Far From the Tree

by Solomon, Andrew

  The local school didn’t want Ben and made his life miserable, and the Lehrs sued the district. Sue told authorities, “You can’t keep him out of your building because he is brown. Tell me where it says you can keep him out because he has autism.” The work at school was modified for him, but he had to do it, though he had little language and couldn’t initiate speech. Some people who cannot produce oral words can communicate in writing, and some who don’t have the muscle control for handwriting type instead, and some who don’t have even the control for typing use other methods. Ben learned facilitated communication, or FC, a system in which someone helped him to use a keyboard by giving his arms nondirected physical support as he typed. There has been great debate about whether what is expressed using FC is really the language of the disabled person or of the facilitator; Ben’s parents are sure that he is controlling his FC utterances.

  As he grew up, Ben would often smash his head on the floor, use knives to cut himself, put his head through windows. “His behaviors were a way of communicating,” Sue said. “Not the best way, but other kids communicate using drugs or driving snowmobiles drunk.” When Ben was a teenager, Bob and Sue took him to RadioShack, his favorite store. He panicked on the escalator, and at the bottom he sat down cross-legged and began smashing himself in the head with his hands and screaming as a crowd gathered. Sue always carried an FC keyboard, and when she took it out, Ben typed, Hit me. “And I thought, ‘Oh, yeah, in the middle of the mall with a security guard, and you’re black and I’m white,’” Sue recalled. “And then he typed out, Like a record player.” Sue suddenly flashed on a stuck needle; she struck him on the edge of the shoulder with the heel of her hand and said, “Tilt.” Ben stood up and they walked calmly on across the mall.

  In high school, Ben began having horrendous behavior problems. “I didn’t like his aide, Willie, an obese, slovenly guy who always wore sweatpants,” Sue said. “But I thought maybe I was just being judgmental. Then he was arrested for raping his own three-year-old daughter. Meanwhile, Ben was typing out that Willie had been hurting him and gave enough details to his speech therapist that she had the principal call the cops. Willie would say, ‘Ben’s having a hard time, so we’ll go up to the weight room and lift weights.’ And that’s where Willie was raping him, while this other guy would watch. So we brought Ben home for a while and nurtured him, to make sure he didn’t think it was his fault.” When he returned to school, Ben developed meaningful relationships with his classmates, assisted by a particularly well-attuned aide. In his senior year, he used FC to write a column for the school newspaper. He invited a nondisabled girl to the prom, and she accepted (somewhat to the chagrin of her boyfriend); at the prom, he was elected to the king’s court. At graduation, when he walked down to get his diploma, the whole audience stood up. Both Sue and Bob began to cry as she described it. “Thousands of people at this graduation. And they all stood up and applauded for Ben.”

  I was struck by the Lehrs’ early decision to help Ben but not to “fix” him. “His sister said to me, ‘Do you ever wonder what it would be like if Ben were normal?’” Sue said. “And I said, ‘Well, I think he’s normal for himself.’ Have I wished that he didn’t have all of his behavior problems? Absolutely. Have I wished that he had better language? Absolutely.” Much of what he types is Delphic. For a while he kept typing, And you can cry. No one ever understood what he meant. Another day he typed, I want to stop those, jerky feelings, jerky hurting. I get upset, then look stupid. Bob described going to conferences and being surrounded by parents desperate for a cure—“It’s going to be all better next year, crap like that. We were avant-garde in saying, ‘No. It’s going to be better right now. Let’s make it as good as possible for him.’”

