Brain research is yielding new information about the development of cognitive and behavioral traits in children—and how development can go off track. This is information parents, teachers, and others will use to help with problems from dyslexia to drug abuse. But with this information, often technical, come many questions. Will knowledge from brain scans be used to label children early in life? How will conﬁdentiality be maintained as diverse providers, some of them commercial, compete for patients? Should government pay for scanning all children to identify not only potential problems but also extraordinary potential? These are questions about values and mutual responsibilities in neuroscience —the domain of neuroethics. The authors consider what an ethical framework for applying the fruits of neuroscience to children —a pediatric neuroethics—might look like.
Could any parent peruse recent reports of the American Academy of Pediatrics without a touch of alarm? The Academy found that in 2004 20 percent of American children and adolescents seen in primary care had “psychosocial problems or mental health disorders,” a rate triple that of two decades ago. Speciﬁcally, the Academy survey found that 13 percent of children and adolescents have anxiety disorders, 6 percent have mood disorders, and 2 percent have substance abuse disorders. High-risk behaviors are also on the rise. The 2001 “Youth Risk Behavior Survey” conducted by the Centers for Disease Control and Prevention found that about one third of youths reported episodic heavy drinking and about 10 percent reported a suicide attempt—suicide is the third-leading cause of death among youths 10 to 19 years of age. I (Judy Illes) have a 17-year-old son who is kind, conscientious, and remarkably trustworthy. Yet the reports leave me with the distinct feeling: “I am one of the lucky ones.”
Faced with these realities, parents, educators, health professionals, and others increasingly express the need for a deeper understanding of what causes a child’s difﬁculties, what risks lie ahead, and what steps might prevent worse trouble. Parents and educators also want to be sure children get the beneﬁt of early and accurate identiﬁcation of any special talents and abilities. Answers to their questions are likely to come from the sciences that study the brain, such as neuroimaging, neurology, psychiatry, psychopharmacology, genetics, and educational psychology. Information is coming at a speed that sometimes surprises and sometimes frustrates. Every answer to a question spawns new uncertainties: uncertainties about the reliability of a diagnosis or prognosis, about the choices that new information may imply, and, crucially, about what options are ethical. New questions are raised: What is the physician’s or the parent’s responsibility? What is fair? What are the respective rights of all involved? How is respect for a child’s autonomy achieved? How are society’s interests, codiﬁed in law or custom, served?
These questions are already engaging scientists, health professionals, technologists, philosophers, jurists, and many others. Their deliberations, bridging the ﬁelds of neuroscience and bioethics, have acquired the label “neuroethics,” a ﬁeld that investigates the ethical, legal, and social implications of advances in brain science. To date, neuroethics has focused mostly on implications of neuroscience that apply widely and generally. But when children (and we include all ages from development of the fetus through the older adolescent) are involved, special considerations enter. Some represent instances of general questions, but many are unique to children at various stages of their lives.
TO SCAN OR NOT TO SCAN?
Modern neuroimaging technology is a rapidly growing source of information about the brain and—whether the potential for serious disorders or the possibility of special opportunities is considered—is increasingly being used to guide decisions about fetuses, newborns, children, and adolescents. Although the ethical challenges raised by using imaging this way are only part of pediatric neuroethics, they exemplify the pressing need to understand the potential risks and beneﬁts of the technology.
Would you want a scan that provided information afﬁrming your child’s giftedness—or the possibility of a brain disease preventing cognitive development, but for which there is currently no cure?
If neuroimaging technology could provide a scan of your child’s brain, conﬁrming or predicting a tendency to antisocial behavior or yielding clues about academic struggles, would you want one? Would you want a scan that provided information afﬁrming your child’s giftedness—or the possibility of a brain disease preventing cognitive development, but for which there is currently no cure? That these questions will arise is not mere speculation; advances in imaging of the brain and central nervous system already promise clues of this kind.
IMAGING THE FETUS AND NEONATE
Neuroimaging can greatly beneﬁt both clinical care and our understanding of human behavior and cognition. But new information brings new risks. Among these risks are possibilities for misunderstanding and misuse, stigma, and the consequences of false-positive (concluding that something is present when it is not) and false-negative (concluding something is not present when it is) ﬁndings. The role of pediatric neuroethics should be to foresee such possibilities and implications and to identify potential ethical, legal, and social conﬂicts in ways that make them manageable from the outset.
