EIF strives to improve children’s life chances and reduce the high costs of late intervention for the government through supporting effective early intervention strategies. However, scientific discoveries in the field of genetics can be presented by the media as a barrier to intervention.
As a first step to addressing the issue, EIF has reviewed the scientific literature on genetics and heritability to test whether the evidence supports this interpretation. We have found that genetics does not prevent effective early intervention and may have the potential, sometime in the future, to usefully inform strategy.
Is the influence of genetics on child development so great as to render early intervention pointless?
It is very clear that the answer to this question is no. There are some rare genetic disorders that can considerably limit the potential for intervention but for the majority of the population and, particularly, for the majority of those for whom early intervention might be considered, genetics creates no substantial barrier to improving outcomes through early intervention. All mental and psychological characteristics, such as intelligence and extraversion, and behaviours, such as parenting and aggression, that have been studied are influenced by both genes and the environment, as well as by interactions between genes and the environment. This raises the question of whether what is currently known can guide intervention approaches.
Can genetics help us improve or better target early intervention strategies?
The emerging science indicates that the answer to this second question may one day be yes but, despite incredible advances in scientific understanding and practice in genetics, the answer as of now is only maybe. The possibilities, while exciting, also raise important ethical concerns, which you can read about in article 6 below.
How can we improve understanding of genetics and the environment?
EIF believes it is important that policy-makers and practitioners engage more directly with these issues. To stimulate this, EIF National Conference 2016 saw delegates from across the range of early intervention-related occupations participating in an ESRC-sponsored session entitled “what can genetics tell us about the possibilities of early intervention?”
Nature or nurture?
Our “nature-nurture” articles (below) explain what makes us the individuals we are, how to critically assess media coverage of the subject and what the implications of the latest research are for early intervention. Further academic papers are set out in the Publications & Resources section of this page.
EIF recognises that policy makers, practitioners and the public are very interested in what the scientific field of genetics has to say about what makes us the people we are. However, we are often confronted with sensationalist and divergent interpretations of the science in the media and the popular press. Understanding the interplay between genes and the environment in child development could dispel myths, improve intervention strategies and enable practitioners to justify the difficult decisions they make.
Children within any population vary in their physical, psychological, cognitive and social characteristics, for example their height, strength, creativity or shyness. The extent to which these population-level differences are due to differences in their genetics (“nature”) rather than differences in their environmental experience (“nurture”) is termed heritability. Whilst heritability is central to our understanding of the relative roles of genes and the environment in development (and the implications for early intervention), it is often misunderstood, which is one of the sources of misinterpretation in the media. (Read more about heritability in article 2 below.)
Genetics has been shown to significantly influence all human characteristics studied so far. These include all key early intervention targets, for example poor mental health, susceptibility to substance abuse and aggression, as well as attributes known to affect life outcomes and social mobility such as intelligence, conscientiousness and resilience. Evidence for this comes mainly from classical “nature-nurture” research: adoption and twin studies. Adoption studies compare the differences between adopted children and their biological and adoptive families. Twin studies compare the differences between identical twins (who share all their genes) to the differences between non-identical twins (who share half their genes, just like ordinary siblings). Learn more about this pivotal research in article 3 below.
Estimates of heritability from twin and adoption studies tell us nothing, however, about which genes affect each characteristic. This would be a more straightforward challenge if each, such as openness or schizophrenia, were influenced by one gene: the “openness gene” or the “schizophrenia gene”, in this instance. Unfortunately, contrary to what is too often proclaimed by the media, no human characteristic fits this simple model; both mental and physical attributes are what are called “complex traits”, likely influenced by hundreds upon hundreds of genes.
Relatively new technology is now allowing whole genomes, that is, all of a person’s genetic information, to be explored at once in relation to complex traits. These genome-wide association studies have been conducted for, among others, schizophrenia, addiction, depression and intelligence. Tiny differences between genomes have been found that go hand in hand with differences found between people for each characteristic or behaviour. The effect of these differences only explains a very small part of the heritability but, when the technology improves, it will have the potential to identify the large set of genes that, together, affect a complex trait. (Read more about this cutting edge science in article 4 below).
