Can Genetic Testing Help Shed Light on the Autism Epidemic?
Autism cases are still on the rise, and scientists don't know why. In April, the Centers for Disease Control (CDC) reported that rates of autism had increased once again, now at an estimated 1 in 59 children up from 1 in 68 just two years ago. Rates have been climbing steadily since 2007 when the CDC initially estimated that 1 in 150 children were on the autism spectrum.
Some clinicians are concerned that the creeping expansion of autism is causing the diagnosis to lose its meaning.
The standard explanation for this increase has been the expansion of the definition of autism to include milder forms like Asperger's, as well as a heightened awareness of the condition that has improved screening efforts. For example, the most recent jump is attributed to children in minority communities being diagnosed who might have previously gone under the radar. In addition, more federally funded resources are available to children with autism than other types of developmental disorders, which may prompt families or physicians to push harder for a diagnosis.
Some clinicians are concerned that the creeping expansion of autism is causing the diagnosis to lose its meaning. William Graf, a pediatric neurologist at Connecticut Children's Medical Center, says that when a nurse tells him that a new patient has a history of autism, the term is no longer a useful description. "Even though I know this topic extremely well, I cannot picture the child anymore," he says. "Use the words mild, moderate, or severe. Just give me a couple more clues, because when you say autism today, I have no idea what people are talking about anymore."
Genetic testing has emerged as one potential way to remedy the overly broad label by narrowing down a heterogeneous diagnosis to a specific genetic disorder. According to Suma Shankar, a medical geneticist at the University of California, Davis, up to 60 percent of autism cases could be attributed to underlying genetic causes. Common examples include Fragile X Syndrome or Rett Syndrome—neurodevelopmental disorders that are caused by mutations in individual genes and are behaviorally classified as autism.
With more than 500 different mutations associated with autism, very few additional diagnoses provide meaningful information.
Having a genetic diagnosis in addition to an autism diagnosis can help families in several ways, says Shankar. Knowing the genetic origin can alert families to other potential health problems that are linked to the mutation, such as heart defects or problems with the immune system. It may also help clinicians provide more targeted behavioral therapies and could one day lead to the development of drug treatments for underlying neurochemical abnormalities. "It will pave the way to begin to tease out treatments," Shankar says.
When a doctor diagnoses a child as having a specific genetic condition, the label of autism is still kept because it is more well-known and gives the child access to more state-funded resources. Children can thus be diagnosed with multiple conditions: autism spectrum disorder and their specific gene mutation. However, with more than 500 different mutations associated with autism, very few additional diagnoses provide meaningful information. What's more, the presence or absence of a mutation doesn't necessarily indicate whether the child is on the mild or severe end of the autism spectrum.
Because of this, Graf doubts that genetic classifications are really that useful. He tells the story of a boy with epilepsy and severe intellectual disabilities who was diagnosed with autism as a young child. Years later, Graf ordered genetic testing for the boy and discovered that he had a mutation in the gene SYNGAP1. However, this knowledge didn't change the boy's autism status. "That diagnosis [SYNGAP1] turns out to be very specific for him, but it will never be a household name. Biologically it's good to know, and now it's all over his chart. But on a societal level he still needs this catch-all label [of autism]," Graf says.
"It gives some information, but to what degree does that change treatment or prognosis?"
Jennifer Singh, a sociologist at Georgia Tech who wrote the book Multiple Autisms: Spectrums of Advocacy and Genomic Science, agrees. "I don't know that the knowledge gained from just having a gene that's linked to autism," is that beneficial, she says. "It gives some information, but to what degree does that change treatment or prognosis? Because at the end of the day you have to address the issues that are at hand, whatever they might be."
As more children are diagnosed with autism, knowledge of the underlying genetic mutation causing the condition could help families better understand the diagnosis and anticipate their child's developmental trajectory. However, for the vast majority, an additional label provides little clarity or consolation.
Instead of spending money on genetic screens, Singh thinks the resources would be better used on additional services for people who don't have access to behavioral, speech, or occupational therapy. "Things that are really going to matter for this child in their future," she says.