Fetal Magnetic Resonance Imaging
The exquisite sensitivity and ever-improving capabilities of MRI for imaging the structure of fetal anatomy have made it an attractive adjunct to ultrasound.
One set of ethical and health policy considerations we face today can be seen in the emergence and implications of fetal magnetic resonance imaging (MRI). Abnormalities of the brain and central nervous system affect some 6,000 newborns each year. Sonography (the use of reﬂected sound to discern shapes, also called ultrasound) has been the standard approach to evaluating prenatal anomalies. But the contrast resolution of ultrasound is limited, so some abnormalities appear only subtle and nonspeciﬁc. The exquisite sensitivity and ever-improving capabilities of MRI for imaging the structure of fetal anatomy have made it an attractive adjunct to ultrasound. The ﬁrst reported use of MRI for fetal evaluation was in 1983; since 1996, when ultrafast MRI sequences for the depiction of fetal anatomy were reported, a steady stream of reports suggest the promise of MRI performed at about 20 weeks of gestational age or later. At times, MRI has led to a different diagnosis than ultrasound and has found conditions not easily visualized on ultrasound, including failure of the brain’s corpus callosum to develop, the presence of abnormal cavities in the brain (called porencephaly), and hemorrhage. Ethical guidelines for managing this newly available information about fetal development, patient concerns and rights, and other implications of using fetal MRI to predict neurologic or developmental outcomes have lagged behind the surge in the technology.
In a recent study by researchers at three medical centers, including our own, an initial effort was made to close the gap between the technologic potential of fetal MRI and consideration of patients’ expectations and views. In this multi-institution collaboration, we sought the voice of the patients about their experiences before and after undergoing fetal MRI. All women in the study had had ultrasound scans that raised suspicions of a problem with the fetus. The women were referred for an MRI by various ultrasound physicians, obstetricians, genetic counselors or geneticists, and neurologists. Most of the 43 women, who averaged 36 years of age, were of European-American ancestry, married, well educated, and relatively well-off ﬁnancially. About half were Catholic; 40 percent were Protestant, Jewish, or Hindu; and the rest did not identify with a religion.
Imaging technology such as MRI that can yield far more information about a fetus (like this one at 36 weeks of age) will change the expectations of patients and raise ethical questions for physicians who present and interpret the results and recommend courses of action. © Simon Fraser / Photo Researchers, Inc.
Almost 9 out of 10 women reported being satisﬁed with the follow-up they received from their health care practitioners in helping them deal with the MRI test results. This ﬁnding was encouraging, especially in view of the wide range of practitioners who had made referrals. The MRI affected the way the women viewed their pregnancy, giving them conﬁdence and reassurance, or, in a few cases, conﬁrming bad news that they reported was nonetheless helpful to learn. As one woman put it: “The MRI results conﬁrmed [my doctor’s] diagnosis...Even though the diagnosis of our baby girl did not change and still showed a poor prognosis, it made us come to terms with the devastating realization that we were to lose a loved one, and decisions needed to be made.”
The MRI test results helped the women make reproductive care decisions, including whether to continue the pregnancy. Perhaps most relevant to the clinical value of fetal MRI, half of the women said results led to changes in treatment, surveillance, or other medical planning. Changes based on the MRI results, such as how the baby was delivered and where, are strong evidence of the utility of this relatively costly technology.
Providers must be careful how they communicate results from new technologies to patients and their families. Some of the women surveyed felt they had received conﬂicting results from their providers; some indicated they had been confused or frustrated when the fetal MRI did not produce a clear prognosis.
Although the power of “knowing” helped these mothers with decision-making, preparation, acceptance, and coping, providers must be careful how they communicate results from new technologies to patients and their families. Some of the women surveyed felt they had received conﬂicting results from their providers; some indicated they had been confused or frustrated when the fetal MRI did not produce a clear prognosis. One patient wrote: “We never felt like we had the information needed to make a clear-cut decision. There were always so many ‘ifs’ and ‘unknowns.’ ” As in the case of genetic testing, the implications of MRI for the viability of the fetus or future of the child can be momentous. Paradoxically, in the case of MRI nothing analogous to genetics counseling is required or usually even available. Fetal MRI will continue to affect the way people experience pregnancy, make choices about whether to continue a pregnancy or abort a fetus, and, potentially, label a child in a way that has signiﬁcant lifelong implications. This is clearly an area that merits more research.