A very positive message for early intervention policy makers and practitioners is that the heritability estimates for characteristics of interest are, in all cases, significantly below the 100% that would indicate no environmental contribution. This means that these characteristics could all be amenable to intervention. Despite receiving disproportionate media attention, finding the responsible genes is not useful in itself. What holds promise for informing early intervention strategy is understanding how particular genes interact with the environment to affect personality, behaviour and brain development in general. It is becoming more apparent that some genes only exert influence, or exert more influence, when in combination with certain environments, and vice versa, and that, additionally, genes can affect parts of our environment, making the environment in some way heritable. (Learn more about the fascinating interplay between genes and the environment in article 5 below.) Research into these areas is only just beginning so applying the science to intervention, whilst exciting, is unfortunately a very distant prospect.
The balance between the influences of “nurture” and “nature” on a particular characteristic in a population is often measured as heritability. Heritability is the proportion of the variation in a characteristic found in a population that is accounted for by differences in people’s genetics. The remaining variation is attributable to environmental differences, such as nutrition, disease, experiences and the social environment. It is measured by comparing people’s physical or mental similarities with their genetic similarities, or relatedness (proportion of genes shared), often using twin or adoption studies (read more about this in article 3 below). There are many misconceptions about heritability, some of which are outlined below. These have led to frequent misleading headlines blaming either “nature” or “nurture” for problems in society, which has inevitable knock-on effects for support for early intervention.
Heritability is not a fixed measurement
The heritability estimate for a characteristic will vary depending on the population studied. Two different estimates do not indicate that one is more correct than the other but that the populations studied were different. This accounts for the many various heritability estimates for the same characteristic reported in the literature (although each study will additionally carry its own particular limitations which also affect the estimates – read more in article 3 below). Most behavioural, psychological and personality characteristics have heritability estimates ranging from 30% to 70%.
That there is no one heritability estimate for a characteristic is explained by the fact that different populations do not share the same levels of genetic and environmental variation, which are the drivers of the variation in the characteristic. For example, in a more genetically similar population, there will naturally be less variation in the genes of the population, so more of the differences between people will be due to environmental variation, generating a lower heritability estimate. Similarly, in a population where there is little environmental variation, more of the differences between people will be due to genetic variation and so, for the same characteristic, a higher heritability will be estimated. Because heritability is specific to the population in which it was measured, interpretations must take into account that for different geographical, socioeconomic and differently aged populations, for example, heritability estimates might well be different.
Heritability says nothing about an individual
Heritability is a measure of differences between people not of what shapes an individual person. This is why it is incorrect to assume an 80% heritability of a characteristic means that the characteristic in each person (or, even, in the population as a whole) is about 80% dictated by their genes and 20% by their environment. This becomes very evident if you imagine placing the same individual into a different population where the measured heritability is quite different (see above). Clearly the input of this person’s genes is not now any different just because the characteristic, in the new population, has a different heritability score.
Heritability does not indicate how predictive genes are
Another common misconception is that high measures of heritability suggest that a person’s genetic make-up can predict their development. High heritability does not mean that a characteristic is determined by genes; the environment can change or be manipulated in ways that alter the characteristic. For example, height is highly heritable (usually estimated at about 80%) but, despite the genes present in the population remaining essentially the same over the past few generations, the average height of people has increased from generation to generation (by about 1 cm per decade between 1920 and 1970). This is very relevant for any future application of genetics research to early intervention practices. Ethical concerns surrounding genetic testing for target characteristics are all the more pertinent if the tests might be prone to be considered predictive (read more about the ethical issues here).
Heritability doesn’t answer the “nature-nurture” question!
As celebrated as a solution to the longstanding problem of “is it nature or nurture?” would be, heritability does not provide the answer. Again, heritability says nothing about the role of “nature” and “nurture” in making us who we are, it can only describe their contribution to the population variation. This becomes easier to appreciate when you consider extreme questions. For example, is having two feet due to genes or the environment? A tiny proportion of babies are born with missing feet and a small percentage of the population undergoes foot amputation later in life. Both of these incidences are due to environmental reasons (birth defects caused by exposure to certain viruses, medications and chemicals in the womb and surgery). Since all the population variation in the number of feet is due to the environment, the heritability of foot number is 0%. However it is quite apparent that, for the vast majority of people, genes are the sole influence in determining how many feet we have; the human genome contains the code for us to have two feet.