Imagine this scenario: A couple is involved in a heated custody dispute over their only child. As part of the effort to make the case of being a better guardian, one parent goes on a "genetic fishing expedition": this parent obtains a DNA sample from the other parent with the hope that such data will identify some genetic predisposition to a psychiatric condition (e.g., schizophrenia) and tilt the judge's custody decision in his or her favor.
As knowledge of psychiatric genetics is growing, it is likely to be introduced in civil cases, such as child custody disputes and education-related cases, raising a tangle of ethical and legal questions.
This is an example of how "behavioral genetic evidence" -- an umbrella term for information gathered from family history and genetic testing about pathological behaviors, including psychiatric conditions—may in the future be brought by litigants in court proceedings. Such evidence has been discussed primarily when criminal defendants sought to introduce it to make the claim that they are not responsible for their behavior or to justify their request for reduced sentencing and more lenient punishment.
However, civil cases are an emerging frontier for behavioral genetic evidence. It has already been introduced in tort litigation, such as personal injury claims, and as knowledge of psychiatric genetics is growing, it is further likely to be introduced in other civil cases, such as child custody disputes and education-related cases. But the introduction of such evidence raises a tangle of ethical and legal questions that civil courts will need to address. For example: how should such data be obtained? Who should get to present it and under what circumstances? And does the use of such evidence fit with the purposes of administering justice?
How Did We Get Here?
That behavioral genetic evidence is entering courts is unsurprising. Scientific evidence is a common feature of judicial proceedings, and genetic information may reveal relevant findings. For example, genetic evidence may elucidate whether a child's medical condition is due to genetic causes or medical malpractice, and it has been routinely used to identify alleged offenders or putative fathers. But behavioral genetic evidence is different from such other genetic data – it is shades of gray, instead of black and white.
Although efforts to understand the nature and origins of human behavior are ongoing, existing and likely future knowledge about behavioral genetics is limited. Behavioral disorders are highly complex and diverse. They commonly involve not one but multiple genes, each with a relatively small effect. They are impacted by many, yet unknown, interactions between genes, familial, and environmental factors such as poverty and childhood adversity.
And a specific gene variant may be associated with more than one behavioral disorder and be manifested with significantly different symptoms. Thus, biomarkers about "predispositions" for behavioral disorders cannot generally provide a diagnosis or an accurate estimate of whether, when, and at what severity a behavioral disorder will occur. And, unlike genetic testing that can confirm litigants' identity with 99.99% probability, behavioral genetic evidence is far more speculative.
Genetic theft raises questions about whose behavioral data are being obtained, by whom, and with what authority.
Whether judges, jurors, and other experts understand the nuances of behavioral genetics is unclear. Many people over-estimate the deterministic nature of genetics, and under-estimate the role of environments, especially with regards to mental health status. The U.S. individualistic culture of self-reliance and independence may further tilt the judicial scales because litigants in civil courts may be unjustly blamed for their "bad genes" while structural and societal determinants that lead to poor behavioral outcomes are ignored.
These concerns were recently captured in the Netflix series "13 Reasons Why," depicting a negligence lawsuit against a school brought by parents of a high-school student there (Hannah) who committed suicide. The legal tides shifted from the school's negligence in tolerating a culture of bullying to parental responsibility once cross-examination of Hannah's mother revealed a family history of anxiety, and the possibility that Hannah had a predisposition for mental illness, which (arguably) required therapy even in the absence of clear symptoms.
Where Is This Going?
The concerns are exacerbated given the ways in which behavioral genetic evidence may come to court in the future. One way is through "genetic theft," where genetic evidence is obtained from deserted property, such as soft-drink cans. This method is often used for identification purposes such as criminal and paternity proceedings, and it will likely expand to behavioral genetic data once available through "home kits" that are offered by direct-to-consumer companies.
Genetic theft raises questions about whose behavioral data are being obtained, by whom, and with what authority. In the scenario of child-custody dispute, for example, the sequencing of the other parent's DNA will necessarily intrude on the privacy of that parent, even as the scientific value of such information is limited. A parent on a "genetic fishing expedition" can also secretly sequence their child for psychiatric genetic predispositions, arguably, in order to take preventative measures to reduce the child's risk for developing a behavioral disorder. But should a parent be allowed to sequence the child without the other parent's consent, or regardless of whether the results will provide medical benefits to the child?
