Today’s Nature has an excellent feature article on the heated scientific debates over why its so hard to link genes to specific mental illnesses.
Genetics is a complex business, but psychiatric genetics even more so, because it attempts to find links between two completely different levels of description.
Genes are defined on the neurobiological level, while psychiatric diagnoses are defined on the phenomenological level – in other words, verbal descriptions of behaviour, or verbal descriptions of what it is like to have certain mental states.
There is no guarantee, and in many people’s opinion, probably no likelihood, that these ‘what it is like’ descriptions actually clearly demarcate distinct processes at the biological level.
It’s a bit like classifying people as heavy metal fans if they have five or more heavy metal albums.
By definition, there’s a biological difference between people who like heavy metal and those who don’t, but it could be a whole number of distinct differences at the level of brain function which are all just recognised as ‘being a heavy metal fan’ in day-to-day life.
Actually, psychiatric diagnosis has an additional problem, in that for some diagnoses, the same classification can be made when the people don’t share any symptoms. For example, two people could be classified as having schizophrenia / being a heavy metal fan, when they have no symptoms / albums in common.
Some psychiatric geneticists just argue that we don’t have enough data yet, because it seems that when connecting genes to psychology each gene contributes very little and the effect is when the influence of many small effect genes add up and interact.
Others argue that we should look for effects on ‘endophenotypes‘ – the cognitive building blocks of more complex mental life. So instead of trying to connect genes to a collection of ‘what it is like’ experiences, we look at how genes influence neuropsychological processes – such as the mechanisms in the prefrontal cortex that control attention.
Increasingly, some researchers are starting to suggest that the genetic results show that existing psychiatric classifications are invalid, and that we should rethink them as new data comes in.
One thing psychiatry has traditionally been very bad at though, is refining diagnoses on the basis of lab studies.
Definitions are often revised to make them statistically more reliable (i.e. so people can reliably agree what is and what isn’t a particular diagnosis), but this is not the same as having something which is a good basis for scientific enquiry.
Unfortunately, psychiatry is (ironically) a bit too emotionally attached to the traditional diagnostic categories because diagnosis is such a core part of what psychiatrists do.
Anyway, the Nature piece is an excellent guide to the debate on whether we should be attempting to link genes to the neuropsychology of mental disorder.
Link to article ‘Psychiatric genetics: The brains of the family’.