FIGURE 1. Social brain system. Adapted by permission of Robert T. Schultz.
Seth’s mother had experienced an uneventful pregnancy, and Seth’s delivery had been without complications. In fact, Seth’s mother had no concerns about her son’s development until he was about 10 months of age, when she noticed that he was not babbling like other babies of similar age. He had experienced many ear infections and chest colds as an infant and young child. His mother always wondered if this had anything to do with his ASD.
Chad had seemed like a perfectly normal baby until a few months before his second birthday, when he stopped saying the few words he had been learning and seemed to lose interest in other people. Later, his family wondered if his onset of ASD was related to the stresses associated with a family move to another home at that time. Were the two connected?
Lauren’s mother had an extremely difficult pregnancy. She had high blood pressure and was placed on total bed rest for 6 weeks. Her amniotic fluid levels were low, and the baby rarely moved. Her doctor was worried about fetal distress and induced labor with Pitocin a month early. Lauren was born weighing only 5 pounds. She required resuscitation at birth and remained hospitalized, on oxygen, for 2 weeks. Could these complications really be just a coincidence, or were they related to Lauren’s difficulties?
The three cases presented above make a couple of key points. First, each child with ASD has a unique developmental history, making it difficult to sort out which factors that are specific to that child’s early development might have contributed to autism, if any. Second, it is natural for parents to draw connections between certain events, such as a difficult pregnancy or family move, and the cause of their child’s ASD. Yet we know that other children who experience similar events don’t usually develop ASD. In most cases, it is difficult, if not impossible, to identify a single factor that could explain why a child developed ASD. However, scientific research has started to provide some general answers about the causes of ASD, and we will be reviewing that knowledge in this chapter. There is clear evidence suggesting that ASD is biological in origin and not caused by parenting or other psychosocial causes related to the parent–child relationship. It is also clear that ASD does not have a single cause. Rather, ASD is a spectrum of varying severity, and different individuals will have different underlying causes for their ASD, which include a combination of genetic and environmental factors.
BRAIN DIFFERENCES IN ASD
When Dr. Kanner first described autism in 1943, he wrote that children with the disorder were born with an “innate,” or inborn, difficulty bonding with people. In the mid-20th century, most doctors were trained in the psychoanalytic tradition, which attributed all behavioral and mental disorders to early childhood experiences. Thus, autism was suspected to be caused by the social environment rather than by biology. Dr. Kanner was influenced by these ideas. Later he and others blamed autism on parents. They described “refrigerator mothers” who were so emotionally cold and rejecting that they caused their children to retreat into an autistic “cocoon” of safety. This view began to lose credibility in the 1960s, however, after Dr. Bernard Rimland published Infantile Autism: The Syndrome and Its Implications for a Neural Theory of Behavior. In this 1964 book, the author attacked parent causation theories and pointed out that there was absolutely no research data to support them. He was the first to suggest that autism was due to differences in how the brain worked. This suggestion has stimulated a good deal of research into possible brain differences in individuals with ASD.
To understand where the information collected to date comes from, you need a rudimentary grasp of the methods used to study the brain. Structural imaging, such as magnetic resonance imaging (MRI), provides a picture of the brain’s anatomy or structures, including the fibers that link one part of the brain to another. Postmortem or autopsy studies examine the brains of people who have died. This method permits scientists to look at the brain in a much more detailed manner. Researchers are actually able to examine individual brain cells (called neurons) rather than the large structures, composed of millions of neurons, captured by MRI. Functional imaging, including functional MRI (fMRI), measures how the brain works. By studying patterns of blood flow while a person is engaged in a task, for example, scientists can study whether the same parts of the brain are active (and working as hard and as efficiently) in individuals with ASD as they are in persons without ASD. Figure 1 illustrates some of the major structures of the brain and how they affect social behavior.
FIGURE 1. Social brain system. Adapted by permission of Robert T. Schultz.
Autopsy Studies
Autopsy studies have uncovered some differences in the brains of persons with ASD. First, they have found that there are too many brain cells (neurons) in parts of the brain that support social and emotional behavior (amygdala) and learning (hippocampus). Moreover, the cells are smaller and more tightly packed together than they are supposed to be. This may mean that they do not have the correct shape and/or enough room to make the connections with other brain cells that allow them to function as they should. Autopsy investigations have also revealed significantly fewer brain cells in another part of the brain called the cerebellum, which is important in both motor coordination and cognitive activities. While these findings are very interesting, they are based on the study of relatively few brains (approximately 25) and were not found in all the brains examined. Thus we are not sure how widespread these brain differences are in people with ASD. Additionally, almost all the brains studied came from individuals with severe autism and intellectual disability, and in some cases epilepsy, and therefore we are not yet sure whether these results pertain to children or adolescents with milder forms of ASD.