Similarly, although schools are required, and may be held accountable for failing to identify children with behavioral disabilities and to evaluate their educational needs, some parents may decline their child's evaluation by mental health professionals. Should schools secretly obtain a sample and sequence children for behavioral disorders, regardless of parental consent? My study of parents found that the overwhelming majority opposed imposed genetic testing by school authorities. But should parental preference or the child's best interests be the determinative factor? Alternatively, could schools use secretly obtained genetic data as a defense that they are fulfilling the child-find requirement under the law?
The stigma associated with behavioral disorders may intimidate some people enough that they back down from just claims.
In general, samples obtained through genetic theft may not meet the legal requirements for admissible evidence, and as these examples suggest, they also involve privacy infringement that may be unjustified in civil litigation. But their introduction in courts may influence judicial proceedings. It is hard to disregard such evidence even if decision-makers are told to ignore it.
The costs associated with genetic testing may further intensify power differences among litigants. Because not everyone can pay for DNA sequencing, there is a risk that those with more resources will be "better off" in court proceedings. Simultaneously, the stigma associated with behavioral disorders may intimidate some people enough that they back down from just claims. For example, a good parent may give up a custody claim to avoid disclosure of his or her genetic predispositions for psychiatric conditions. Regulating this area of law is necessary to prevent misuses of scientific technologies and to ensure that powerful actors do not have an unfair advantage over weaker litigants.
Behavioral genetic evidence may also enter the courts through subpoena of data obtained in clinical, research or other commercial genomic settings such as ancestry testing (similar to the genealogy database recently used to identify the Golden State Killer). Although court orders to testify or present evidence are common, their use for obtaining behavioral genetic evidence raises concerns.
One worry is that it may be over-intrusive. Because behavioral genetics are heritable, such data may reveal information not only about the individual litigant but also about other family members who may subsequently be stigmatized as well. And, even if we assume that many people may be willing for their data in genomic databases to be used to identify relatives who committed crimes (e.g., a rapist or a murderer), we can't assume the same for civil litigation, where the public interest in disclosure is far weaker.
Another worry is that it may deter people from participating in activities that society has an interest in advancing, including medical treatment involving genetic testing and genomic research. To address this concern, existing policy provides expanded privacy protections for NIH-funded genomic research by automatically issuing a Certificate of Confidentiality that prohibits disclosure of identifiable information in any Federal, State, or local civil, criminal, and other legal proceedings.
But this policy has limitations. It applies only to specific research settings and does not cover non-NIH funded research or clinical testing. The Certificate's protections can also be waived under certain circumstances. People who volunteer to participate in non-NIH-funded genomic research for the public good may thus find themselves worse-off if embroiled in legal proceedings.
Consider the following: if a parent in a child custody dispute had participated in a genetic study on schizophrenia years earlier, should the genetic results be subpoenaed by the court – and weaponized by the other parent? Public policy should aim to reduce the risks for such individuals. The end of obtaining behavioral genetic evidence cannot, and should not, always justify the means.
When was the last time you made a pro-con list? Carefully considered all factors and weighed them against each other before you made a choice?
Chances are that most of your decisions do not follow this rigorous process. They are made quickly, subconsciously, and often do not adhere to any strict logic. Rather, your decisions are influenced by your mood, your relatives and friends, and a range of other factors that scientists are still unraveling.
When the shoppers were asked why they chose that bottle of wine, almost none of them noticed the music or believed it influenced their decision.
Influencing your choices is also the holy grail of marketing. Companies spend vast amounts of time and money creating product designs and ads. These ads are often tested in focus groups or individual interviews to ensure that they will do well in the market.
Traditional methods of market research rely on self-reports. The participants are asked which ad they find more appealing and why. But there are a few problems with this approach.
For one, the participants might not fully understand their true preferences. They might think that the green design looks more appealing when they compare choices, but then pick up the orange one when they mindlessly wander through the supermarket. It's well known that we humans often do not act rationally, so why would we accurately predict our own behavior?
Another issue is that we like to think of ourselves as logical. Even though our choices are at least partially made subconsciously, we have a tendency to rationalize them after the fact. For example, when supermarkets play French music, the shoppers are 3-4 times more likely to buy French wine. Play German music and German wine sales go up. But when the shoppers are asked why they chose that bottle of wine, almost none of them notice the music or believe it influenced their decision. Instead, they say that they preferred the label or price.
