Class 3: 2 to 6 months of
age. Class 4: 6 to 12 months of age. If an individual has more than one observation in a
class, we compute his/her mean head circumference percentile. Bold numbers indicate
individuals with more than 25-point increase in head circumference percentile from
Class 1 to 2.
106PEDIATRIC NEUROLOGY Vol. 32 No. 2
may be a risk factor associated with the development of
autism spectrum disorders in some individuals.
In this retrospective study, 19% of the participants were
found to have macrocephaly; among individuals with at
least two head circumference records, 35% were found to
have an abnormal acceleration rate in early development.
Both of these values are significantly higher than expected
in the general population. However, in these data the
abnormal rate in head growth occurs more often than
macrocephaly in individuals with autism.
Our results are consistent with the hypotheses that
macrocephaly, or brain overgrowth, is not usually present
at the time of birth  and that in some individuals with
autism brain development is associated with accelerated
growth before 2-3 years of age resulting in brain enlarge-
ment in childhood [2,17]. The results of the present study
also replicate the recent findings of a sudden and excessive
increase of head size in individuals with autism between
the first and second months of life . Both studies
suggest that abnormal acceleration of head growth be-
tween the first and second months of life represents one of
the earliest physical findings in the autism population.
This replication of the results of Courchesne et al. 
clearly points to the need for further investigation of early
accelerated head growth in autism.
An additional point of interest lies in the finding that
abnormal acceleration of head circumference in early life
was associated with better functioning on select clinical
measures. These data support findings that macrocephalic
individuals with autism were higher functioning relative to
a normocephalic subgroup of individuals with autism .
In the current study, the group with accelerated head
growth starting at birth was less clinically impaired. This
observation raises the question of whether the accelerated
head growth in early childhood is a protective reaction in
response to pathognomic neurodevelopmental processes
that contribute to autism. However, these results conflict
with those of Courchesne et al.  who reported that
individuals with autism and accelerated growth were more
severely affected. This discrepancy may be a function of
small sample sizes in conjunction with clinical differences
between the groups (i.e., the Courchesne sample was more
broad with respect to autism severity). To answer this
question, early investigation of abnormal acceleration in
head growth in a broader range of individuals with autism
in comparison with either nonaffected sibling control
individuals or the normal population is required.
Given the consistency of accelerated head growth in
select groups of individuals with autism, this clinical
variation may be a potential stratification measure, both in
genetic and clinical studies. Increasingly, investigators are
exploring clinical variables to use as the basis for mean-
ingful subgroups that may share some common etiologic
underpinning. For instance, abnormal acceleration in head
growth may be used to identify a homogeneous subgroup
that contributes to overall linkage in a given chromosomal
location and help to more precisely map and localize
susceptibility genes. The link between abnormal acceler-
ation in head growth in early development and brain
overgrowth may be accounted for by genetic mechanisms.
For instance, recent work suggests that brain-derived
neurotrophic factor may be elevated in children at risk for
autism . This neural growth gene is a plausible
candidate with respect to early brain overgrowth.
There were several limitations in this study that may
affect interpretation of the results. It was not possible to
assess the reliability of the retrospective head measure-
ments from medical records as they were taken by differ-
ent individuals. To avoid potential bias toward incorrect
values, head circumference measurements were averaged
for individuals for a given period of time. Also, head
circumference measurements were made at different times
for different individuals. Measurements were grouped into
age classes to compensate for the different ages at the time
of measurement. Because this was a retrospective study in
a sample ascertained for a genetic linkage study, head
circumference measurements at early ages were not avail-
able for most participants. The study was also limited by
the lack of other physical measurements such as height,
weight, and nutritional diet. Also, head circumference
measurements were not obtained for parents or unaffected
siblings, which prevents us from ruling out familial
influences on head size. An additional concern was the
inclusion of seven (3%) individuals who were microce-
phalic. Several recent studies documented the association
of this phenotypic characteristic with autism. Fombonne et
al. reported the presence of microcephaly (head circum-
ference ?3%) in 15% of their sample of individuals with
autism , a number significantly higher than one would
expect in the normal population. Miles et al.  identi-
fied microcephaly in 7% of their sample of individuals
with autism and suggest that the pathogenesis of autism in
microcephalic individuals is different from other autistic
individuals, perhaps involving different genes. The current
study did not exclude microcephalic individuals from
consideration. However, similar and even stronger results
were obtained when these individuals were excluded
(results not shown).
In conclusion, regular observations of not only head
circumference but also its growth rate between measure-
ments during early stages of life may be important factors
for identifying children at risk for developing autism. The
finding of accelerated head growth can be detected before
the first clinical behavioral signs of autism occur (e.g.,
language delay, repetitive behaviors). Further, retrospec-
tive examination of head growth patterns in children with
autism and other neurodevelopmental disabilities may
provide useful clinical and research information. Clearly,
there is a need for prospective investigations of this
phenomenon to understand the precise nature of associa-
tion between the accelerated head growth and autism, as
well as other neurodevelopmental disabilities.
107 Dementieva et al: Accelerated Head Growth in Autism
The authors thank the patients with autism and their families who agreed
to participate in this study, the personnel of the Duke Center for Human
Genetics for their input on this project, and Dr. Margaret Pericak-Vance
for constant support and helpful discussions. This research was supported
by National Institutes of Health grants RO1 NS36768 and PO1 NS26630,
the DeCamp Foundation, the National Alliance for Autism Research, and
the WV-BRIN project.