In fact, if you consider that each of us is infinitely more similar to all other humans than we are to any other animal, then it is clear that all our characteristics are principally genetically influenced (as humans evolved, our genes gradually became different from those of our ancestors and other animals). It is only because we are interested in the differences between people that we are weighing up the contributions of genes and the environment at all. So it does not make sense to consider the relative contributions of genes and the environment to variation among people, the heritability, as an answer to questions of “nature or nurture”. With current understanding, the entire “nature-nurture” debate is, in truth, very much an obsolete and meaningless concept.
Most of the information we have about the heritability of personality and other mental characteristics comes from twin and adoption studies. These naturally occurring experimental situations are very useful as they control, to a certain extent (although see criticisms below), the environmental and genetic influences. This allows their relative contributions to be calculated.
Identical and non-identical twins share different amounts of genes (100% and 50%, respectively) but they share, to the same degree, childhood environments. So any greater similarities between identical twins, as compared to between non-identical twins, are considered to be caused by genes.
Calculating heritability from twin studies:
Based on the assumption that identical twins share 100% of their genes but non-identical twins share only half their genes, if a characteristic were only influenced by genes, all identical twin pairs would correlate for that characteristic while only half the non-identical twin pairs would be expected to correlate. So heritability is twice the difference between the identical and non-identical twins’ correlation for a characteristic. As a fabricated example, if 80% of identical twin pairs show equal aggression compared to only 60% of non-identical twin pairs, there would be a difference of 20%. The heritability estimate for the characteristic of aggression in this case would be approximately double that: 40%.
Twin studies can also estimate the relative influences of the shared environment and non-shared environment. This can be important for early intervention strategists as it indicates which environment might most effectively be altered.
- individual-specific experiences
- contributes equally to the differences between twins in identical and non-identical pairs
The similarities between identical twins are due to genes and shared environment. Therefore the proportion of identical twin pairs who do not share the characteristic represents the amount of variation in the population of the characteristic attributable to the non-shared environment: non-shared environmental contribution = 100% – % correlation in identical twins or, using the fabricated example above, non-shared environmental contribution = 100% – 80% = 20%.
- affects both twins
- contributes equally to the correlation of a characteristic in identical and non-identical twins
- prenatal environment (for example maternal nutrition or stress during pregnancy) and the home environment (for example parental attitudes and economic circumstances)
Once the heritability of a characteristic within the population has been calculated (as above), then the shared environment accounts for the remaining population variation in the characteristic: shared environmental contribution = % correlation in identical twins – % heritability or, using the fabricated example above, shared environmental contribution = 80% – 40% = 40%.
Adopted children share different amounts of genes with their biological parents and adoptive parents (50% and 0%, respectively) but all the environmental influences come from their adoptive parents. So any greater similarities between adopted children and their adoptive parents, as compared to between adopted children and their biological parents, are attributable to their environment.
It would be inconceivable to conduct behavioural genetic research on humans, for example by raising children in controlled laboratory environments. So, for nearly 100 years, twin and adoption studies have provided the best method of estimating the relative contribution of genes and environment to human variation within the population. Twin studies are considered so important that a number of countries have set up ambitious twin databases, including extensive information collected from thousands of twins and, in some cases, biological samples and complete genome sequencing. However, there are a number of criticisms of twin and adoption studies and some people condemn the over-reliance of behavioural genetics on this research.
Criticisms of twin and adoption studies
One of the main limitations of twin and, even more, adoption studies is that it is difficult to recruit enough twins and adoptees (and their families) to make the results convincing. Other weaknesses of this type of research give biased results leading to under or over-estimates of heritability.
