Awareness of autism as a significant public health problem, propelled largely by parent advocacy and recent prevalence estimates of 1 in 150 and 1 in 200 (
1
), has moved autism research into the mainstream of neuroscience. Analogous to other broadly defined and heterogeneous neurobehavioral conditions, such as mental retardation, autism spectrum disorder (ASD) likely represents many distinct conditions with numerous etiologies (2
). Autism spectrum disorder is also highly heritable, and recent genetic findings, including linkage and association studies and studies of rare genetic variation due to chromosome copy number variation (CNV), further emphasized its etiologic heterogeneity (3
, 4
). This etiological and clinical diversity has led to the formulation of the term “the autisms” to describe ASD (2
).The extent of this initially unforeseen heterogeneity presents significant challenges to those working in the field at all levels, from identification of its underlying neurobiological bases to development and evaluation of more effective treatments. This heterogeneity makes it difficult to conceive of models, whether cognitive, behavioral, or physiological, that capture common features of the autisms under one conceptual umbrella. The most popular models (e.g.,
5
, 6
), ranging from “weak central coherence,” deficits in “theory of mind,” “local hyperconnectivity” or “long-range underconnectivity,” “developmental dysconnection,” to the “male brain,” also reflect divergent levels of analysis and different disciplines, from anatomical, physiological, or cognitive, to more popular psychology—not easily amenable to integration. Nonetheless, these diverse approaches and synthetic attempts represent true progress and highlight several points of convergence in the current literature. It is in this context that the seven studies in this special issue of Biological Psychiatry are timely. Here, I briefly focus on two topics in common to a few of these articles: 1) the issue of enlarged heads (brains) in autism (Hobbs et al., pages 1048–1055, and Sacco et al., pages 1038–1047, both in this issue), and 2) physiological phenotypes (Orhekhova et al., pages 1022–1029, and McCleery et al., pages 1007–1014, both in this issue). Both share the common theme of looking for quantifiable features of autism (7
), endophenotypes that are closer to underlying biological processes than the broad cognitive and behavioral features that currently define the autisms.Head Size
Large heads and atypical head growth trajectory (e.g.,
8
) have been observed in ASD in several studies. Nearly half of the children originally described by Leo Kanner had large heads (), and macrocephaly, defined as head circumference (HC) greater than the 97th percentile is observed in about 20% of autistic children. Recent studies have also shown that macrocephaly is familial in ASD and thus likely inherited in this subset of approximately 20% of ASD cases (Spence and Geschwind, unpublished data, 2007). Because autism is thought to arise during prenatal development, Hobbs et al. studied head and body size during fetal development. They analyzed midgestation fetuses using retrospective assessment of ultrasound data in 45 autistic children and compared this with 222 typical children, finding no evidence of enlarged fetal HC in children with ASD, parallel with previous studies that have reported normal HC at birth. This agrees with the observations of Courchesne et al. (8
) who initially described normal HC followed by an increased rate of postnatal head growth in children with ASD. The current study by Hobbs et al. provides the first study of fetal HC in ASD and provides another tier of evidence that abnormal brain size is primarily related to postnatal developmental processes.This is not to imply that fetal development is entirely normal in ASD. It certainly is plausible that early processes, such as neurogenesis or process growth, are involved but manifest later as the brain matures. Additionally, Hobbs et al. did find a marginally significant increase in biparietal diameter (BPD) relative to HC in ASD subjects, as well as soft signs of developmental abnormalities of the renal system, the latter of which is considered a marker for increased risk of genetic disorders. These may be important clues and it will be important to replicate in larger samples of patients, understand their distribution in the population, and study the subset of potentially higher risk patients for specific genetic anomalies or mechanisms, such as CNV or chromosomal abnormalities.
Neurophysiological Phenotypes and Local Versus Global Processing
Although the core deficits of autism involve higher cognitive functions, such as social cognition, language, and repetitive and restrictive behaviors, alterations in the function of primary sensory systems have also been observed, consistent with decades of patient and parent reports of abnormal sensitivity to sensory stimuli (review,
10
). Children with ASD may perform better at detection of local details relative to global properties of a stimulus, supporting the notion that integration of primary order perceptions into higher order concepts is altered (11
). Whether this is a top-down deficit, as proposed by the theory of weak central coherence or due to heightened primary processing remains to be convincingly demonstrated (11
).Based on the idea that first-degree relatives of ASD probands will share aspects of heritable quantitative endophenotypes, albeit in a less severe form than their siblings with ASD, McCleery et al. studied the early developing magnocellular visual pathway in a small cohort of 6-month-old infants with autistic siblings. These infants are also at a more than twentyfold risk for developing ASD over those without autistic siblings, so that 10% are expected to develop a diagnosis of ASD by 2 to 3 years of age. The investigators used observation of the infants’ visual preference to measure luminance sensitivity (magnocellular [M]) and chromatic sensitivity (parvocellular [P]) and found an early enhancement in luminance sensitivity, consistent with a genetically based alteration in visual system development in the ASD infant siblings. Rather than identifying a deficit, these data indicate an enhancement at the level of this surrogate measure of the M pathway.