 Aylward EH, Minshew NJ, Field K, Sparks BF, Singh N. Effects
of age on brain volume and head circumference in autism. Neurology
Bailey A, Luthert P, Bolton P, Le Couteur A, Rutter M, Harding
B. Autism and megalencephaly [letter]. Lancet 1993;341:1225-6
Fidler DJ, Bailey JN, Smalley SL. Macrocephaly in autism and
other pervasive developmental disorders. Dev Med Child Neurol 2000;
Bolton P, Macdonald H, Pickles A, et al. A case-control family
history study of autism. J Child Psychol Psychiatry Allied Disciplines
 Davidovitch M, Patterson B, Gartside P. Head circumference
measurements in children with autism. J Virol 1996;11:389-93.
Fombonne E, Roge B, Claverie J, Courty S, Fremolle J Micro-
cephaly and macrocephaly in autism. J Autism Dev Disord 1999;29:
 Lainhart JE, Piven J, Wzorek M, et al. Macrocephaly in children
and adults with autism. J Am Acad Child Adolesc Psychiatry 1997;36:
Parmeggiani A, Posar A, Giovanardi-Rossi P, Andermann F,
Zifkin B. Autism, macrocrania and epilepsy: How are they linked? Brain
Skjeldal OH, Sponheim E, Ganes T, Jellum E, Bakke S.
Childhood autism: The need for physical investigations. Brain Dev
Stevenson RE, Schroer RJ, Skinner C, Fender D, Simensen RJ.
Autism and macrocephaly. Lancet 1997;1744-5.
Woodhouse W, Bailey A, Rutter M, Bolton P, Baird G, Le
Couteur A. Head circumference in autism and other pervasive develop-
mental disorders. J Child Psychol Psychiatry 1996;37:665-71.
 Folstein SE, Santangelo SL, Gilman SE, et al. Predictors of
cognitive test patterns in autism families. J Child Psychol Psychiatry
 Piven J, Arndt S, Bailey J, Havercamp S, Andreasen NC,
Palmer P. An MRI study of brain size in autism. Am J Psychiatry
Bailey A, Luthert P, Dean A, et al. A clinicopathological study
of autism. Brain 1998;121:889-905.
Courchesne E, Muller RA, Saitoh O. Brain weight in autism:
Normal in the majority of cases, megalencephalic in rare cases. Neurol-
 Courchesne E, Karns CM, Davis HR, et al. Unusual brain
growth patterns in early life in patients with autistic disorder: An MRI
study. Neurology 2001;57:1-10.
Kanner L. Autistic disturbances of affective contact. Nervous
 Sparks BF, Friedman SD, Shaw DW, et al. Brain structural
abnormalities in young children with autism spectrum disorder. Neurol-
Courchesne E, Carper R, Akshoomoff N. Evidence of brain
overgrowth in the first year of life in autism. JAMA 2003;290:337-44.
 Baker P, Piven J, Sato Y. Autism and tuberous sclerosis
complex: Prevalence and clinical features. J Autism Dev Disord 1998;
 Tierney E, Nwokoro NA, Kelley RI. Behavioral phenotype of
RSH/Smith-Lemli-Opitz syndrome. Ment Retard Dev Disabil Res Rev
Trottier G, Srivastava L, Walker CD. Etiology of infantile
autism: A review of recent advances in genetic and neurobiological
research. J Psychiatry Neurosci 1999;24:103-15.
 Hendren RL, De Backer I, Pandina GJ. Review of neuroimag-
ing studies of child and adolescent psychiatric disorders from the past 10
years. J Am Acad Child Adolesc Psychiatry 2000;39:815-28.
Gillberg C, Coleman M. Autism and medical disorders: A
review of the literature. Dev Med Child Neurol 1996;38:191-202.
Lord C, Rutter M, Le Couteur A. Autism Diagnostic Interview–
Revised: A revised version of a diagnostic interview for caregivers of
individuals with possible pervasive developmental disorders. J Autism
Dev Disord 1994;24:659-85.
American Psychiatric Association. Diagnostic and statistical
manual of mental disorders (DSM-IV). Washington, D.C.: American
Psychiatric Press, 1994.
International Molecular Genetic Study of Autism Consor-
tium. A full genome screen for autism with evidence for linkage to a
region on chromosome 7q. Hum Mol Genet 1998;7:571-8.
 Schopler E, Reichler RJ, Renner BR. The Childhood Autism
Rating Scale. Los Angeles: Western Psychological Services, 1988.
 Sparrow SS, Balla D, Cicchetti D. Vineland Adaptive Behavior
Scales, Interview Edition. Circle Pines, MN: AGS Publishing, 1984.
 National Center for Health Statistics. CDC growth charts:
United States percentile data files with LMS values. Centers for Disease
Control and Prevention; 2003.
Roche AF, Mukherjee D, Guo SM, Moore WM. Head circum-
ference reference data: birth to 18 years. Pediatrics 1987;79:706-12.
 SAS Institute. SAS/STAT user’s guide (version 8). Cary, NC:
SAS Institute, 2000.
Lord C, Schopler E. Differences in sex ratios in autism as a
function of measured intelligence. J Autism Dev Disord 1985;15:185-93.
Filipek PA, Accardo PJ, Ashwal S, et al. Practice parameter:
Screening and diagnosis of autism: Report of the Quality Standards
Subcommittee of the American Academy of Neurology and the Child
Neurology Society. Neurology 2000;55:468-79.
 Gillberg C, de Souza L. Head circumference in autism, As-
perger syndrome, and ADHD: A comparative study. Dev Med Child
Nelson KB, Grether JK, Croen LA, et al. Neuropeptides and
neurotrophins in neonatal blood of children with autism or mental
retardation. Ann Neurol 2001;49:597-606.
 Miles JH, Hadden LL, Takahashi TN, Hillman RE. Head
circumference is an independent clinical finding associated with autism.
Am J Med Genet 2000;95:339-50.
108PEDIATRIC NEUROLOGYVol. 32 No. 2