Drawbacks leading to overestimates of heritability:
- identical twins might have a more similar social environment to each other than non-identical twins due to others’ perception of them and being treated as a pair rather than as individuals (this is the principle criticism of twin studies and some scientists believe it invalidates all twin study-derived heritability measures (which is most of them))
- gene-environment interactions mask environmental influence (where the environmental factor is shared, that is, the home or prenatal environment affecting both twins)
- many adoptees are not adopted at birth so, in addition to the prenatal environment, they can have varying degrees of environmental contribution from biological parents in their early life
- range of environments is narrower than that of the population
- twins are not treated as normal siblings and also often face early difficulties such as preterm birth and developmental delays
- adoptive families are more likely to be well-adjusted and affluent
- biological families of adoptees are more likely to suffer from social, psychological and medical problems
Drawbacks leading to underestimates of heritability:
- identical twins do not have completely identical genomes (imprinting, copy number, mobile DNA, mitochondrial DNA, mutations)
- gene-environment interactions mask genetic influence (where the environmental factor is non-shared, that is, experiences, usually outside the home, affecting only one twin)
- parents might pick each other based on similarities, particularly, for example, with regard to intelligence or, in some cultures, parents might be cousins or otherwise related, meaning their children, including non-identical twins, will share more than 50% of their genes
- characteristics perceived to be undesirable (such as aggression or lack of conscientiousness) might be under-reported (not admitted to) in studies
- adoptive families are not randomly selected for each adoptee and, in some cases, might share environmental similarities (geographical area, ethnic community, religious values) or, in some very poorly-controlled studies, might even be relatives of the biological family
Some of these criticisms have been addressed by additional related research. This includes: studies of siblings (siblings have been found to share personality characteristics to the same extent as non-identical twins, showing that twins are adequately representative); studies of mischaracterised twins (twins that were mistakenly thought to be non-identical were not less alike than had they been raised as identical twins, and vice versa, showing that twins are raised in comparable environments regardless of their twin type); simultaneous adoption and twin studies (each is based on different assumptions yet their results are very similar); plus studies of relatives, children coming from donated eggs and twins adopted apart, all of which break down some of the assumptions of twin and adoption studies.
GCTA as a “molecular twin study”
A new alternative approach to estimating heritability comes in the form of genome-wide complex trait analyses, which examine the molecular genetic data of unrelated people with a measured characteristic. The similarity of their genomes (proportion of SNPs shared (read more about this in article 4 below)) indicates the (albeit extremely low) “relatedness” of individuals (number of genes shared). This information can be used in the same way that the known relatedness of identical and non-identical twins (100% and 50%, respectively) and their similarity for a particular characteristic is used to calculate heritability in twin studies. By analysing a very large number of people, the heritability of the characteristic in question can be determined; the more heritable it is, the more it will be shared among individuals who are more “related”. One of the main advantages of this type of research is that the number of individuals included in the analysis is not limited by the need to find recruitable twins and adoptees, although the statistical methods involved require thousands if not hundreds of thousands of, albeit unrelated, people. However, heritability results can easily be overestimated if those included in the study do not accurately reflect the population as a whole.
Single gene versus complex trait
Each of our genes carries the information required to make a protein in our body. A rare abnormal variation in one gene can lead to the person suffering from a disease such as cystic fibrosis (faulty CFTR gene) or Duchenne’s muscular dystrophy (faulty dystrophin gene). But most of the millions of normal variations in our genes which make us different from each other lead to small changes in proteins (or big changes in less crucial proteins). These changes can affect our appearance or behaviour. For example, a slightly imperfect version of a particular protein that helps produce melanin, the pigment which darkens our skin, might lead to slightly lighter skin. But there are very few single gene changes that give rise to discernible characteristics. Using the same example, many different proteins contribute to the melanin production process, so it is a combination of all the variations in the genes that code for these proteins which results in the particular tone of our skin. Apart from blood type (AB, A, B or O), virtually all characteristics that vary between humans are complex traits, influenced by many genes (and that includes many characteristics commonly, but wrongly, thought to be determined by a single gene such as eye colour, tongue rolling and chin clefts!).
Candidate gene research
One approach to finding the genes responsible for complex traits is to investigate genes that seem to be “likely suspects”. These genes would code for proteins which function in biological processes relevant to the trait in question. For example, scientists might investigate the link between intelligence and a gene known to be involved in brain development, or between alcoholism susceptibility and a gene affecting how the body deals with alcohol. This type of research typically involves identifying the different forms (called polymorphisms) of the candidate gene found in the population, usually containing just small differences in the genetic code. Association studies are then performed to find out whether people similar for a characteristic are more likely to have a particular version of the candidate gene. It is these types of studies that have a tendency to be reported in the media to have discovered a “happiness gene” or a “smart gene”. Such declarations are always exaggerated, as the genes found make tiny contributions to the heritability of the characteristic, and often premature, since findings that cannot be repeated by different researchers, which is frequently the case, are unconvincing.