Structural Imaging Studies
This research method has found a wide variety of differences among brains of persons with ASD as compared to those with typical development. One difference is found in the ventricles of the brain, which (in all people) contain fluid instead of brain tissue. Some studies have found that the ventricles are larger than normal in some people with ASD, which may mean that brain tissue around the ventricles has been lost. This finding is not specific to ASD, however; it has been found in a variety of other syndromes. It appears to be a marker for an abnormal brain, rather than unique to ASD.
Other studies have found that there are differences in the way the brain develops in children with ASD, and these differences can be seen very early in life. During the preschool years, young children with ASD usually have increased brain size, especially in the frontal and temporal lobes. These brain regions are known to be involved in social and language functions. However, this accelerated brain growth starts to level off by the early elementary school years. By adulthood, the brain size of persons with ASD does not differ from that of typical individuals. This unusual pattern of brain growth might be related to the development of the individual cells in the brain (neurons). During normal brain growth and development, there is at first a period of tremendous overproduction of neurons and the connections between neurons: the brain grows more cells and makes more connections than it really needs. Later, many of the neurons die and the connections between neurons that aren’t being used much are weeded out. Some scientists believe that the large size of brains in some individuals with ASD indicates that this “pruning” mechanism has failed. This might mean that there is more background “noise” (or static) in the brain, which prevents it from functioning most efficiently. This is currently only a hypothesis, and we do not yet know what causes the brain to be larger or how this directly affects the way the brain functions in people with ASD.
A recent study suggests that differences in structural brain development can be seen even before ASD is diagnosed. By following a group of infants who were at risk for ASD (because they had older siblings with ASD), researchers found that 6- to 12-month-olds who later developed ASD themselves had atypical development of the fiber tracts that connect one part of the brain to another (also called white matter). These connections are important because complex behaviors such as social interaction and language require many different parts of the brain to work together. As the brain develops, the different parts of the brain become functionally connected through these fiber tracts. If these connections are not developing normally, this might help explain why individuals with ASD have difficulty with social and language skills.
Functional Imaging and Other Studies of How the Brain Works
Two specific areas of the brain have been the focus of investigations into whether the brains of those with ASD work differently from the brains of those without ASD.
Frontal Lobes
Since ASD always involves social deficits and repetitive behaviors, the areas of the brain that control these functions have been a focus of neuroimaging studies. In the late 1970s, two American neurologists, Drs. Antonio Damasio and Ralph Maurer, published a paper that pointed out behavioral similarities between people with autism and patients with damage to their frontal lobes (the region in the front of the brain, just behind our eyes and forehead). Both groups had difficulty controlling their emotions, would get very upset by small changes, were compulsive (wanting things to be “just so”), and were rigid in their solutions to problems, seeing things in a concrete, black-and-white manner. This led to a theory, still influential today, that if the frontal lobes do not develop correctly, this could cause autism. Functional imaging studies have found differences in how the frontal lobes are working in people with ASD. For example, for individuals with ASD, there is both less blood flow to and less electrical activity (firing of neurons) in this region, suggesting that their frontal lobes are not as active as they are supposed to be. In people without ASD, it is typical that multiple brain regions are needed to perform a task. Some studies of individuals with ASD have found that the activity of the frontal lobes is not well coordinated with other parts of the brain during task performance. Studies of normal volunteers and people with frontal damage have shown us that the frontal lobes are important to planning, flexibility, organization, behavioral control, and reasoning. If they are not working as efficiently or as well as they should, this could explain some symptoms of ASD.
Temporal Lobes
A second region of specific interest in ASD is the temporal lobes, which lie on the sides of the brain, roughly at the level of the ears. The specific part of the temporal lobes that seems to be involved is their very inner lining, lying closest to the center of the brain. This region is known as the medial (or middle) temporal lobes. Some of the structures in this area include the amygdala and the hippocampus. These parts of the brain are important for recognizing the emotional significance of a stimulus (for example, an angry face), interpreting emotions and facial expressions, such as where another person is looking, and memory. Some autopsy and structural MRI studies have shown differences in the temporal lobes (for example, being larger than normal early in life or having too many small, densely packed neurons). Several studies, including those conducted by two of us (G. D. and J. M.), have shown that individuals with ASD do in fact have problems in processing basic aspects of social information, including face recognition and discrimination of facial expressions—processing that is controlled by parts of the temporal lobes.