Fetal Functional MRI
Unlike MRI, which produces an image of anatomy, functional MRI (fMRI) captures a picture of brain functioning by measuring the relative demand for oxygen of a particular area of the brain. What is called “blood oxygenation level-dependent contrast” in an fMRI image signals the relationships between the activity of the neurons and the response of the blood vessels (arteries carry more or less oxygen to parts of the brain depending on the level of activity in those areas). Among adults, in whom fMRI has been tested extensively, differences in regional microvasculature, neuroanatomy, and resting cerebral blood ﬂow can confound accurate assessment of the timing, location, and level of neural activation. Functional MRI studies of fetuses tend to compound such inaccuracies, because the brains of fetuses are different than adult brains: Fetuses have more gray matter, less white matter, more synapses, different metabolic rates, and smaller skulls. Nonetheless, given the poor predictive value of ultrasound for detecting delayed fetal growth, for example—one of the most common and complex problems in modern obstetrics—fetal fMRI is being advanced as a new noninvasive method for assessing fetal well-being in at-risk pregnancies.
In animal models, fMRI has shown promise for detecting changes in fetal oxygenation as a marker for deﬁcient oxygenation of the blood, a condition called hypoxemia. In normal-term human fetuses, visual, acoustic, and vibrational stimuli produce measurable neuronal responses in the temporal lobes and frontal cortex of the brain. But fetal fMRI studies have generally been of low resolution and have encompassed the entire abdomen of the mother. Newer studies focus on the ﬁeld of view of the fetal brain and exclude most of the mother’s abdomen. Over the past year here at Stanford University, Roland Bammer, Ph.D., Laura Pisani, Ph.D., and Gary Glover, Ph.D., have been using fMRI to obtain fetal images that show responses to various stimuli and have a higher image quality than previously achieved. Challenges still exist in managing the quality of images because of motion of the fetus, and therefore, to minimize movement of the fetus’s head, the studies are performed on women who are at least 36 weeks into their pregnancies. Results from these experiments will not alter management of the women’s pregnancies; the data will be analyzed after the baby is born to see whether fetal fMRI measurements correlate well with the health of the new baby. If so, fetal fMRI promises better timing of delivery in at-risk pregnancies, for example, by alerting physicians to the presence of hypoxemia. When technical improvements such as thinner “slices”—pictures of more cross sections—are incorporated and imaging at earlier gestational ages becomes possible, the diagnostic potential of fetal fMRI may be unparalleled.
What about the prognostic power of this technology? Will imaging information affect a woman’s choice to continue a pregnancy, as can genetic testing?
What about the prognostic power of this technology? Will imaging information affect a woman’s choice to continue a pregnancy, as can genetic testing? In adults, fMRI has been used in many studies to try to identify cognitive traits, personality traits, emotional makeup, incipient brain disorders, and much more. With technology pushing the envelope beyond boundaries considered realistic, at least for now, it is intriguing to speculate whether advanced neurotechnologies will one day permit parents-to-be to check whether a new Mozart or Einstein is developing inside the womb. In addition, even now, ultrasound images are marketed directly to consumers, who can purchase a printout as a souvenir. With whole-body scanning centers that accept self-referral common in some parts of the country, little stands in the way of selling fetal fMRI scans as souvenir keepsakes, too.
Out of the Womb: Diffusion-Tensor Imaging
Another scanning technology, which has beneﬁtted from developments in MRI over the past two decades, is diffusion-tensor imaging (DTI). DTI, which is used not prenatally but after a child’s birth, shows the microscopic structure of the brain and can detect subtle processes not easily seen with conventional MRI. It has raised new issues about the role of scanning in assessing cognitive potential. For example, studies that used DTI have shown how development is altered in disorders such as dyslexia and in certain problems of cognition and behavior. Improvements are continually being made in the amount of information that can be obtained by processing scan results in various ways.
If scientists can now use DTI to assess cognitive potential, then, by implication, they can also identify cognitive impairment—but what does cognitive impairment mean in terms of neuroimaging data?