Genome-wide association studies (GWAS)
To search for the genes that influence a complex trait like a personality characteristic, which could well be into the hundreds or thousands, as many genes as possible need to be investigated at the same time. Genome-wide association studies are currently the best way to achieve this. Millions of tiny variations in our genes (called SNPs) have been identified that occur commonly in the population (in more than 1 in 100 people). In two individuals, some SNPs will be the same, some different. The more related two people are, the more SNPs they will share. Genome-wide association studies compare all the SNPs across a large number of people. They can then identify the SNPs which are more often found in people with a certain characteristic, for example, people who suffer from depression. The whole team of SNPs influencing a characteristic (this is called a polygenic score), together, can then be used as a predictor of the characteristic. For the same example, it would mean that a baby’s genome could be used to predict whether or not they are likely to be susceptible to suffering from depression later in life. Clearly this could be useful information for early intervention practitioners, who could endeavour to minimise the contributing environmental influences in a highly targeted portion of the population.
So far, however, this approach has had very limited success. The influential SNPs discovered have only explained a very small proportion of the differences among people (sometimes called the “missing heritability” – see below). One explanation is that genes do not always contribute to a characteristic little by little, some counteract each other’s contribution. This effect, called epistasis, could mean that genes which are actually very important contributors to a characteristic, might not be easily discovered because, in many of the people that have that gene, there is a protective or shielding effect from other genes. There is also the masking effect of the environment with genes that are only influential when combined with certain environments (read more in article 5 below), which would make genome-wide correlation studies less powerful. The current technology also only looks at particular types of genetic differences (for example SNPs) and, then, only the common ones (found in more than 1 of every 100 people) whereas there are other potentially important types of variation in the genome and many scientists believe that the under-investigated rare genetic differences could play important roles. The next generation of genome studies will compare the codes for the entire genomes of individuals rather than just predefined points of the code where common differences have been found, which should answer some of these questions.
The “missing heritability” of a characteristic is the discrepancy between its heritability as estimated from traditional behavioural genetic methods, such as twin studies, and the heritability accounted for by the DNA variations found to be influential in genome-wide association studies (often around 1%, so regularly 50 times smaller than expected!) It is often used to criticise, both genome-wide association studies for failing to find influential genes, and twin studies for overestimating heritability.
As described here, there are several reasons why genome-wide association studies are currently under-detecting the DNA variations that contribute to the heritability of characteristics. Much higher estimates of heritability come from studies that consider all the genome’s SNPs simultaneously to determine to what extent people with more similar genomes are more similar in a particular characteristic. This suggests that future, more powerful genome-wide association studies to identify influential genes will partly reduce the “missing heritability” gap.
In reality, there is much more funding for genome-wide association studies of complex trait diseases (such as asthma and diabetes) than of complex traits that are relevant to early intervention. Since research has so far failed to uncover useful gene targets for even these diseases, it is unlikely that we will have a clear picture of the genes that influence the early intervention-relevant characteristics in the near future. In terms of early intervention strategy, this means that we are too far away from being able to predict, by testing a person’s DNA, whether their life outcomes might be positively affected by early intervention, to even say whether that will ever be a practical reality.
The most promising area of research for impacting early intervention is the interplay between genes and the environment. Appreciating the role of genetics in child development in conjunction with the role of the environment could, in the future, allow interventions to be both be better targeted (identifying who is at most risk and in what circumstances) and better designed (identifying the best environmental changes to make). Some of the ways the environment and genes can interact are explained below.
Genes influence the environment (gene-environment correlation – rGE)
Gene-environment correlation is based on the concept that people shape their own environment through their behaviour and the choices that they make. These, in turn, are influenced to some degree by their genes. It is important to consider this connection because it both affects measurements of heritability (leading to underestimates of the contribution of genes to the variation between people) and also, crucially, could inform intervention strategies. This is particularly the case with the parenting environment, which is a focus for early intervention (read more about the EIF’s research on the parent-child relationship here).
One form of gene-environment correlation is individuals intentionally selecting their own environment (known as active rGE). For example, a child with behavioural problems might purposely seek conflict with parents or other children, which leads to their experiencing harsh parenting or social isolation, or a person with high novelty-seeking behaviour might deliberately choose friends and surroundings more likely to introduce them to alcohol or drugs. Recognising that the environments influencing their life outcomes might, in reality, be a product of their genetics, should help focus intervention strategies.
A related type of gene-environment correlation (called evocative rGE) involves an individual’s genetic make-up influencing how others treat them. A child who is extroverted and engaging is likely to gain a more positive reaction from people than a child who is shy and fearful and so their social environments will be differently nurturing, in part, therefore, due to their genes. Likewise, a child whose character makes parenting them more challenging, might negatively affect the feelings, discipline approaches and marital harmony of their parents. Understanding the gene-environment correlation in this case might help steer intervention efforts; equipping parents with techniques to recognise and cope with the child’s challenging behaviours might have a positive knock-on effect on the child’s parenting environment that is actually more effective or economical than attempting to change the parenting directly.