Finally, participants might truly know their preference but choose not to disclose it. Imagine sitting in a focus group watching a TV spot that makes fun of somebody's misfortune. You might be too embarrassed to admit that this is the funnier and more appealing spot, because you're afraid of being judged.
Results from traditional market research are therefore unavoidably subjective and biased.
In the hope of overcoming these limitations, newer ways of market research have been developed, among them neuromarketing, which applies neuroscience to marketing.
Today, neuromarketers focus their efforts on three main stages: to aid product ideation, evaluate the finished product or prototypes, and develop the best marketing strategy. In all cases, they want to find the option with the most "favorable" brain response – but exactly how this brain response is defined varies vastly between studies.
Perhaps the most promising of all non-traditional techniques is functional magnetic resonance imaging (fMRI). This neuroimaging technique measures brain activity indirectly by tracking changes in blood flow. In short, active brain areas receive more oxygen-rich blood. The fMRI scanner picks up the difference between oxygen-rich and oxygen-poor blood and can therefore measure which brain areas are more active than others. But is there truly an untapped potential in the human brain that can be unlocked using neuroimaging?
A number of studies claim that functional neuroimaging has been successfully applied to marketing scenarios. For example, when researchers tried to predict the success of 6 different ads for chocolate bars, the brain response of 18 women was reportedly more predictive than their self-reported preference. The ad that was rated best in interviews was actually the least successful in a real supermarket. In contrast, the neuroimaging algorithm correctly predicted the top two selling ads.
One of the biggest fears is that the potential insights from neuromarketing studies could be used in new, disturbing ways for consumer manipulation.
This study has a number of limitations, which are representative of the majority of neuromarketing research. The field is full of experiments that are conducted with small samples or using suboptimal protocols, with a lack of appropriate control conditions. While a small number of academic researchers are using rigorous protocols, most studies are conducted by neuromarketing companies or funded by the corporations whose products were tested. Such set-ups raise the risk of biased reporting, calling into question the reliability of the findings. Publication bias – the tendency to publish only positive results which leads to a skewing of reported results in the literature – is especially common for industry-funded studies.
One of the biggest fears is that the potential insights from neuromarketing studies could be used in new, disturbing ways for consumer manipulation. If a new product or ad campaign is designed to target our subconscious decision-making better than ever before, are we less able to resist the purchase? We might believe that we all have a healthy amount of self-control, but when we're in the supermarket after a stressful day or we're struggling to manage the self-control of someone else, like a small child, is it ethical for corporations to tap our unconscious decision-making?
As with any technology, the deciding factor is how it will be used. While there are many dangerous applications that might make unhealthy products one day impossible to resist, there are also some more optimistic scenarios. For example, brain scans have been used to predict the success of an antismoking campaign. If such public health interventions that are notoriously ineffective could encourage more people to make healthier lifestyle choices, don't we all benefit? Or is this still a step too far toward manipulation and propaganda?
The conduct of the studies themselves is another problematic area. Academic researchers must go through a rigorous process before they can start a study, which involves review by an ethics board. In contrast, there are barely any regulations for corporate studies. This is not only relevant for the experience of the participants, but also for how the data are being used. Take an extreme case – the brain scan reveals that the participant has a tumor. Universities have protocols in place for how to deal with these situations – often, the scans would be reviewed by a neuro-radiologist and the participant would be informed. Commercial organizations are under no such obligation.
Neuromarketing carries great potential to nudge positive behavioral change, though it also carries the risk of abuse.
Neuromarketing is now a highly competitive field with many different vendors. The Advertising Research Foundation compared 8 vendors that used neuroscientific methods or biometrics for the research of ad campaigns and found that there were differences in methodology and approach; most were proprietary and vendors were not willing to disclose what they measured and how. This lack of transparency is slowing down progress, as researchers cannot contrast and compare different approaches to optimize them.
Despite these methodological challenges, neuromarketing carries great potential to nudge positive behavioral change, though it also carries the risk of abuse. Where one ends and the other starts will need to be clearly defined. It's time to start a public debate now to inform future laws and regulations for the neuromarketing industry, as these technologies will eventually affect us all.