One important issue is: If scientists can now use DTI to assess cognitive potential, then, by implication, they can also identify cognitive impairment—but what does cognitive impairment mean in terms of neuroimaging data? That is, how should the concept of cognitive impairment or likely cognitive impairment be deﬁned when fMRI is resticted to blood oxygenation information and DTI to showing microstructural change? Indicators of potential developmental disability that may be seen— such as those associated with premature newborns with very low birth weight—certainly may enable families to initiate early interventions and to plan for the future. But what if relevant interventions have not yet been developed to respond to the information provided by DTI? Furthermore, how should false negatives or false positives—potentially equally devastating—be handled ethically?
Beyond problems a baby may be born with, environmental inﬂuences, such as the richness of infant and early childhood experiences or exposure to neurotoxins, clearly affect different systems of the developing brain. For example, researchers at the University of Pennsylvania have shown that cognitive stimulation inﬂuences the development of language, whereas social or emotional nurturing affects the development of memory but not language. The interaction of these variables and how they are revealed biologically, such as on brain scans, will bear on ethical and legal debates about the role of this technology.
ELEMENTARY SCHOOL-AGE CHILDREN AND ADOLESCENTS
Neuroimaging research is racing ahead to explore the brain for both cognitive and behavioral traits. Obviously, for the parents of elementary school-age children and adolescents, both areas are of intense interest.
Pictures of Intelligence and Talent
It seems inevitable that neuroimaging research of school-age children in the laboratory with fMRI and DTI will be translated to the critical area of educational tracking and intervention for children and adolescents. Already, in the open marketplace, the ability of parents to arrange, without a doctor’s referral, for SPECT (single-photon emission computed tomography) scanning for diagnosis of attention deﬁcit hyperactivity disorder and learning disorders is feeding an unregulated industry.
For pediatric neuroethics, the issues are compelling. How might new information from functional neuroimaging experiments change our deﬁnition of normal cognitive functioning in children—or moral responsibility for actions in 16- and 17-year-old adolescents? Should we be concerned about risks associated with scan results that might frighten or discourage young subjects? Once individual differences in functional imaging are better understood, how will scientists deal with individual patterns that appear anomalous as compared with group norms? Obviously, our responses to these questions will need to be worked out before the problem is upon us by examining a wide range of issues. Here, a few examples must sufﬁce:
- Can neuroimaging be used reliably to track and quantify individual academic performance?
- Once the technology is fully validated for measurements from individuals, how might neuroimaging markers predict below- and above-average threshold learning capabilities?
- How might neuroimaging inform decisions about who should receive educational interventions?
- How might neuroimaging guide strategies of intervention that are adjusted for each child? Such strategies might be behavioral or pharmaceutical, or even involve transcranial magnetic stimulation that may boost sluggish performance.
- What are suitable endpoints for re-testing children?
- How should we deﬁne and manage the risks of medicalizing conditions not formerly considered pathologic, such as shyness?
- What ﬁnancial resources should be devoted to public policy responses to these questions? How can we ensure that experts and lay people who provide oversight are free of conﬂict of interest?
This MRI scan shows the frontal section of a healthy infant’s brain. Brain scans may have the potential to give parents clues about the intelligence, talents, and behavioral traits of their child. How might this information be used—and how should it be used—in guiding a child’s education and, where necessary, intervening? © CNRI / Photo Researchers, Inc.
And, of course, the big question is whether we should turn to neurotechnology for making decisions about cognitively normal children at all. Many of these same questions can be asked, with suitable modiﬁcation, about exceptional talent and giftedness identiﬁed at stages of the childhood and adolescent life span. If the day comes when scientists can understand and measure giftedness by using advanced neuroimaging, the effect on learning in children and performance through adulthood could be profound. But these beneﬁts, too, will come with costs and questions. For example, a child with truly special talent may be identiﬁed at an earlier age than ever before and be given the advantage of special educational opportunities. Should all eligible (however deﬁned) children have access to such technology? Should government bear responsibility for the costs of scanning? Will students be preselected for schools or special programs according to their brain scans? An ethical understanding of these thorny questions and the risks and beneﬁts posed by the answers are essential for appropriate application of such promising technology.
