• The Nuffield report [1]only considers genetic influences on behaviour in the normal range.
• Individual genes interact with one another and with environmental influences to determine health and behaviour. Simple tests are therefore rarely predictive in any meaningful way.
• Prenatal and pre-implantation diagnosis for behavioural traits should not be permitted.
• Genetic testing for behavioural traits should not be used in employment or insurance decisions.
• Genetic information about behaviour in the normal range does not absolve an individual from moral responsibility for their actions, and is therefore not valid as a legal defence.
• Research in behavioural genetics should continue despite some public concerns (validity, abuse of findings, don't want to know) because it has the potential to advance our understanding of human neurobiology, psychology and behaviour.
• It is important to create safeguards and discuss issues in an open and multi-disciplinary way now, before they are upon us. The aim of the Nuffield report is to provide a framework for such discussions: to "provide an intellectual toolkit to watch this space".
• Because of the perceived power of genetics, particularly in media reports, emphasis on a genetic link may give an opinion spurious legitimacy. There is clearly work to do in demystifying genetics and placing genetics in context: genetics is "only one of the pantheon of tools that make us what we are". If we over-emphasise its place we are prone to run into eugenics, and big trouble.

(Professor Ian Kennedy, the chairman of the council)

• The Nuffield Council on Bioethics, established in 1991, is an independent body that examines ethical questions raised by advances in biology and medicine. It publishes reports to stimulate debate and contribute to policy making.

• The report covers the following behavioural traits, focussing on variation within the normal range: intelligence; personality (including anxiety, shyness and attention seeking); antisocial behaviour; and sexual orientation.
• The report was produced by a working party established in Nov 2000 that met 11 times (in both the US and the UK), commissioned research and held a public consultation (in 2001).

(Professor Martin Richards, member of the working party)

• Concern was expressed during the consultation about whether behavioural genetics research was robust and reliable, and whether it caused environmental factors to be eclipsed and ignored.
• There was a fear of behavioural genetics leading to eugenics (efforts to improve human stock by encouraging certain 'fitter' members of society to have more children and dissuading or preventing others from reproducing). Behavioural genetics has strong roots in the eugenics movement although today's research is not necessarily or commonly eugenic. However it's important that we learn from the past and recognise that information from behavioural genetics can be misused in this way. Historical and philosophical studies can help to draw lines of acceptability.
• Other fears concerned the misuse of information derived from behavioural genetics in employment and educational decisions, increasing stigma and discrimination and in manipulating embryos to select favourable traits.

(Dr Paul Pharoh, member of the working party)

• Several, if not many, genes affect each trait; each having a small effect. And the same genes may affect more than one trait. Environmental influences are significant. Genetic influences on behaviour are less well described than genetic influences on disease.
• Researchers study the influence of genetics on behaviour by: studying the range in populations and comparing groups (e.g. twins) - quantitative genetics; examining specific genes and correlating these with behaviour - molecular genetics; and by manipulating genes in animals, examining the consequences in their behaviour and making inferences as to effects in humans.
• Some clues to genetic influences in behaviour have been discovered (e.g. novelty seeking behavior and dopamine receptors, and anxiety and seratonin receptors). However there has been little replication of findings and no genetic variant has been shown conclusively to influence specific behavioural traits.
• The effect of genes is therefore not inevitable. Genes do not determine behaviour in a fatalistic way and the term 'a gene for', although widely used, is extremely misleading and inappropriate. Irresponsible reporting that misrepresents the results of studies is harmful. For example, despite press claims to the contrary, there is no 'gay gene'.

(Dr Tom Shakespeare, member of the working party)

