Genetics of psychiatric disorders

Standard methods for investigating genetic contribution to disease

Assessing the extent of genetic versus environmental contribution was first done using twin studies and adoption studies. Certain classical modes of transmission, such as X-linked recessive, can be recognised by studying the segregation of diseases through pedigrees.

Genetic markers can be used to localise susceptibility genes to particular chromosomal regions. Linkage analysis identifies markers which segregate with disease through pedigrees and give a broad localisation. Association analysis identifies marker alleles which tend to occur more frequently in cases than controls and give a narrower localisation.

Genome-wide association studies (GWASs) use hundreds of thousands of markers to identify common variants associated with disease. GWAS SNPs are also used to estimate heritability, the proportion of the variance in liability caused by genetic factors.

Copy number variant detection identifies deletions and duplications of chromosomal segments associated with disease.

Whole genome or whole exome sequencing identifies every variant present to see if any are commoner in affected subjects.

Detection of genetic variation associated with a disease can be followed by studies to determine the effect on gene expression and/or functioning in animal models and cell cultures.

Familiality in the major psychoses

Examining the diagnoses which occur in relatives of probands with a given disorder may throw light on the classification of psychiatric illness. For example, in general manic depression and schizophrenia tend to "breed true", supporting the Kraepelinian distinction between these disorders. Relatives of manic depressive probands are also at increased risk of unipolar depression, and relatives of schizophrenic probands have increased rates of schizoid and schizotypal personality disorder, suggesting that the diagnoses may sometimes represent less severe forms of the same underlying disease process. Although it can be difficult to classify puerperal psychosis clinically, the fact that relatives tend to suffer increased rates of affective disorder suggests that in most cases puerperal psychoses are essentially atypical forms of affective psychosis.  

Manic depression / bipolar disorder

Twin and adoption studies indicate a substantial genetic contribution to risk

Although it is possible to find some densely affected pedigrees the mode of transmission is unclear and it is assumed that a number of different genes may contribute to risk.

Linkage studies implicated candidate regions but did not produce conclusive results.

The genes most consistently implicated by GWASs are CACNA1C and ANK3. CACNA1C codes for a subunit of a voltage-gated calcium channel and there are results suggesting the involvement of other ion channel genes. ANK3 is located at the nodes of Ranvier and is involved with the localisation and functioning of sodium channels. Another GWAS implicated gene, NCAN, is involved in neuronal adhesion and neurite growth. Mice knocked out for either ANK3 or NCAN demonstrate mania-related behaviours such as overactivity or impulsivity.

Although there is a genetic contribution to the susceptibility to unipolar depression, it is less marked than for bipolar disease. There is little evidence from linkage or association studies to demonstrate which genes might be involved, and although large GWASs produce significant hits their biological interpretation is unclear.


Twin and adoption studies have demonstrated a genetic influence on predisposition. Also, children of the normal monozygotic cotwins of schizophrenics have a similar risk of schizophrenia as do the children of the schizophrenic parents (supporting a genetic contribution to aetiology). A number of environmental risk are also identified, including maternal influenza, maternal famine, birth trauma, winter birth, cannabis use.

The mode of transmission of schizophrenia is unclear and it is likely that a number of genes are involved.

A number of regions generated positive linkage results, although with little consistency.  

A large Scottish pedigree has been described in which many subjects have a translocation between chromosomes 1 and 11 and suffer from schizophrenia or other psychiatric illness. The breakpoint on chromosome 1 goes through a gene named DISC1 ("Disrupted in schizophrenia 1"), and association studies of DISC1 have been positive in other samples. Abnormalities of DISC1 may interfere with neuronal development.  

A deletion of part of chromosome 22 causes velo-cardio-facial syndrome (VCFS) and some patients with this also suffer from a psychosis indistinguishable from schizophrenia. Indeed in some cases the psychosis is the only abnormality which comes to attention and in a cohort of unselected schizophrenics approximately 1% will turn out to have VCFS. VCFS thus represents a rare genetic cause of schizophrenia. It is possible that genes in this region may influence susceptibility to psychosis in other cases.

Studies of other copy number variants (CNVs - deletions and duplications) show that in addition to the VCFS deletion CNVs are generally more common in subjects with schizophrenia than in controls. CNVs affecting NRXN1 (neurexin 1) are associated with schizophrenia. Deletions at 15q11.2 and duplications at 16p11.2 increase risk. Altogether, CNVs might account for around 1-2% of cases of schizophrenia.

