Dr Paul Morrison is a private psychiatrist in London with over 20 years experience in the treatment of mental illness. Paul’s clinical philosophy is to create a space in which the patient and their psychiatrist can work together in close collaboration to enhance insight and recovery. Find blogs on psychiatry.
The previous post in this series described the symptoms of schizophrenia. Here we turn to the causes of schizophrenia. There has been major progress in this area over the last twenty years. A number of factors have been identified which carry a risk for schizophrenia. Some of these factors are genetic, others impact during the course of life.
Usually schizophrenia emerges in late adolescence or early adulthood, as the intellect, personality and neural networks are being sculpted. The population risk is slightly less than 1%, with a slight excess of male sufferers (1.4:1). Males also tend to show a more severe pattern of illness, with more impoverishment of the personality and psychological decline.
Risk Factor Categories
The risk factors for schizophrenia can be grouped into several categories (Figure 1). The perinatal category includes hypoxic and nutritional insults to the developing brain in-utero. The second category includes being brought up in city environment, particularly for immigrants. The third category includes drugs of abuse, specifically strong cannabinoid CB1 receptor agonists. Finally, there is the genetic category, which can be subdivided into single nucleotide polymorphisms (SNPs) and copy number variants (CNVs).
Figure 1. Risk factors for schizophrenia.
Genetic risk factors
It has long been recognized that schizophrenia runs in families (Figure 2), but until the last decade attempts to identify specific genes floundered. Technological advances have revolutionized the field. It is now feasible to screen an individual’s DNA at every base pair (A, T, C, G) in every chromosome. Variants (say the substitution of an A for a T) are called single nucleotide polymorphisms (SNPs) when they occur in at least 5% of the overall population.
In the technique known as GWAS (genome wide association study) tens of thousands of patients are compared against tens of thousands of controls. So far, 145 SNPs have been shown to confer risk for schizophrenia (Figure 3). Each SNP on its own carries a very small risk, but they are common in the population, and their effects are additive. Two SNPs of considerable interest are the gene for a calcium channel (CaV1.2) and the gene for a protein called complement C4a.
Figure 3. Single nucleotide polymorphisms which confer a risk for schizophrenia.
The second major breakthrough in schizophrenia genetics are copy number variants (CNVs). Copy number variants are deletions or duplications of a long stretch of DNA, typically incorporating half a dozen genes or so. So far eight CNVs which confer a risk for schizophrenia have been identified. Each of these CNVs carry a very high risk. A CNV of considerable interest is NRXN1 (Figure 4).The NRXN1 protein forms a physical bridge which stabilses synaptic connections in the brain. The NRXN1 story provides strong evidence for a long-held theory that the pathology of schizophrenia stems from abnormal connectivity within neural networks.
Figure 4. Copy number variants which carry a risk for schizophrenia.
We can recap. A number of factors confer risk for the development of schizophrenia. These can be categorized into several categories – perinatal, environmental, cannabinoid CB1 drugs, and genes. The gene category includes SNPs (such as, complement C4a, the calcium channel CaV1.2) and CNVs (such as NRXN1). In the next post in this series we will look at the neurobiology of these components and cannabinoid CB1 drugs.
Schizophrenia has long been the heartland of psychiatry, but can be as confusing now as it was 100 years ago. Lay opinion is that schizophrenia is commensurate with hearing voices and paranoia, but this is not true. Hearing voices and paranoia are non-specific phenomena which can occur in normal psychic life.
So what exactly is schizophrenia? Well it is also not commensurate with psychosis. In fact, there is a long list of conditions in which psychosis can occur (Figure 1).
Psychosis we can define as a fundamental shift in a person’s experience of lived reality, affecting the highest faculties of mental life – perception, thinking, beliefs, self-hood.
Figure 1: There are numerous causes of psychosis, not just schizophrenia.
Lay opinion also makes the mistake in formulating the psychotic shift as breach from consensual reality, the shared reality of the group. It is not. The psychotic shift is a breach from one’s previous way of being in the world. One aspect of psychosis is key. The sufferer is unaware of the falsity of their new reality, a fact which is obvious to friends and family.
But what is schizophrenia?
Again we return to the question, what is schizophrenia? Here we grasp for an answer. The most important point is that schizophrenia involves a loss. Not just a loss of the previous way of perceiving and thinking about the world, but something deeper. The loss encompasses – human relationships, interests, intellectual pursuits, ambition, motivation, emotional life. At its most extreme, it is the spark of mental life itself which is lost (or markedly impoverished) – the personality, drive, speech, thinking, the will.
In modern psychiatry such phenomena are referred to as the negative syndrome, which obviously denotes loss. The negative syndrome maps onto poorer intellectual abilities, across multiple domains – working memory, episodic memory, processing speed, social cognition – and is associated with functional disability in daily life.(Figure 2).
Figure 2: The different domains of schizophrenia.
We can recap. Voices and paranoia are not commensurate with psychosis. Psychosis occurs in a multitude of disorders, not just schizophrenia. Finally, schizophrenia is so disabling because of the loss (or diminution) of all those aspects which make human mental life so special and unique.