  After high school, Bob and Sue gave Ben the down payment for a house eight miles from their own. His Social Security check covered his mortgage and most of his utility bills. He earned money by making wooden tables to sell at craft fairs. Someone was with him constantly, either a trained aide or a lodger who shared the house in exchange for caretaking. Because water is Ben’s passion, the Lehrs found him places to swim and bought him a hot tub. A decade later, Sue’s mother died, and the Lehrs took their inheritance and went on a three-month family camping trip to Europe. “Each person in the family got to pick one thing they wanted to do,” Sue said. “Ben picked swimming in every body of water he could find. So he’s been in the Mediterranean, he’s been in the Aegean, he’s been in pools and lakes and streams. We have a picture of him in Athens, sitting on the top of a stone wall, the highest point in Athens. He’s got his little drumsticks and he’s tapping on the stones and he’s got a look of sheer joy on his face.”

  When they returned from Europe, Bob was diagnosed with Alzheimer’s, which had advanced considerably by the time I interviewed him for this book. For two years, Bob didn’t want anyone but Sue to know, but Ben would type out, Daddy’s sick. Observing that Sue was upset, he’d type, Mommy is broken. Finally, Bob sat down and explained that Ben was right, Daddy is sick, but he wasn’t going to die right away. In the face of this diagnosis, the Lehrs woke up anew to the profound effect Ben had had on them. “I absolutely handled the news differently than I would have if we hadn’t had Ben,” Bob said. Sue said, “I think I’d learned a lot from Ben about reading people, trying to understand what they’re thinking or feeling that they can’t articulate. About treating someone as a human being even when his thoughts and feelings are mixed up. How do we make you feel safe, loved, okay? I learned the way it works by having Ben. And so I had it ready when Bob needed it.”

  • • •

  Autism is associated with underconnectivity between hemispheres and an overabundance of local connections; the neuronal pruning that helps the average brain avoid overload does not appear to occur in autism. Many autistic children are born with smaller heads than the norm, but by six to fourteen months, many have larger heads than the norm. The brains of autistic children are often enlarged by 10 or 15 percent, a condition that appears to resolve as the children grow. The human brain consists of grey matter, where thought is generated, and white matter, which conveys that thought from one area to another. In autism, inflammation has been observed in areas of the brain that produce white matter; too much is produced too soon, creating terrible noise, much like what you might get if every time you picked up your telephone, you heard not only the voice of the person you were calling, but also a hundred other voices all on the line at the same time. The fact that you and the other person were both speaking clearly would get lost in the cacophony. In autism, neurological losses have been observed also in the cerebellum, the cerebral cortex, and the limbic system. Autism genes may alter brain levels of neurotransmitters at crucial stages of development.

  It seems likely that autism is a blanket term. Autistic behavior may prove to be a symptom of a variety of causes, much as epilepsy can be caused by a genetic defect in brain structure, a head injury, an infection, a tumor, or a stroke; or as dementia may be the result of Alzheimer’s, cerebrovascular degeneration, Huntington’s, or Parkinson’s. No single gene or consistent set of genes causes the syndrome, although many genes that have been identified are functionally connected to one another, forming a network in the brain. It is not yet clear whether autism-related genes always or sometimes require environmental triggers to become active, nor, if so, what the triggers may be. Researchers have studied many possible developmental influences: prenatal hormones; viruses such as rubella; environmental toxins such as plastics and insecticides; vaccines; metabolic imbalances; and drugs such as thalidomide and valproate. Autism may be genetic, determined by spontaneous new mutations or through inheritance; it is strongly correlated with paternal age, possibly because of germ line de novo mutations that occur spontaneously in the sperm of older fathers. In a recent study, the rate of autism increased fourfold when researchers compared fathers in their thirties to those in their twenties, and the situation appears to be more drastic for fathers in later stages
of life. Researchers have also hypothesized that autism is caused by mother/child genetic incompatibilities that play out during gestation. Others have proposed a theory of assortative mating, suggesting that people with particular personality types find one another more readily in our mobile, Internet-enabled era, so that two people with mildly autistic tendencies—“hypersystemizers”—produce children together in whom those traits are concentrated.

