May 21 2013

The Genetics of Mental Illness

The new Diagnostics and Statistical Manual, DSM-V, is out. Not surprisingly, it has sparked some controversy. Psychiatry deniers are proclaiming that this is the collapse of the mental-illness fraud (I believe reports of the death of psychiatry are exaggerated).

What the DSM-V does represent, to some degree, is an attempt to advance psychiatry to the next stage of our understanding of illness. It seems that we are not quite ready for this step in psychiatry, but the effort is sincere and interesting.

For background, the DSM (now in the fifth edition) is essentially a list of official psychiatric diagnoses, based upon clinical criteria. For mental illness and disorders we mostly lack clear biological markers or pathology, and so we have had to make do with clinical descriptions – lists of signs and symptoms. This is very much a descriptive phase of scientific understanding.

What almost every popular article I read on the subject gets terribly wrong, however, is in characterizing this as a unique feature of psychiatry, unlike the rest of medicine. A recent Wired article, for example, writes:

In most areas of medicine, diagnoses are based on the cause of illness. Heartburn and heart attacks both cause chest pain, but they’re different diagnoses because they have different underlying causes.

At least they added the qualifier “most”, but even that is misleading.

In fact most disorders and medical illnesses begin their life as a description of signs and symptoms – a purely clinically defined entity. Scientists then investigate possible causes, with the full spectrum of success. For some illnesses we have very little idea, nothing but guesses, about the cause and pathology. In others we have a completely fleshed out model of what is happening, down to the most reductionist level.

The Wired article also notes:

What doctors now diagnose as schizophrenia may in fact be several disorders with different causes that happen to produce an overlapping set of symptoms.

True, but his is also true of many medical illnesses. ALS (Lou Gehrig’s disease, which causes progressive weakness), for example, is a clinical syndrome. We don’t know the ultimate cause, so it is entirely defined clinically. It is very likely to be multiple pathophysiological diseases with a common manifestation.

Migraine headaches are another favorite example. They are diagnosed by a list of symptoms, just like DSM diagnoses. Migraine is likely many different underlying biological entities that all manifest in a similar fashion. It is also possible that underlying biological traits manifest in some people as classic migraines, in others as a different type of headache, and in still others with no symptoms.

Medical diagnoses span the entire spectrum from a pure description of clinical features, to some knowledge of mechanism, to fairly complete pathophysiological description.  Mental illness is not unique for being at the clinical description end of the spectrum.

What neuroscientists focusing on mental illness are seriously trying to do is advance psychiatric diagnoses to the next step, from pure clinical description to at least classification by underlying mechanism. No one thinks this is going to be easy. The brain is very complex, and the higher cognitive manifestations of the brain are subject to a host of influences. Teasing apart those influences to determine their relative contribution to a mental disorder is a herculean task, but not impossible.

Possible influence include genes, epigenetic factors, developmental factors, biological factors such as nutrition, and all possible environmental factors.

Scientists understand this complexity, but you would not know that from reading many popular treatments of the current DSM and efforts to advance our understanding of mental illness. For example, a recent Slate article, Double Inanity, claims right in the subtitle that “twin studies are pretty much useless.” It opens with this howler of a straw man (actually two).

One of the main messages of science over the last couple of decades is that genes are destiny. With every new issue of a psychology journal, it seems that the portion of your life governed purely by your own free will gets smaller and smaller. Genes determine 50 percent of the likelihood that you will vote. Half of your altruism. One-quarter of y our financial decisions.

A rebuttal is Psychology Today points out that no serious scientist claims that genes are destiny. This is just plain nonsense. Further, the author, Brian Palmer, completely confuses the interpretation of genetic influence. They write:

However, the degree of genetic influence varies from trait to trait, the mechanisms of genetic influence are highly complex and dependent on environmental input, and genes alone are never determinative of anything, except perhaps for rare single gene disorders like Huntington’s Disease.

Further, he confuses population data with individual data.

