Schizophrenia Is Actually Eight Different Disorders

Research shows the promise of genetic medicine.

For decades, schizophrenia has been understood as a group of disorders with similar symptoms. Those with paranoid schizophrenia, for example, experience what are called “positive” (or inappropriate compared to normal behavior) symptoms such as delusions and hallucinations. Those with disorganized schizophrenia, another subtype, may be unable to communicate coherently. Still another subtype experiences “negative” (or absent compared to normal behavior) symptoms, such as the inability to express appropriate emotion or take interest in everyday activities.

Now, researchers have found that schizophrenia may be a group of genetically distinct disorders — by mapping their genetic basis. In the past, researchers, using identical twin studies and family histories, have found that schizophrenia is highly inheritable. Current researchers are using DNA studies to identify the basis for inheritance.

In the growing field of bioinformatics, computers are used to analyze the masses of information now available about the human genome, the “map” of the human DNA first completed in 2000. Researchers at Washington University in St. Louis used such techniques to carefully match symptoms of schizophrenia against the DNA of about 4,100 people with the disorder and 3,800 healthy controls culled from the Molecular Genetics of Schizophrenia study.

The researchers looked at the DNA of participants, analyzing nearly 700,000 variations in gene base pairs called single nucleotide polymorphisms (SNPs — or, as scientists call them, “snips”). Since human genes are 99.9 percent identical from person to person, the 0.01 percent of the genome SNPs play an important role in the discovery of the causes of illness and disease.  

The group mapped sets of such variations to the symptoms each participant experienced. They were able to identify eight subclasses of schizophrenia, each with more or less severe positive and negative symptoms, associated with genetic variations.

One set of variations had a 100 percent correlation with schizophrenia — meaning those who have it are guaranteed to develop the disorder. Others had risk levels in the 80 to 90 percent range. In all, the researchers found 42 SNP sequences that indicated a greater than 70 percent risk of schizophrenia.

The researchers then replicated their results using DNA databases from two other major studies of schizophrenia, the Clinical Antipsychotic Trial of Intervention Effectiveness, or CATIE, and the Portuguese Island studies from the Psychiatric Genomics Consortium. The identified SNPs they’ve identified mapped to similar symptoms in all three studies, they found.

Those genetic variations may combine with environmental influences to produce the disorder, says Igor Zwir, PhD, a principal researcher on the project. Zwir is a research associate at Washington University and associate professor in the Department of Computer Science and Artificial Intelligence at the University of Granada in Spain.

The study explains a longstanding mystery in schizophrenia research: why efforts to tie the disorder to a single gene have never worked in the past. Instead, the researchers say, it is clusters of gene variations that create the disorder.

“The idea is that, if you mix all of these patients together, you lose the ability to identify the heritability of the disease,” Zwir explains. “However, if you separate them into coherent subsets that match the symptoms with their genetic variants, you can discover basically what is the risk of the disease. If you put everything in the same bag, you cannot identify the risk. But if you separate them into pieces, you are able to identify the risk in groups and the genetic variance of risk.”

Next, Zwir says, the team is working on a test to diagnose schizophrenia. No such test exists, although a method of identification based on eye movement shows promise. They’re also developing an app to help doctors diagnose schizophrenia and its subtypes. Soon, he says, a group of doctors from Washington University, the University of California–San Francisco, the University of South Florida, and Veterans Affairs hospitals will test the app, hopefully later this year.

Washington University’s study, first published online last September, attracted attention worldwide. Zwir says he received thousands of letters from people with schizophrenia. “It’s the first time that they see … genetics connected to the real symptoms that they have.”

Indeed, the team’s techniques are one of the first attempts to create the oft-heralded future of “personalized medicine” — healthcare tailored to your individual genetic makeup.

For that to occur, simply identifying the name or number of a gene isn’t enough. “You cannot jump from genetics to clinics or to translation of science,” Zwir says. “However, if I tell you that there is a subset of real patients, that have these particular features, with these particular symptoms, and they are associated with these particular genes, not just one gene, but a subset of genes … [that] is the difference here.”

Mapping specific genetic signatures against symptoms, the researchers point out, could also help with other complex disorders, such as heart disease or high blood pressure, which are also known to be heritable to some degree. No single gene to explain them has yet been found. It could be that they, too, appear in concert with specific combinations of genetic and environmental factors.

“Genes don’t operate by themselves,” said C. Robert Cloninger, MD, PhD, one of the study’s senior investigators. “They function in concert much like an orchestra, and to understand how they’re working, you have to know not just who the members of the orchestra are but how they interact.”