Long stretches of DNA, from 1000 to several million base pairs, can be deleted or duplicated within a chromosome. These changes are known as copy number variants (CNVs). It is now recognised that eight CNVs are associated with schizophrenia. For a person carrying a CNV, their risk of developing schizophrenia is increased by 3-58 times compared to the general population. It is also known that approximately 2.5-5% of people suffering from schizophrenia will carry at least one of these CNVs.
A new paper by Danish researchers serves as an excellent introduction to this rapidly developing and fundamental topic. The authors detail the characteristics of each of the eight CNVs.
The effect of CNVs: Neurexin
The majority of the CNVs harbor multiple genes. The exception is deletion of a stretch of DNA which harbors a single gene coding for a protein called neurexin (NRXN)1.
Synapses are held together by cell adhesion molecules, one of which is neurexin (NRXN1). Deletion in the gene for neurexin is a risk factor for schizophrenia and autism.
Researchers are beginning to understand how the the seven other CNVs might operate to confer risk for schizophrenia, and indeed many other psychiatric syndromes. A property known as pleitropy means that a genetic change actually confers risk for a range of disorders. The CNVs associated with schizophrenia are also risk factors for autism, ADHD, epilepsy and intellectual disability. Indeed, all eight schizophrenia CNVs confer risk for autism spectrum disorders (Table 1).
Copy number variants (CNVs) which confer risk for schizophrenia and other neuropsychiatric syndromes.
Using animal models, researchers can decipher how the CNVs alter brain and behviour since DNA stretches are conserved across mammalian species. Animal models of five of the eight schizophrenia CNVs are now available. Those CNVs have been shown to cause deficits in cognition, social behavior, information processing and synaptic plasticity.
Further developments in this area are bound to reveal much more detail about how altered neuronal dynamics give rise to schizophrenia and other major psychiatric syndromes.
Psychiatric genetics can be daunting for the non-expert. But it is so important for all mental health researchers and clinicians to have some understanding of where this field is at. Unlike much of the rest of psychiatric research and theory, modern genetics represents a firm foundation of valid and reliable knowledge. That knowledge is slowly unfurling how we think about psychiatric disorders such as ADHD, autism, depression, OCD, substance abuse, schizophrenia and bipolar.
It has long been known that psychiatric illness runs in families. The heritability of psychiatric disorders (i.e. the degree of variance in a trait in a population which can be explained by genetics alone; a figure between 0 and 1) ranges from 0.3-0.4 for PTSD and depression, up to 0.7-0.8 for ADHD, autism, schizophrenia and bipolar. (Figure 1. Orange diagonal).
For the more genetic disorders (ADHD etc.) susceptibility very rarely comes down to one gene. Far more commonly, hundreds of individual genes are involved. Each gene, on its own, carries a tiny, almost negligible effect, at least in clinical if not statistical terms. But when a collection of risk genes is inherited, the chance of developing a psychiatric disorder starts to increase. For experts in psychiatric genetics, this is known as the polygenic risk score (PRS) the summed value of all the individual risk genes. The PRS is an important measure in modern psychiatric genetics.
Genetics researchers are now turning their attention to how a collection of genetic variants that increase the chances for one disorder (the polygenic risk score, PRS) may also increase the risk for other psychiatric disorders. This is pleitropy at a higher level. The early findings again point to crossover between disorders. Kendler and colleagues elegantly illustrate the headline findings (Figure 1). The light blue squares show the genetic correlation between disorders using the methods of modern molecular genetics. The light orange squares show the genetic correlation between disorders using the more historical methods of family and twin studies.
For the present the main research effort will be to gather and pool more whole genome data from individual patients (and controls). Sample sizes of >100,000 will find more and more gene variants which confer risk for psychiatric disorders. Groups such as the multinational psychiatric genomics consortium (PGC) co-ordinate this task. Data-sets and computing resources are freely available to any researcher. Whether the traditional diagnostic systems collapse completely or remain in a different form cannot be know at present, but with modern molecular genetics, psychiatry is at last on a firm empirical and theoretical ground.
One goal in modern medicine is to use genetics to determine the best treatment strategy for an individual patient. A simple, cheap genetic test which predicts response to a particular treatment has clear utility for patients.
To date, psychiatry has lagged behind general medicine, no doubt due to the complexity of CNS tissue and the sheer number of components involved in CNS processing, but the picture is changing. One particular area of interest is the treatment of alcohol dependence.
The pharmacogenomics of topiramate
The drug topiramate started off as an anticonvulsant, but was also been found to be effective in reducing alcohol consumption in heavy drinkers. An American study has advanced our knowledge. Natural variation in the DNA coding for a type of glutamate receptor called the kainite receptor was shown to determine if problem drinkers responded to topiramate treatment.
In those carrying the CC variant at rs2832407, topiramate (200mg/d) markedly reduced the number of heavy drinking days compared to placebo (Fig. 1). Heavy drinking decreased from 5 days per week to approximately 1 day per week in CC subjects treated with topiramate.
In a 12 week study, heavy drinkers with the CC genotype at rs2832407 in the kainite receptor showed significant reductions in heavy drinking when treated with topiramate.
A larger study is now underway, hoping to confirm the initial genetic findings. If the result is replicated, genetic testing may translate to routine clinical practice for heavy drinkers considering topiramate treatment.
This would represent a significant advance in the management of alcohol disorders.