It is also important for practitioners to be aware of the third category of gene-environment correlation (called passive rGE), which refers to a child’s behaviour that is genetically influenced being attributed to environmental causes, because the child’s parents provide both the genes and the environment conducive to development of that behaviour. This gene-environment correlation does not involve a child’s genes influencing their environment and so is unlikely to be informative for intervention strategy, but can lead to misidentification of the cause of the problem.
Genes interact with the environment (GxE)
Perhaps with even more implications for early intervention, a different type of interplay, called gene-environment interaction (GxE) is drawing increasing interest. It involves a person’s genes affecting their sensitivity to a particular environment or vice versa. Essentially, genes and the environment are not always independent in their contribution to child development but can also work in combination. This means that genes can provide protection from an adverse environment or experience, or that a genetic predisposition to a certain behaviour, personality characteristic or psychological disorder might only be realised under specific environmental conditions.
Gene-environment interaction explains why not everyone who faces traumatic experiences, or who is raised in adverse conditions, goes on to suffer poor mental health or life outcomes. People’s differing genetic susceptibility to particular environments also accounts for why some people respond better to intervention than others. Advancing understanding of which genes and environments interact, and exactly how, has the potential to better the targeting of interventions and improve the tailoring of those interventions to the individual.
The “how” of gene-environment interaction is increasingly thought to be attributable to a relatively new branch of genetics, called epigenetics. Our DNA can be altered by the cells of our body, not by changing the genetic code itself that we inherited from our father’s sperm cell and our mother’s egg cell, but by changing the “visibility” of the code (epigenetics). Some alterations to our DNA (called epigenetic tags) can compact the DNA, folding, winding and squashing it together so that the cell can no longer “see” it. Other epigenetic tags can unravel and expose DNA to make it very easily “visible” (and, naturally, there are also plenty of situations in between). The cell will be able to make proteins (which make up and operate, or create other things that make up and operate, the whole of our body) from only the genes on exposed DNA, and the more exposed the more readily this will happen. This phenomenon is called gene expression: high expression of a gene makes lots of the protein and low expression very little.
Epigenetics explains a lot about how our genome, a long piece of code essentially the same in each one of the trillions of cells in our body, can be put to use so differently in different cells. Epigenetic modifications to the DNA in a stomach lining cell, for example, would give high expression of genes encoding proteins useful for digestion (digestive enzymes) and low or zero expression of genes encoding the proteins used for communication between brain cells (neurotransmitter receptors), and vice versa. Importantly, with respect to early intervention, epigenetics also is beginning to explain how the same genetic code can have different effects at different times of our life and in different ways in different people. Research with animals and some human studies have suggested that methylation, an epigenetic tag that decreases gene expression (“silences” genes), can be caused by environmental experiences, particularly early in life, and has been linked to anxiety, schizophrenia and obesity.
The more we discover, the more complex the “nature-nurture” puzzle becomes
The research into these ways genes and the environment combine to influence our development is likely just scratching the surface of the complexity of the problem. How our genetic information is used in our bodies (epigenetics) influences the development of our brain, and our brain influences our choices in life, which shapes our environment (rGE). This, in turn, could affect the way our genetic information is used, as well as further impacting our behaviour, psychology and life outcomes. The situation is further complicated when you remember that our environment is made up of other people (how a person develops and their life choices also affects other people and their development) and also that the particular political, cultural and economic circumstances into which each person is born are also important. In fact, some experts believe there are so very many interdependent influences on our development in play that we can’t learn very much at all from the research so far, where studies have generally concentrated on one influence at time and there have been few attempts to integrate the results of these separate studies. It will be a future challenge to further understand what makes us who we are and to determine which interventions will be most effective by simultaneously considering as many of these influences as possible. (For a detailed exploration of how current thinking and research is not even close to solving the “nature-nurture” puzzle, read this thought-provoking paper.)
Whilst neither genes nor gene combinations that influence personality and other mental characteristics have yet been unquestionably identified, the many ethical issues that surround this controversial area of research cannot be ignored; science will inevitably progress.