  If we knew what goes on in the brain during autism, it would help establish which genes are implicated. If we knew which genes were implicated, we might be able to figure out what is happening in the brain. If we have only fragmentary knowledge of each thing, both goals are elusive. Up to two hundred genes may be implicated in autism, and some evidence suggests that you need several to manifest the syndrome. Sometimes, epistatic, or modifier, genes influence the expression of primary genes; sometimes environmental factors influence the expression of these genes. The closer the relationship between genotype (what genes you have) and phenotype (what behavior or symptoms you manifest), the easier it is to discern. In autism, some people with a shared genotype don’t share a phenotype, and some with a shared phenotype don’t share a genotype. Genetic research has demonstrated “variable penetrance” in autism—that is, one can possess known risk genes and not be autistic, and conversely, one can be autistic without having any known risk genes.

  If one identical twin has autism, the chances are 60 to 90 percent that the other twin will be autistic as well, though the second twin may have a much milder or a much more severe version of autism. This indicates a strong genetic basis for the disorder. While traits such as eye color or Down syndrome are always shared by identical twins, many other characteristics are not shared absolutely, and the correlation for autism is the highest for any cognitive disorder—higher than for schizophrenia, depression, or obsessive-compulsive disorder.

  If one fraternal (nonidentical) twin has autism, the chances are 20 to 30 percent that the other twin will have autism. Fraternal twins do not have identical genetics, but they do have near-identical environments. Nontwin siblings of children with autism are some twenty times as likely to have the condition as members of the general population. Even unaffected close relatives of people with autism are likely to have some subclinical social difficulties. All this suggests that there are strong genetic factors in autism, but that genes alone do not explain all instances of the condition.

  A common disorder may be caused by a single anomalous gene. So anyone who has Huntington’s disease, for example, has the aberrant Huntington’s gene. Autism is the opposite of Huntington’s in this regard. Hundreds of different genetic anomalies can predispose someone to autism. No individual rare gene variant occurs in very many people, but much of the population has a variant of some kind. The genome is full of hot spots, areas that mutate more easily and frequently than others. Some diseases—breast cancer, for example—are linked to a small number of specific gene mutations, each of which occurs on a particular stretch of a particular chromosome, and they are easily traced because women who have them frequently reproduce. Autism genetics are harder to map because there seem to be many rare gene variants associated with autism that are not usually inherited. They are sprinkled all over the genome. As Matthew State, codirector of Yale’s Program on Neurogenetics, has said, “Saying you have found an autism linkage peak on the part of the genome you are studying is like saying you live near Starbucks. Who doesn’t live near Starbucks?”

  NIMH director Thomas Insel said, “It takes five thousand genes to grow a normal brain, and conceptually, any of them could go wrong and cause autism.” According to Michael Wigler at Cold Spring Harbor Laboratory, no single mutation is associated with more than 1 percent of instances of autism, and many of the genes implicated have yet to be discerned. It’s not clear whether the complex symptoms of autism arise from a number of separate genetic effects—compromising language separately from social behaviors, for example—or whether one genetic effect, brought about by multiple genes, cascades to various brain regions to generate the characteristics of the syndrome. Most genes associated with autism are pleiotropic, which means that they have multiple effects. Some of these effects are linked with conditions that often co-occur with autism, such as ADHD, epilepsy, and gastrointestinal disorders. Most demonstrate small effect sizes, which means that a gene may boost your chance of developing autism by 10 or 20 percent—not boost it tenfold, as would happen for many disease-risk alleles.