For example, in a study published in 1999, Dr. Simon Baron-Cohen and his colleagues measured brain function (using fMRI) while people with and without ASD looked at pictures of eyes. Their job was to decide what emotion the eyes conveyed. These researchers found that adults without ASD relied heavily on both the amygdala and the frontal lobes to perform this task. In other words, these two regions seemed to be most important to processing the social and emotional information conveyed by the eyes. In contrast, adults with ASD used the frontal lobes much less than the normal adults and did not “turn on” the amygdala at all when looking at the pictures of eyes. Instead, they used other parts of the brain that are normally not active during this task. Another study, led by Dr. Robert Schultz at Yale University, found that people with ASD used the part of the brain that normally makes sense of objects when they looked at faces. Dawson and colleagues discovered that some very young (3- to 4-year-old) children with autism fail to show recognition of a familiar face, but show normal recognition of familiar objects. The brain systems responsible for interpreting information from others’ faces come on line very early in life, offering the possibility that face-processing impairments may turn out to be one of the earliest indicators of abnormal brain development in autism. These findings don’t mean that your child does not recognize you. Instead, they suggest that your child may rely more on cues other than facial features for recognition (such as touch and voice).
These findings suggest that one reason people with ASD may make less eye contact and may have so much more trouble understanding others’ emotions, thoughts, and intentions is that critical regions of the brain are not working when they should be. Even when people with ASD can figure out what someone’s eyes or face conveys, they do so in a different way than everyone else, which may be less efficient or take more time. Indeed, as you can see in the picture of the brain on page 60, many areas of the brain are involved in social behavior. Researchers are actively discovering which parts of this complex system do not function properly and therefore account for the difficulty that those with ASD have in relating socially.
Differences in the Connections among Brain Regions
As we mentioned earlier, structural imaging studies have found that the fiber tracts, or white matter, that connect different parts of the brain are not developing normally in persons with ASD. Studies using functional imaging have also supported the idea that ASD is associated with abnormal patterns of brain connectivity. Using fMRI, as well as measures of electrical activity in the brain, researchers have shown that, when people with ASD perform complex tasks, such as language, the different areas of the brain that are required for processing language are not functioning in a synchronized fashion. This may help to explain why it is often difficult for even high-functioning people with ASD to perform well when they are required to understand and respond to rapid and complex language and social information. Interestingly, this unusual pattern of brain connectivity might also help explain some of the special talents and strengths that many people with ASD have. These strengths often rely on the specialized expertise of one brain region, such as the ability to recall visual detail.
Summary
The field has come a long way since parents were considered to be the cause of ASD. Study after study finds brain differences in people with ASD. These differences appear very early in life and may help to explain both the difficulties associated with ASD, such as impairments in complex behaviors like social interaction, and the strengths. The sophistication of imaging tools is increasing rapidly but has likely not yet reached anything close to their eventual capacity. The techniques used a decade ago, when much of the research just reported was performed, were far less powerful and may have been able to identify only the most obvious brain differences. The next decade holds great promise for finding more answers about the brain differences in people with ASD and how they may cause the condition. In contrast, much progress has already been made in understanding how genes may contribute to the development of ASD.
GENETIC INFLUENCES IN ASD
In the 1970s, two eminent psychiatrists made an important discovery that opened the door to the study of genetic influences in ASD. Sir Michael Rutter is a famous British child psychiatrist who was knighted by the queen of England for his important research contributions to the understanding of ASD and other childhood disorders. Dr. Susan Folstein is a prominent child psychiatrist in the United States who has published many papers on the genetics of autism. Both scientists noted that, while the rate of autism in siblings was small, it was much, much higher than the rate in the general population. This led to the birth of a new field, the study of genetic contributions to autism. There is now very strong evidence that in most (but perhaps not all) families, genetics plays some role in the development of the condition. Unfortunately, it is already abundantly clear that untangling the genetics of autism will not be simple. We now know that hundreds of genes are involved in increasing the risk for ASD, and different families carry different sets of genes. It also appears that the genes have broader effects than just ASD itself. A variety of different conditions, such as language delay and learning disability, seem to run in the families of children with ASD that may all be caused by the same genes. Thus, ASD is just one of several possible outcomes of having these genes.