• There are no genetic tests for behaviour yet and future tests are likely to be of low predictive capacity. Given this there is unlikely to be much cause for concern any time soon. However it is sensible to begin discussion of the ethical issues in advance.
• Misunderstood science and pseudoscience has a history of driving bad policy so there's a danger if we don't investigate and consider applications of novel science.
• Increased genetic understanding might be useful in several ways.
• Gene therapy, where faulty genes are repaired or replaced is only just beginning to be possible for simple genetic diseases and there is no prospect of this being useful in behavioural change. Germline gene therapy is widely viewed as illegitimate and this group endorses that view.
• Environmental change to compensate for genetic predisposition is conceivable (e.g. remedial education for dyslexia, change of diet to overcome shyness). However this is still speculative, although no clear reason why would be problematic.
• Drug development relies on finding drug targets. Genetics will identify factors involved that may represent such targets. However we don't want to move to situations where those with behaviours within the normal range (e.g. shyness) take strong pharmaceuticals. There are concerns about inappropriate marketing to the socially vulnerable.
• Reduction of social complexity to simple models does have some social benefits. For example less parental blame, less judging of alcoholics once they are acknowledged to have an illness. However, if normal range shyness and risk taking are redefined as medical problems there may be decreased tolerance and increased social pressure to conform and accept medical 'treatment' without considering other options.
• Selecting traits in the as yet unborn is particularly controversial for several reasons: fear of consumerism; concern about a move from accepting the random assortment of genes one currently gets in one's baby; and a fear that we are moving to a situation where we can construct a baby from parts.
• Prenatal diagnosis (PND) is only possible currently for well defined genetic diseases. Since the options after obtaining adverse results are morally serious (termination) and genetic tests (that don't yet exist) for behavioural traits are likely not to be very predictive PND is not appropriate for behavioural selection.
• Using pre-implantation genetic diagnosis (PGD) we can only select from among the selection of in vitro fertilised embryos that we have, and even then only for one or two simply defined genetic characteristics. Even this limited selection though may lead to problems of acceptance/ expectation between parents and children. The view of the working group is that PGD should not be extended to include behavioural traits in the normal range such as intelligence, sexual orientation and personality.

Professor Bob Hepple, QC, chairman of working party

• Genetic information about behaviour in the normal range does not absolve an individual from moral responsibility for their actions (we're not talking here about insanity or psychiatrically recognisable personality disorders). Freewill is not compromised by genetics: there should be no genetic fatalism.
• There is no 'gene for' criminality in general, or for likelihood to rape or assault.
• However, along with social and other reports, genetic predispositions could be taken into account when sentencing. There may be genetic susceptibilities (e.g. monoamine oxidase inhibitor (MAOI) deficiency) that in concert with childhood abuse predispose to violent behaviour. Not currently accepted by courts although US defence lawyers representing clients on death row are trying!
• Genetic information, because of its poor predictability of behaviour, should not be demanded by employers. The current psychometric tests are also pretty poor. In any case, consent to genetic testing by employers is unlikely to be freely given, more likely to be under economic or other duress.
• In future there may be educational benefits to understanding genetic components of behaviour, if extra teaching might mitigate.
• The five year moratorium on insurers asking for genetic information should also cover behavioural genetics.

• It is important to engage and inform school children so they can come to their own conclusions. Since they won't read the report perhaps there needs to be support of teachers, who are often anxious about discussing ethics.
• The report is not intended to educate people as to what they should think but to raise awareness and generate debate. Sometimes it's best to come down on one side rather than the other, where arguments appear to be on that side, and then invite criticism.
• The real dangers are not of appearance of actual tests but of over-interpretation of current genetic research. A lot of the aim of this Nuffield group's work was to demystify genetics and explain why it can't be used in behavioural profiling even where it can be used in research. Because of the complex interactions among genes and with environmental factors, genetics is less deterministic than we think.
• It was suggested that for certain medical conditions we should view genetic factors in the same way as biochemical testing (e.g. cholesterol, blood pressure and sugar in urine). However testing on the basis of the actual behaviour is more powerful than testing for genetic influences.
• Why shouldn't one be allowed to select characteristics in a baby if it does the child's health no harm? There might however be consequences for the parent/child relationship. Maybe there will be resentment from the child that they were predefined in some way, that their future was restricted, that they were channeled.
• Non-identical twins show that genetic influences are very important. However there is a fear that emphasis on the genetic component and behaviour that correlates with behavioural genetics may lead to skewed measuring of only some influences on behaviour.

How can access to selection technology be made available to all regardless of income? In a scarce healthcare environment it is hard to justify widespread availability. We may have to accept it is like private education, which we allow.
 

 
1 The report can be downloaded from http://www.nuffieldbioethics.org/publications/pp_0000000015.asp [return]

2 See http://www.nuffieldbioethics.org [return]