Pictures of Behavior
In the laboratory, fMRI can show differences between brain activation patterns of extroversion and introversion, deception and candor. It can show how vivid imagining can lead to false memories. Scanning can reveal the cortical and subcortical regions that are active with anger, decision making, goal-oriented behavior, moral judgment, and emotion: the qualities that deﬁne each of us as a unique person. But what is the appropriateness, and what are the risks, of seeking such neurobiologic information about individuals’ propensity to conduct disorders, suicide, obsessive-compulsive disorders, or alcohol and drug use disorders— especially in at-risk adolescents? Critical ethical thinking about these questions will guard against a new kind of stigma associated with a brain-imaging result suggestive of aggression or addiction as well as conceivable bragging rights in the bold and brash.
In preparing to encompass neuroimaging into the diagnostic and treatment plan, providers will need to explicitly emphasize the whole context of the young patient’s world so that medication does not become the only perceived solution in isolation from other complementary and potentially beneﬁcial ones such as psychotherapy.
From early results of a study our team has recently completed about the receptivity of mental health providers to fMRI, we preliminarily observed that the technology would be a welcome adjunct to traditional methods of clinical diagnosis of adult major depressive disorder. We also found that the use of fMRI would likely improve patients’ willingness to accept and comply with treatment with psychopharmaceuticals. While we must be exceedingly cautious in translating results from adults to children, acceptance of such advanced diagnostic methods might be welcome since, according to the American Academy of Pediatrics, about half the time medical treatment for mental illness is terminated prematurely. A positive effect of neuroimaging on non-pharmaceutical forms of therapy was less evident, but understanding in advance how the technology might affect patients of any age should enable providers to recognize problems early. The results of our study suggest that, in preparing to encompass neuroimaging in the diagnostic and treatment plan, providers will need to explicitly emphasize the whole context of the adult and young patient’s world so that medication does not become the only perceived solution in isolation from other complementary and potentially beneﬁcial ones such as psychotherapy.
Today, fMRI is being used to identify brain changes that may indicate specific types of behavior, as in a recent study showing the activation of certain brain regions in young people in intense romantic relationships (above). Scanning technology might be effective in spotting adolescent behavioral problems from drug abuse to suicidal tendencies. Should there be systematic screening for such problems? © Courtesy of Lucy L. Brown, Albert Einstein College of Medicine
An imperative for pediatric neuroethics is to establish guidelines and an ethical framework for moving neuroimaging and related neurotechnologies from the laboratory into clinical and eventually educational uses. As Howard Gardner and his colleagues have described, a new kind of educator—a neuroeducator—may be needed. New knowledge for early identiﬁcation of suspected disease and effective intervention shows great potential for beneﬁting children and adolescents, but a highly skilled specialist will be needed. It is an altogether different matter, though, whether we really want general screening of the population. In addition to the cost and expense of such screening, there are worrisome possibilities that businesses could begin to seek young mental health customers without adequate controls on the use of information obtained from tests—information that could be misused by mislabeling or proﬁling children and adolescents. Neuroethics will need to carefully consider how proﬁt-seeking can be aligned with standards for appropriateness of procedures, accuracy, conﬁdentiality, and protections against discrimination.
TOWARD A PEDIATRIC NEUROETHICS
Ethical challenges posed by pediatric imaging are only one relatively limited arena of neuroethics, but they serve to highlight the issues that newly minted neuroethicists have begun to encounter. Imaging is moving fast, appears frequently in the media, and tempts a public always hungry for innovation. The ﬁeld has acquired a high-tech halo not easy to ignore. In reviewing the issues raised about children across the prenatal, childhood, and adolescent life span, certain themes emerge that cut across age divisions. To the extent that they do, they may be seen as general issues facing an emerging pediatric neuroethics. Let us look at six of these themes.
This souvenir montage made for parents shows two profiles of their baby: an ultrasound scan performed during the eighth month of pregnancy and a photograph taken immediately after the baby’s birth. New imaging techniques provide much more detail about the brain, but with that detail come ever-more complex ethical challenges. © LADA / Photo Researchers, Inc.
Power of the Image
How much do new forms of neuroimaging data tell about individual brain functioning? A picture may be worth a thousand words, but one must be mindful of the complexity of a brain-imaging experiment, the statistical processing of the data, and professional and cultural attitudes that the experimenter brings to design and interpretation. If apparently normal mood shifts in adolescents resulting from hormonal ﬂuctuations were not sufﬁciently confounding, changes in blood oxygenation (and so scanning data) are also related to gender and day-to-day variations in physiology and blood ﬂow.