Genetic and environmental explanations of behavior apply only to differences among individuals. It is meaningful to say that 80% of differences in height among individuals in the modern world are associated with genes, the statement that 80% of a single person’s height is explained by genes is completely meaningless.

If neuroscientists are going to go beyond clinical descriptions of mental disorders, which we know are placeholders pending deeper understanding, then we will have to rethink our categorization of mental disorders. This too is not unique to psychiatry.

For example, muscular dystrophies were first described as clinical syndromes – defined by the distribution of weak muscles (so we have fascioscapulohumoral muscular dystrophy, and limb-girdle). It was not until decades later that we started to discover the actual genetic mutations that cause many dystrophies, and this forced a re-categorization along genetic lines. Some clinical diagnoses survived (FSH is, in fact, a discrete genetic disorder) while others didn’t (limb-girdle is a mixture of many genetic disorders).

The hope is that the same will happen in psychiatry. If neuroscientists are successful in deepening our understanding of mental illness, then in 50 or 100 years the list of mental disorders may be entirely different than what it is today. That would be a good thing, not a sign that mental illness is not real.

There are two main efforts to better understand mental illness. Playing off the new information provided by the genome project, some scientists are trying to find specific gene variants that are linked to mental disorders. No one expects to find “the schizophrenia gene.” What scientists are looking for are large numbers of genes that each affect the risk of developing a specific mental disorder (while recognizing that our list of specific mental disorders may not break down the same way that the genetics do).

There has been some success. Autism research, for example, has yielded a large number of associated genetic variants.

With this kind of research, however, we are likely to generate more questions than answers for some time. This is expected, and is not a sign that the entire endeavor is misguided.

The second major effort to redefine mental illness uses the latest (and continually developing) technology to image brain function. If we can identify the modules and networks in the brain that underlie specific mental functions, and then further identify networks that are behaving differently in patients with certain mental disorders, this may provide yet another way to reclassify and understand mental illness.

In this effort neuroscientists are stepping back from the DSM. The DSM is useful clinically, because it describes the problems with which people present. But perhaps it does not reflect the actual underlying brain malfunction. So neuroscientists studying mental illness are trying to think about disorders in terms of basic mental functions, then identify the network in the brain that correlates with that function, and perhaps identify how it is misbehaving in people with a specific deficit.

In order words, rather than thinking about full clinical syndromes, neuroscientists are trying to reduce them to specific and well-defined neurological phenomena. In a recent interview on the SGU with one such researcher, Heather Berlin, she described how in her research of obsessive compulsive disorder she is looking at the brain responses to disgust. OCD patients, it turns out, have an increased disgust response.

Perhaps, then, we may eventually understand a subset of what are now called OCD patients as hyperactive disgust response disorder.


Mental function of the brain is one of the most, if not the most, complex thing that humans study. Researchers who spend their career studying mental function and disorders of mental function know this. They understand the limits of our current understanding and research methods and the complexity of influences involved.

Progress is therefore understandably slow, when compared to some other areas of research. The challenges faced by neuroscientists focusing on mental disorders, however, are not unique. Criticisms of mental diagnoses, while often pointing to legitimate limitations, are also not unique to psychiatry.

There is a subculture of psychiatry denial, however, that utilizes all the methods of denial to argue that the limitations, challenges, and slow progress of psychiatry mean it is not a legitimate science, and some even argue that mental disorders therefore do not exist.

Ironically they are now using the fact that scientists are pushing the study of mental disorders forward by trying to explore the underlying genetics and specific brain functions that are manifesting as what we describe as mental disorders, as evidence that the entire endeavor is flawed and not legitimate.

This is an identical strategy to that of the creationists who argue that changes in evolutionary theory over the last 150 years is evidence that evolution is wrong.

As with the study of evolution, the study of mental illness is a useful scientific paradigm. We are now venturing into a new era of increased genetic and neuroscientific understanding – progressing from the earliest phase of pure clinical description. This is a good thing. It is a sign of scientific robustness – not evidence that the house of cards is collapsing as some would have you believe.

598 responses so far