With respect to early intervention, the benefits of finding the genetic contribution to the relevant characteristics are rooted in how they might interact with the associated environmental factors so that these might be better manipulated. Other welcome genetic discoveries would be those where a medical intervention might be possible, such as finding an influential gene for schizophrenia, which might be targeted by a drug, and those where proving a genetic link might reduce stigma, for example, relating to sexuality or mental illness. However, what has been learnt so far suggests that each contributing gene has only a very small effect, making effective medical therapy unlikely, and that all personality characteristics are influenced significantly by both genes and the environment, a fact unlikely to greatly dispel stigma.
Ethical problems might arise simply from progress in the identification of genes involved in personality formation, before the science has even influenced medicine or policy. A genetic cause for a characteristic that can lead to poor life outcomes might well relieve society of responsibility for those outcomes and lessen the drive for improvement, widening the social divide and increasing poverty and suffering. Additionally, the reputation of DNA evidence (genes) for being accepted as “scientific fact”, arguably in contrast to social sciences data, might focus the spotlight away from the environmental contributing factors thereby diverting crucial funding and policy attention. These problems may be partly avoided by responsible reporting of the research. Unfortunately, this has already proved lacking with an absurd disparity between how little the genetics of mental characteristics are currently known and the journalistic culture of “gene for x discovered!”
Possibly the most feared policy response to progression in this field is eugenics. Extreme eugenic practices such as enforced sterilisation that persisted in some countries until the 1960s are, today, certainly ethically unsound. However the advancement of in vitro fertilisation and prenatal diagnosis technologies leads to questions about the genetics of mental characteristics being used for “designer babies” and selective abortions, respectively. “Designer babies” in the reality of the foreseeable future, concerns embryo selection based on preimplantation genetic diagnosis, so can be thought of, rather than as “designing” embryos, as choosing to “give” life to certain (“natural” and extremely early) embryos over others. It is therefore perhaps more ethically acceptable than prenatal diagnosis which involves the “taking away” of life from unwanted foetuses based on their genetics. The Nuffield Council on Bioethics concludes that this action of a parent is “not the unconditional, loving acceptance of whatever child one turns out to have”.
If DNA testing for personality and other mental characteristics ever became possible or, even less likely, commonplace, there would be numerous everyday ethical concerns to address. Most widespread might well be an increase in negative discrimination, including in employment, education, policing and insurance. There is additionally the problem of self-discrimination, whereby the genetic “diagnosis” of a person’s predispositions, might result in a fatalistic attitude and increase the influence of genes over environment.
Mental characteristics are usually widely varying between individuals which presents the problem of determining which behaviour lies outside the normal range and requires intervention. Once a genetic influence on a characteristic has been identified, a trend could begin for medicalising the characteristic and “treating” people inside the normal range. With the potential advent of treatments to target the genes responsible, the distinction of therapy (for disorders) versus enhancement is liable to blur. Treatments might be environmental, pharmaceutical or, in the distant future, genetic (gene therapy) but any of these might be a cause for concern, especially because the principal targets would be children, a highly vulnerable section of society. How tests would be provided, controlled and acted upon would be a very contentious issue; government issued testing might be considered a symptom of a “nanny state” whereas privately available testing might only serve to increase the advantages of the privileged and lessen social mobility.
Justice is an area of ethical consideration that is already being affected by advancements in the field of behavioural genetics. Preventative incarceration of individuals genetically susceptible to engaging in criminal activity would be an unthinkable use of this research. Genes, just like environmental factors, cannot be used as predictors even if they can be shown to increase the likelihood of committing a crime, in the same way that, at the very extreme, although having a Y chromosome makes a person twenty times as likely to end up in prison, we do not detain all boys at birth. However, gene discovery might raise questions about diminished responsibility with regard to sentencing, just as social and environmental factors are currently taken into account. An under-functioning monoamine oxidase A (MAOA) gene, which has been linked to aggressive behaviour, particularly in combination with certain adverse environments such as childhood abuse, has been used for this purpose, sometimes with success, in the defence cases for numerous violent defendants.
The EIF is very grateful to the Economic and Social Research Council (ESRC) for funding this project. We would also like to thank our panel of academic experts in genetics and social sciences who have helped quality assure and guide our work.
The main body of the work has been conducted by Rachel Latham (University of Sussex), Steve Peck (University of Michigan-Ann Arbor), Sarah Farmer and EIF staff Leon Feinstein and Leslie Gutman.