The Grand Theory

Last week there was a bit of hubbub as a new paper was printed in The Lancet, Identification of risk loci with shared effects on five major psychiatric disorders: a genome-wide analysis. 

Django Django Default

The paper is a dense discussion of the higher maths of a large genome-wide association study. There’s all sorts of statistical wizardry going on with triple pseudocontrols and a number of things that are a bit above my pay grade, with my single statistics class from medical school and my genetics knowledge from the 20^th century. But the gist of the paper is the following: after hacking the genome (using variations in genome markers, called SNPs, or “snips” to use the genetic parlance, meaning “genome-wide single nucleotide polymorphisms”) of many tens of thousands of cases and controls for people with autistic spectrum disorders, schizophrenia, bipolar disorder, major depressive disorder, and attention deficit hyperactivity disorder, several gene clusters were found that seemed to increase the risk of ALL FIVE of those disorders. Most of the genes found to increase the risk of these disorders were located within the coding for two L-type voltage-gated calcium channel subunits.

What? Okay, so we know that psychiatric diagnosis is a mess. We have the DSM (soon to be in its 5^th iteration) which is basically a recipe list of symptoms. Find enough of your symptoms on the list, and voila, a diagnosis. The DSM is atheoretical ON PURPOSE. They make no pretensions at looking for pathology or brain chemistry issues or whatnot because, for the most part, back in 1980 when the DSMIII came out no one knew what the heck the actual pathology was. The earlier renditions did have some “causes” which led to some gems like the “schizophrenogenic mother,” and so instead of making horrible gaffes, the psychiatrists writing the DSMIII (and IV and V) really strove to see if there were clusters of symptoms that seemed to have some diagnostic validity without venturing into causes at all, so that researchers and clinicians and everyone could more or less speak the same language when talking about patients and trying to come up with evidenced-based treatments. The first MRI was built in 1977, and it’s only been since then (and really, since the invention of functional MRIs, advanced genetics with rapid PCR and PET scans and SPECT scans and the like) that we’ve really been able to see a bit more what is going on in a human brain in a living, thinking subject.

One of the frustrating (and great) things about psychiatry (and, frankly, medicine in general) is that you will almost never see a textbook case of anything. Most of my patients who show up at my office for the first time meet criteria for several diagnoses simultaneously. It’s easier to use symptoms as a benchmark rather than the full diagnoses, though for FDA purposes, treatments are researched based on diagnosis, not symptom. More recently genetic studies have shown links between various diagnoses (most notably, I think, bipolar disorder and schizophrenia, and schizophrenia and autism). While the diagnostic categories are quite clear, sometimes it can be quite difficult in practice to distinguish schizophrenia from bipolar disorder (so much so that there is a midway in between diagnosis called schizoaffective disorder), and many antipsychotic medications are approved for both treating the psychosis of schizophrenia and treating the manic episodes of bipolar disorder.

So what the new paper tells us is that these 5 different disorders of adult and childhood-onset diseases may well all be related. Voltage-gated calcium channels are little places in the cell membrane that decide, based on the voltage of the membrane at the time, whether to let a parade of positively charged calcium ions through the membrane. We need the particular calcium channel subunits coded by the genes found in the study for many brain processes, including functions involved with memory, planning, emotional processing and regulation, and attention. Another one of the calcium channels identified by the study helps other calcium channel subunits be turned on and off, whether by helping them travel to the membrane to set up shop, assist in their regulation by other molecules, or increase their function in other ways. 

A commentary on the study published at the same time in The Lancet spoke to the methods used in the study (and some of the weaknesses, geneticists, by all means, pull the papers and have a look). But it also tried to make some sense of the overall picture here. Basically we have many, many interacting factors that convey risk for psychiatric disorders.  

Prenatally, we have infectious disease, drugs, alcohol, nicotine, nutritional deficiencies, maternal stress and disease interacting with the genetic make-up of the baby. The postnatal environment of trauma, living conditions, nutrition, disease, etc can effect epigenetic changes which interact with brain plasticity to influence, along with the genetic make-up, disease vulnerability. So someone with an abnormal calcium channel gene in the wrong place at the wrong time (along with some other genetic and environmental issues) will get symptoms of ADHD. Another person (with the same calcium channel gene issue but different other genes, environment, etc.) will end up with schizophrenia.  Some with the gene will have no psychiatric disorder at all, I’m sure. 

Are we getting closer to upending the DSM? To define disease from the pathology upwards rather from the symptoms down? I hope so.

See also: Toward a Grand Unified Theory of Psychiatry