The authors suggest that this enhancement of a sensory stimulus could have negative consequences for the further upstream areas, negatively impacting higher order processes. An alternative interpretation is that this reflects a true functional enhancement for processing of local detail and primary percepts that alone is not sufficient to cause autism, because the majority of these children will develop typically; hence, its adaptive value and propagation within the normal population. From this perspective, the elements that lead to M pathway changes must be compounded by other factors, genetic or otherwise, that co-occur in the autistic sibling, leading to true central nervous system (CNS) dysfunction, resulting in ASD. Still, this does not diminish the potential value of M pathway alterations as a relevant endophenotype for genetic studies. It will be very interesting to determine the generalizability of these findings using other measures of M pathway function and across the autism spectrum, as well as the relationship to other areas of dysfunction along the life span in ASD.
There is a growing literature reporting primary sensory deficits in ASD subjects, not only in visual processing but in auditory and other modalities as well (
11
). As with the M pathway enhancements described by Hobbs et al., it remains possible that these other sensory abnormalities may not be due to primary sensory processing but be the result of more upstream processing dysfunction that feeds back to more primary visual processes. Conversely, when global or local deficits in neuronal synchrony are ascribed to processes such as weak central coherence or disrupted connectivity (e.g., 12
, 13
), the question arises as to whether these are causal or due to more elemental disturbances in sensory processing (11
).In this issue, Orekhova et al. studied electroencephalogram (EEG) at rest in two different populations of boys with ASD and identified increased gamma band (high frequency, “global”) oscillations in both groups of ASD subjects, providing independent validation of their findings. Importantly, the extent of gamma band increase was positively correlated with the extent of global developmental delay (mental age) as measured by various intelligence quotient (IQ) measures. This observation certainly suggests relevance of this quantifiable, physiological phenotype to the cognitive functioning of the subjects studied. However, IQ deficits are not core deficits or strongly heritable in ASD, and IQ measures do not measure the severity of the specific features of ASD, such as social communication, which is among its core features. Rather than solidifying a clear connection to ASD, this correlation to mental age more likely signifies that increased resting high-frequency activity is reflective of more global brain dysfunction. Although ASD has overlap with general intellectual disability (mental retardation [MR]), a key issue is how ASD is distinct from MR and other neuropsychiatric diseases thought to have a significant developmental component, a point that will be critical to understanding the relationship of this phenotype to ASD and ASD pathobiology in general.
Some Conclusions
From a neurobiological perspective, autism research is in its fetal stage. The studies in this issue add to rapidly growing literature showing that the tools are in hand to understand the relationships between different levels of analysis, from molecular and physiological to behavioral and different levels of the nervous system, from pathways to regions to circuits, to demystify the neural bases of the autisms.
It is now critical to distinguish the factors in autism that differentiate it from other related neurodevelopmental disorders, in addition to quantifying etiologically relevant phenotypes across diagnostic boundaries. Furthermore, when any specific physiological deficits are identified in a primary sensory system, their potential roles in causing ASD need careful scrutiny; such deficits, although providing significant clues from an etiological standpoint, may be peripheral to the core features. Additionally, primary and integrative disturbances need not be mutually exclusive; from a cellular neurobiological perspective, both may coexist (
13
). Because elements of human cognition and behavior, such as language or social relatedness, are based on the integrated activity of specific brain regions, including higher order association cortices, it remains likely that specific deficits in regional functioning or connectivity will be involved.I thank Judith Piggott, Ph.D., for her critical reading of the manuscript and Li Hong for her editorial assistance.
Dr. Geschwind receives funding from National Institutes of Health (NIH) and Autism Speaks for his own group’s research on autism.
References
- Prevalence of autism spectrum disorders–Autism and Developmental Disabilities Monitoring Network, 14 sites, United States, 2002.MMWR Surveill Summ. 2007; 56: 12-28
- Autism spectrum disorders: Developmental disconnection syndromes.Curr Opin Neurobiol. 2007; 17: 103-111
- Mapping autism risk loci using genetic linkage and chromosomal rearrangements.Nat Genet. 2007; 39: 319-328
- Strong association of de novo copy number mutations with autism.Science. 2007; 316: 445-449
- Autism: A window onto the development of the social and the analytic brain.Annu Rev Neurosci. 2005; 28: 109-126
- Understanding autism: Insights from mind and brain.Philos Trans R Soc Lond B Biol Sci. 2003; 358: 281-289
- The endophenotype concept in psychiatry: Etymology and strategic intentions.Am J Psychiatry. 2003; 160: 636-645
- Evidence of brain overgrowth in the first year of life in autism.JAMA. 2003; 290: 337-344
- Autistic disturbances of affective contact.Nerv Child. 1943; 2: 217-250
- Annotation: What do we know about sensory dysfunction in autism?.J Child Psychol Psychiatry. 2005; 46: 1255-1268
- Vagaries of visual perception in autism.Neuron. 2005; 48: 497-507
- What do disturbances in neural synchrony tell us about autism?.Biol Psychiatry. 2007; 62: 190-191
- Children and adolescents with autism exhibit reduced MEG steady-state gamma responses.Biol Psychiatry. 2007; 62: 192-197
Article info
Publication history
Accepted:
September 3,
2007
Received:
August 31,
2007
Identification
Copyright
© 2007 Society of Biological Psychiatry. Published by Elsevier Inc. All rights reserved.