  Many genetic diseases occur because a particular gene is abnormal in its structure. In some others, however, a gene is missing entirely; in yet others, there are extra copies of a gene. So let’s consider the sentence “I am happy” as a stand-in for a sequence on the genome. The most frequent model for a disease would be for the sentence to come out as “I am harpy” or “I ag happy” or some other such disruption. In a rare case, though, it might come out as “I m hpy” or, alternatively, as “I amamamamamam happpppy.” Wigler and his colleague Jonathan Sebat have looked primarily at these copy-number variations. A basic principle of genetics is that we have two of each gene, one from our mother and one from our father. But sometimes, a person actually has three, four, or as many as twelve copies of a gene or group of genes; or in the case of deletions, only one copy of a gene or group of genes, or none at all. The average person has at least a dozen copy-number variations, generally benign. Certain locations on the genome appear to be linked with cognitive disorders. Repetitions in these locations are associated with vulnerability to schizophrenia, bipolar disorder, and autism. However, deletions in the same region are linked only to autism. Wigler has found that many of his autistic subjects possess large deletions, lacking as many as twenty-seven genes. Sebat is now studying whether people with autism and a repetition have the same syndrome as those with autism and a deletion. He has found some significant correlations—for example, that the people with a deletion consistently have larger heads than those with a duplication in the same spot.

  The ultimate goal is to map these genes, describe their function, develop model systems, clarify molecular and cellular mechanisms, and then, finally, devise practical applications of findings. We are still identifying the rare variants; we are at the tip of the iceberg. Wigler pointed out that even when we’ve got all the information, we will have to contend with gene interactions that are not always subject to mathematical mapping. “There is probably an interplay between personality and the deficit,” he said. “You and I could have similar deficiencies, but we would make different choices. It sounds odd that a two-year-old may be making a choice about what he can and can’t handle, but they probably do. You could have two kids that grow up in the same impoverished environment, and one joins the priesthood and the other becomes a thief, right? I think that can happen internally.”

  “We are at the place now where we were twenty-five years ago with cancer genetics,” said Daniel Geschwind, codirector of the Center for Neurobehavioral Genetics at UCLA. “We know about twenty percent of the genetics; given how late the work started compared with research on schizophrenia and depression, the progress is remarkable.” Autism is a catchall category for an unexplained constellation of symptoms. Whenever a subtype of autism with a specific mechanism is discovered, it ceases to be called autism and is assigned its own diagnostic name. Rett syndrome produces autistic symptoms; so, often, do phenylketonuria (PKU), tuberous sclerosis, neurofibromatosis, cortical dysplasia-focal epilepsy, Timothy syndrome, fragile X syndrome, and Joubert syndrome. People with these diagnoses are usually described as having “autistic-type behaviors,” but not autism per se. If autism is defined by behavior, however, it seems counterproductive to describe as “not autistic” those whose autistic behavior has a known origin.

  Until recently, researchers devoted limited energy to these infrequent syndromes, but some have now turned their attention to them with the thought that if we could understand why such conditions cause autistic behavior, we might be able to access the larger me
chanisms of autism.

  Rapamycin, an immunosuppressant drug usually used in organ transplants, has suppressed seizures and reversed learning disabilities and memory problems in adult mice with tuberous sclerosis; it might have a similar effect on some human beings with the condition. Dr. Alcino Silva of UCLA said of that work, “Memory is as much about discarding trivial details as it is about storing useful information. Our findings suggest that mice with the mutation cannot distinguish between important and unimportant data. We suspect that their brains are filled with meaningless noise that interferes with learning.” This evokes the sensory experiences described by many autistic people; “noise” may be a major mechanism of the syndrome.

  Fragile X and Rett syndrome are both single-gene mutations. People with fragile X have a gene mutation that encodes a protein that in turn blocks an important brake on protein synthesis in the brain. While the mechanism by which the mutation causes intellectual and behavioral deficits is not known, a current theory is that these symptoms result from excessive protein production. Mice artificially bred with the fragile X mutation overproduce protein, and show learning problems and social deficits. One therapy for fragile X syndrome would be to block the mGluR5 receptor, which is a major stimulus for protein synthesis in the brain. Drugs that do so have reduced the excessive protein, suppressed seizures, and normalized behavior in fragile X mice. The genetics and mechanism of Rett syndrome differ from those associated with fragile X, but mice artificially bred with the Rett syndrome mutation have likewise responded to drugs that target a pathway affected by their mutation.