Increased Rate of Head Growth During Infancy in Autism

In 1943, Kanner published "Autistic Disturbances of Affective Contact."¹ Kanner had been chosen to lead the first child psychiatry service in a pediatric hospital in the United States, located in the Harriet Lane Home for Children in The Johns Hopkins Hospital, Baltimore, Md. With only medical school training in pediatrics and child psychiatry but with the gift of astute observation, Kanner soon recognized a disorder previously not described: autism. In addition to his remarkable recognition and conceptualization of the clinical signs shared by children with autism, he also observed how they differed. One of the differences he noticed was that although the children with autism appeared physically healthy, some of them had relatively large heads.

Now, more than 50 years after Kanner's observation, macrocephaly is known to occur in approximately 20% of individuals with idiopathic autism.² The finding of macrocephaly in autism is one of the most replicated neurobiological findings in the disorder. However, even though rates of macrocephaly are increased in autism, macrocephaly is usually not present at birth but seems to develop during childhood.³ This original finding was quickly replicated⁴ and stimulated research interest in the disorder. Some neuroimaging and autopsy studies suggested that macrocephaly in children and adults with autism was associated with an enlarged brain.^{5 - 9} More recent neuroimaging studies show that brain volume is increased in 2- to 5-year-old children with autism but not in adolescents and adults with the disorder.^{10 - 12} Although mean head circumference is increased in older individuals with autism and these individuals with autism remain significantly impaired, brain volume appears to normalize after early to middle childhood.^{10 ,12} Autism is now hypothesized to be, at least in part, a complex disorder involving brain growth and perhaps brain maturation. Other evidence suggests that brain microanatomic and functional organization and connectivity are also disturbed.^{7 ,13 - 15}

In this issue of THE JOURNAL, Courchesne and colleagues¹⁶ provide additional information about abnormalities of head size and brain volume in autism. If increased head size is not present at birth or infancy but brain volume is increased by 2 to 5 years of age, a logical inference is that rates of head and brain growth must be increased during the first 2 years in children with autism. This appears to be the case in the group studied by Courchesne et al. The investigators used retrospective birth data and early childhood head circumference data obtained from children diagnosed with autism. They compared the autism data with reference data on typically healthy developing children in the Fels Longitudinal Study.¹⁷ Between birth and 6 to 14 months, head circumference increased at a significantly greater rate in the children with autism than in the reference sample. Fifty-nine percent of the children with autism, but only 6% of the reference sample, had a 2 or more SDs increase in head circumference between birth and 6 to 14 months. Increased rate of head circumference growth during infancy was associated with increased whole brain volume and gray matter and increased cerebral gray matter during early childhood in autism.

Courchesne et al also report decreased mean head circumference at birth in autism. Previous studies had found no difference between mean birth head circumference in male infants with autism and the reference sample.³ But the reference data used in earlier studies and in the current study by Courchesne et al differ. In the earlier studies, the distribution of birth head circumferences in autism did not appear normally distributed and ranged from microcephaly to macrocephaly.^{3 - 4} A recent large population-based study also showed that the risk of autism is not associated with head circumference at birth.¹⁸

Courchesne et al conclude that increased rate of head circumference growth during infancy appears to indicate an increased risk for autism.¹⁶ However, risk requires the establishment of precedence.¹⁹ If increased rate of head growth in infancy is a risk factor for autism, it must precede the onset of the disorder. It is not yet clear if this is the case. Although increased rate of head growth during infancy may precede parental and clinical recognition of signs of autism, recognition of a disorder is not always coincident with its onset. Autism begins before 3 years of age, but exactly when it begins is not known. Some signs of developmental abnormality are usually present during infancy in children later diagnosed with autism.²⁰ In addition, findings of postmortem studies suggest that the brain abnormalities begin before birth in at least some cases of autism.^{7 ,14} Therefore, increased rate of head growth during infancy in autism may be an important concomitant or correlate of autism (ie, a physical symptom of the disorder) rather than a risk marker. Another recent population-based study shows that macrocephaly in infants aged 5 to 12 months is significantly associated with the development of the behavioral manifestations of autism spectrum disorders.²¹ The odds ratio for an autism spectrum disorder in children who were macrocephalic vs normocephalic in infancy was 5.44.

Can increased rate of head growth during infancy help in the early identification of children with autism? To explore this possibility, consider that the prevalence of autism is 16.8 per 10 000 infants,²² and the frequency of increased head circumference growth (≥2 SDs) during infancy, as reported by Courchesne et al, is 59% for children with autism vs 6% for typically healthy developing children. Assuming that the sample frequencies are the same as population frequencies would suggest that for every 10 000 infants, 600 would be expected to have increased rates of head growth and typical development, and 10 would be expected to have increased rates of head growth and autism. Some infants with other neurological and genetic disorders will also have increased rates of head growth during infancy. These estimates suggest that the findings of Courchesne et al are most useful to researchers attempting to define the underlying developmental neuropathology of autism rather than to physicians trying to identify young children with autism.

The study by Courchesne et al provides important direction for what studies need to be performed next. The findings should be confirmed in a larger sample of individuals with autism. In addition, brain growth in infancy and the relationship between patterns of early head and brain growth and the early course of autism need to be studied. It is premature to conclude that increased rate of head growth is a universal feature of autism. There may be different trajectories of head growth in autism³ and of brain growth. Because genetic factors are important in early childhood brain growth, different patterns of head and brain growth in autism may elucidate different genes involved in the disorder. Sophisticated and suitable experiments are also needed to describe the macroscopic and microscopic anatomic reasons for aberrations of early brain growth in autism.

Brain development during the first 2 years in typically healthy developing children is characterized by organizational events essential for the establishment of the elaborate circuitry and functional complexity of the brain.²³ Neuronal proliferation and migration are completed well before birth. An exuberant overproduction of dendrites, axons, and synapses begins before birth and continues during the first 2 years. In addition, programmed cell death, selective elimination of synapses and axonal process, as well as the effects of experience, quietly refine neural circuitry and neural communication and in essence help shape the brain. Nonradial glial cells are proliferating and differentiating. The organizational events in the brain during the first 2 years set the neurodevelopmental stage for the acquisition of language and the capacity for inference, a sense of self-awareness,²⁴ and, eventually, complex information processing.¹³

Although much more is now known about autism than when Kanner first described the disorder, many fundamental questions remain unanswered. Advances in genetics, neuroimaging, neuroscience, pathology, and other disciplines will hopefully reveal the answers to the many remaining questions about what happens in brain development in children with autism.