One piece of evidence that ASD has a genetic basis comes from twin studies, which have found that the likelihood that twins would both have autism was much higher if the twins are identical, that is, shared all their genes, than if they are fraternal, that is, shared on average only about half of their genes. Studies also showed that while many co-twins have full-blown ASD, others demonstrated language, cognitive, or social difficulties that fell short of an ASD diagnosis but were nevertheless significant clinically. Many scientists now believe that the genes involved do not cause ASD itself but instead cause a variety of language and social differences, of which ASD is the extreme form.
Some children with known genetic disorders have ASD, providing another piece of evidence that ASD can have genetic origins. Fragile X syndrome and tuberous sclerosis are two genetic conditions that are readily diagnosed through genetic testing. The specific mutations in DNA that cause the conditions are known, and it is possible to diagnose both disorders before a child is born and to counsel couples who carry the genetic mutations for future pregnancies. A proportion of children with both fragile X syndrome and tuberous sclerosis develop autistic symptoms,1 suggesting that the genes involved in causing these disorders may also be involved in causing ASD. Because it is now possible to identify single genes that are known to contribute to ASD in some cases, the American Academy of Pediatrics recommends that genetic testing be conducted on all individuals with a diagnosis of ASD. Two types of tests are done. One test, called microarray analysis, examines all of the individual chromosomes for mutated or missing genes. The second test specifically tests for fragile X syndrome, one of the most common single-gene causes of ASD.
As mentioned earlier, there appear to be a variety of traits that run in the families of people with ASD, especially traits related to the areas of language and social abilities. Higher rates of language delay, articulation problems, learning difficulties, social difficulties, and social anxiety are found in relatives of people with ASD than in family members of people with other disabilities. Studies indicate that these milder difficulties show up in about 10–20% of siblings of individuals with ASD and often in parents as well.
The strengths of people with ASD can run in their families too. Parents and siblings often have talents and interests similar to those of people with ASDs. British researcher Simon Baron-Cohen proposed a theory that people in families with a member with an ASD would be especially skilled at understanding mechanical topics (such as how machines work), physical cause-and-effect mechanisms, and visual–spatial problems (as in puzzles). He and others have tested the theory that not only characteristic difficulties but also characteristic strengths run in families. Dr. Baron-Cohen’s team found that the parents of children with ASD were more likely to be engineers, physicists, and mathematicians than the parents of other children (other scientists have also found elevated rates of accounting and science careers in families of individuals with ASD). Dr. Baron-Cohen’s research group also surveyed over a thousand university students majoring either in literature or in math, physics, or engineering. They found that the rate of ASD was significantly higher in the families of the math, physics, or engineering students than in the families of the literature students. Finally, this British group of researchers tested the parents of children with ASD directly, giving them tests of both social understanding and visual–spatial skill. They found that the parents of students with ASDs were more skilled at solving puzzles and finding hidden shapes in complex drawings than control parents, as well as slightly less accurate than the other parents at interpreting the expressions on people’s faces. Some more recent research projects have confirmed the finding that strong visual–spatial, mechanical, and memory skills are often found in the families of individuals with ASDs.
These findings all converge on one conclusion. What appears to be genetically transmitted in the families of people with ASD is not ASD per se, but a certain distinctive style of thinking about, relating to, and reacting to the world that brings with it both strengths and challenges. This underscores the future limits we envision in the ability to provide genetic counseling. If a person carries certain genes, it does not necessarily mean that he or she will have ASD. ASD may represent the most extreme outcome of the genes, but it may be just one of several possibilities, many of which are strengths. A further limitation in genetic counseling is that in some cases the genetic alteration that might have caused an individual to be vulnerable to ASD is not always inherited from the parent. Studies have found that sometimes these mutations occur spontaneously in the egg or sperm.
WHAT ABOUT ENVIRONMENTAL CAUSES?
Over the past several years, there has been increasing interest in understanding how the combination of genetic risk factors and environmental influences can increase risk for ASD. We will review some of these findings in this section. It is important to keep in mind that it is highly unlikely that one environmental risk factor, by itself, is the cause of an individual’s ASD. Rather, several environmental factors, in combination with genetic risk factors, likely explain an individual’s ASD. This has been likened to a bucket of water to which individual risk factors, represented as drops of water, are added. As these risk factors, or drops of water, accumulate, it finally reaches a threshold at which the water will overflow. Similarly, as the number of genetic and environmental risk factors increase for a given individual, the likelihood of reaching that threshold that results in ASD will increase. Thus, it is usually not possible to point to a single factor that explains why a given person develops ASD.