GWAS results indicate a large number of regions associated with a small effects on risk. These effects can be summed to produce a "polygenic risk score", which also shows that some genetic risk factors are shared with other psychiatric disorders.

Exome sequencing studies imply that multiple rare variants are involved but have not identified specific genes definitively. Although cumulative evidence points towards glutamatergic transmission, calcium channels, synaptic functions and histone modification.

C4 (complement component 4) variants imputed from GWAS SNPs show modest but statistically highly significant association between predicted expression of C4A and schizophrenia risk.

Extremely rare loss of function variants of SETD1A produce a high risk of schizophrenia (or intellectual disability). It codes for a component of a methyl transferase for Lys-4 residue of histone 3 (H3-K4). It seems that disruptive variants in a gene called RBM12 can also cause schizophrenia but its function is unknown.


Family, twin and adoption studies have demonstrated a genetic influence on the predisposition to alcoholism. The mode of transmission is unknown and it is not clear whether single gene effects or polygenic effects occur.  

Note the following example of the influence of genetic polymorphism on the development of psychiatric illness:  

A mutation common in orientals inactivates mitochondrial aldehyde dehydrogenase 2 (ALDH2), which is involved in the normal metabolism of ethanol. In such individuals the consumption of small amounts of alcohol produces circulating acetaldehyde leading to unpleasant symptoms (facial flushing, etc.) and this causes them to avoid alcohol. So the mutation which deactivates ALDH2 indirectly protects against the development of alcoholism.  

A variant in the gene for alcohol dehydrogenase 1B causes a gain in function of the enzyme, leading to excess acetaldehyde production and, again, the avoidance of alcohol.

It is expected that there will be other genetic variants which influence susceptibility to alcoholism through psychological mechanisms such as reward-seeking, compulsivity. For example, GABA receptor variants can strongly influence alcohol preference in mice.

Attention deficit hyperactivity disorder (ADHD)

There is good evidence for a genetic contribution to the risk of ADHD. There have been reports that large, rare CNVs are commoner in cases. No specific genes are definitively implicated.

Intellectual disability

Intelligence is in general normally distributed but with a swollen left tail reflecting specific syndromes which can cause profoundly low IQ.

40% of cases have unknown aetiology. Can be divided into neurodegenerative, syndromic and non-syndromic. Many chromosomal (Downs syndrome) and genetic (phenylketonuria) abnormalities cause learning disability.

Note the example of phenylketonuria showing the way in which genes and environment (phenylalanine in the diet) can interact.  

The mutation responsible for the fragile X syndrome has been identified and the gene in which it occurs has been named FMR1 (Fragile-X Mental Retardation). Its transmission is complex, since although the mutation can act as an X-linked recessive it "gets worse" as it passed on through different generations - a trinucleotide repeat sequence (CGGn) enlarges.

There are over 2000 genes in which are variants can produce intellectual disability as part of the phenotype. Increasingly, whole exome and whole genome sequencing are used to identify causes and currently yield diagnoses in about 50% of cases. Severe intellectual disability is often caused by de novo mutations.


Autism was originally described as a combination of impaired social functioning, odd use of language and stereotypical behaviours in subjects with intellectual disability. Subsequently it has been broadened to include subjects without intellectual disability and with milder  deficits. Autistic features can occur as part of a number of defined genetic syndromes. There is an excess of de novo mutations - variants seen in cases but not their parents. Rarely cases are due to mutations affecting the neuroligin or neurexin 1 genes. The autism phenotype can also occur with fragile X syndrome and mutations of MECP2 which more usually cause Rett syndrome. 1% of subjects with autism have a deletion at 16p11.2 and other CNVs have also been identified. Exome sequencing implicates  synaptic formation, transcriptional regulation and chromatin-remodelling pathways. A growing list of specific genes have been implicated including CHD8, EHMT1, ANKRD2, FOXP1, TBR1, RAI1, SYNGAP1, SHANK2, SHANK3. NRXN1, NRXN2, GABRB3, SCN1A, CNTNAP2. These are involved in chromatin remodelling, gene transcription regulation, cell growth and proliferation, ubiquitination, synaptic organization and activity, dendritic morphology and axonogenesis.

Dave Curtis (