What fetal, child, and adolescent populations are appropriate for scanning in the laboratory to reﬁne the technology and what populations are appropriate for outside, real-world applications? When scientists are searching for biomarkers for aggression, for example, a natural temptation would be to use vulnerable children and adolescents who are institutionalized or resident in group homes. Under any circumstances, minors are considered to have limited capacity for making certain decisions; if cooperation by minors who are wards of the state buys them extra privileges, such as 15 extra minutes at the ping-pong table, then their participation is dangerously close to being coerced. All basic principles of the common rules for informed consent, privacy, and conﬁdentiality must be upheld, whether the research is conducted by investigators in the academic or private sector.
How can the risks of false-positive and false-negative results be mitigated? This is an especially compelling question for the developing mind, given that a ﬁnding predictive of dysfunctional behavior or substandard performance can become a self-ﬁlling prophecy.
In focus groups conducted with neuroscientists, thinkers about ethics, and others concerned with priorities and principles for the next generation of imaging neuroethics, our team learned of another concern related to selection of the child population. This one arises from the constraints of the imaging environment (the small bore of the scanners can be claustrophobic and the noise loud), coupled with the inherent squirminess of children. In describing current imaging research that involved children with autism-spectrum disorders, one psychiatrist lamented: “In autism, the spectrum is huge (from low intellectual functioning to high). [The] high functioning are easy to deal with. So technology is very much high functioning oriented —[the] low functioning are being neglected.” This is a real and ethically difﬁcult challenge.
How might neuroimaging information change a child’s life? Will anxious authorities deprive adolescents of the choice to accept or reject drugs and alcohol if a brain or genetic marker is found to predispose them to addiction? Will their sense of moral responsibility be placed at risk or might they abdicate responsibility to their biology to excuse socially unacceptable behavior?
Missed and False Results
How can the risks of false-positive and false-negative results be mitigated? This is an especially compelling question for the developing mind, given that a ﬁnding predictive of dysfunctional behavior or substandard performance can become a self-ﬁlling prophecy.
Resource Demands, Education, and Public Engagement
What will be the response to parents and educators who obtain new information from technology and demand new resources to help them cope? When no useful therapy or other strategy is available, and the respective beneﬁts and risks about the potential for such behavior are not fully understood, should we delay the introduction of technology that provides such knowledge? How do we move to achieve better education and effective public engagement?
An ethical framework that embraces a pragmatic approach to critical thinking will bring us the greatest success in responding to these issues. It will enable us to ask hard questions without the fear that we are making tough issues even more problematic, and ensure an openess and innovation in responding to them. As Dana Brain Alliance member Giorgio Innocenti, M.D., commented to me (Judy Illes) during my recent visit to the Karolinske Institute in Stockholm, Sweden, “All children deserve the right to fully achieve their genetic potential.” Close alignment of ethical values—existing and projected—with the tremendous potential of neuroscience research is a critical step in guaranteeing that right.
In this era of “No child left behind,” we hope for this kind of integrated approach but we must seek clarity about how to achieve it. We cannot leave parents to face the lure of promises and hope advertised in the direct-to-consumer healthcare marketplace alone.
Managing and responding to ethical questions will require close coordination among parents, pediatricians, and educators; emphasis on creating a healthy environment for all children; and clear rules brought about by professional self-regulation and well-deﬁned expectations. In this era of “No child left behind,” we hope for this kind of integrated approach but we must seek clarity about how to achieve it. We cannot leave parents to face the lure of promises and hope advertised in the direct-to-consumer healthcare marketplace alone. Even if the information they obtain from such sources proves reliable, in the absence of any regulation, far too many uncertainties exist about how it will be used.
Today, none of us is vouchsafed an exemption from these matters. When I (Judy Illes) ask my techno-savvy son (or, in a few years, my now-12-year-old daughter), “What are you doing today?” and he replies, “There’s a two-for-one sale over at Neurodating, Inc., in the mall. I was thinking that my new girlfriend and I would go over and get scanned for our functional neurocompatibility,” what shall I say?