So what are some of the environmental factors that are associated with risk for ASD? Two factors have been associated with decreased risk for ASD. The first is being a female. Males are four to five times more likely to develop ASD than females. Studies suggest that being a female has a “protective” influence against genetic risk factors for ASD. It appears to take many more genetic “hits” before ASD develops if one is a female. The second factor associated with decreased risk for ASD is the mother’s nutrition right before and during pregnancy. Specifically, taking prenatal vitamins, especially folic acid, has been associated with lower rates of ASD.
Although much of the research on environmental factors that might increase risk for ASD has only recently been published and needs to be replicated, a few findings are starting to emerge. One finding is that parents who are older when they conceive are slightly more likely to have a child with ASD. Although we don’t know exactly why this is true, it is known that as we age, all of us accumulate tiny deletions in our chromosomes, and it is possible that this might contribute to increased risk for ASD. Recent studies suggest that exposure to high levels of toxins, such as air pollution, may be associated with increased risk for ASD. Also, certain pregnancy and birth complications are associated with increased risk for ASD. These include exposure to severe infections during pregnancy, such as the flu with high fever; being born prematurely with very low birthweight; and experiencing birth complications that restrict oxygen to the brain. It has also been suggested that such complications are not causes of autism but consequences of it. This interesting hypothesis speculates that obstetric problems may occur in pregnancies in which something has already gone wrong with fetal development. Evidence for this comes from children with genetic disorders like Down syndrome, whose mothers have higher than average rates of pregnancy and delivery complications. Down syndrome is determined at the moment of conception. Therefore, there is something different about the growing baby long before the obstetrical complications take place. Some scientists have wondered if a similar scenario could account for the slightly elevated rate of prenatal and birth difficulties seen in individuals with ASD. They reasoned that genetic factors, as discussed earlier in this chapter, act early in fetal development to in some way weaken or jeopardize the fetus so that the pregnancy and delivery fail to progress normally. In other words, the complications follow, rather than lead to, the ASD. As we mentioned earlier, none of these factors—by itself—is likely to explain a given child’s ASD. Rather, these factors have been shown to slightly increase the risk for ASD, especially in genetically vulnerable people.
In recent years, there has been increased interest in whether ASD could result from an inherited immune system deficiency that makes children more susceptible to viral or bacterial infections. Unable to clear the organism quickly, it has been hypothesized that the fetus or infant is at increased risk that the infection might damage the brain directly. Another possible mechanism is that early infections might trigger an autoimmune response, in which the body’s immune system turns on itself and attacks its own parts as if they were foreign invaders (like viruses). It is thought that a similar breakdown of the immune system’s “self-recognition” mechanisms leads to other autoimmune illnesses, such as diabetes. In diabetes, for example, the immune system is triggered by an infection, but instead of just fighting the virus or bacterium, it also attacks the body’s own pancreas, killing the cells that produce insulin. The lack of insulin then causes the symptoms of diabetes. It has been proposed that some children with ASD might undergo such an autoimmune process, but that the organ of the body attacked by the immune system in the case of ASD is the brain rather than the pancreas. A few studies have found evidence of antibodies (immune proteins usually produced to combat infections) that “recognized” brain cells as “foreign” in some children with ASD. One further prediction of the autoimmunity theory is that children with ASD should have a higher rate of other autoimmune illnesses, such as asthma, allergies, arthritis, diabetes, multiple sclerosis, and the like. Some studies have indeed found an elevated rate of some of these difficulties in both children with ASD and their family members.
The environmental factors discussed above are ones that could potentially influence the development of the fetus during pregnancy. One factor that would have an influence later in a child’s development that has received much interest and controversy is vaccines, especially the measles–mumps–rubella (MMR) vaccine that is administered at about the time that some children with ASD regress and develop symptoms. The many studies that have been conducted to date do not support this hypothesis.
This chapter has discussed a wide range of potential causes for ASD, including genetic and environmental risk factors. There are many different ways for an individual to end up with ASD. Even when causes are found, each cause will apply only to a small group of children. New theories are generated every few months, some of which prove to be fruitful and others of which lead to dead ends. Parents often ask if they should have new medical tests performed as new theories about cause appear. Typically, doctors suggest waiting, because it will take time for a useful theory explaining ASD to evolve, as research centers around the world tackle the question. The search for answers is a universal human need that affects both parents and scientists, so there is no doubt that the search for causes will continue and intensify in the future.
____________________
1 However, only a very small proportion of children with autism spectrum disorder have either fragile X syndrome or tuberous sclerosis. Some doctors routinely screen all children with ASD for these conditions, but others suggest genetic testing only if the child with ASD has certain physical features common in fragile X syndrome or tuberous sclerosis, as mentioned in Chapter 2.