The decades-long attempt to locate the gene or genes for schizophrenia has failed, according to a new article in Psychiatric Research by prominent schizophrenia researcher E. Fuller Torrey.
In the article, Torrey reviews the history of the Human Genome Project, their hopes for identifying the genetic basis for schizophrenia, and how those hopes have been dashed by the complete failure to find anything of the sort.
Torrey writes, “Over twenty years later not a single gene has been identified to cause schizophrenia, despite the expenditure of almost $8 billion in genetic research by NIMH. Nor have any new treatments become available from this research.”
This paper is surprising since Torrey has long argued that schizophrenia is a brain disease to be treated biomedically.
Who Is E. Fuller Torrey?
Torrey is a psychiatrist and a researcher on schizophrenia and bipolar disorder. In research circles, he’s known as the founder and executive director of the controversial Stanley Medical Research Institute, which has spent more than $550 million on biological research on schizophrenia and bipolar disorder over the past few decades.
Torrey is one of the primary researchers studying the possibility of immunological causes of schizophrenia. In particular, his research focuses on the parasite toxoplasma gondii, commonly carried by cats and which easily passes to humans. Some studies have found that exposure to that parasite, and exposure to cats in childhood more generally, is associated with a heightened risk of schizophrenia. However, other studies have not found a statistically significant correlation, so it is still considered theoretical. Even if a connection exists, it likely explains only a fraction of schizophrenia cases.
To most, though, Torrey is more well-known as the founder of the Treatment Advocacy Center (TAC), which lobbies politicians to support forced treatment for people with schizophrenia. For this reason, Torrey is a controversial figure: Psychiatric survivor group Mind Freedom International labeled him “one of the most feverishly pro-force psychiatrists in the world.”
Torrey and the TAC have been at the forefront of politicians’ plans for forced treatment, including shepherding Kendra’s law in New York as well as working with Mayor Adams on a plan to sweep up the unhoused population into psychiatric institutions.
However, studies since the 1990s, including a Cochrane review in 2017, have found forced treatment to be ineffective, concluding that it doesn’t reduce rehospitalization or criminality, or improve social functioning, quality of life, or even treatment adherence. Worse, research has found that involuntary treatment increases the risk of suicide. In addition, laws around involuntary treatment have been criticized for racial bias.
The United Nations has called for member states to ban forced treatment, saying that it “may well amount to torture” and that it violates human rights. And the World Health Organization has called for a transformation of mental health services to focus on person-centered and rights-based approaches.
The Human Genome Project
Despite advocating for the idea that psychosis has biological causes, and despite his presence on research teams involved in heritability studies, Torrey has long taken a somewhat skeptical view of genetics research. Yet laypeople, and many mental health professionals, still believe that schizophrenia is a genetic disorder. Indeed, the NIH’s MedlinePlus website lists schizophrenia as a “genetic condition,” even while admitting that its causes are “not well understood” and calling it “an active area of research.”
In the 1990s, the Human Genome Project—a massive effort to understand every human gene—promised to solve this riddle by uncovering the genetic basis of the disorder.
A few rare diseases are directly caused by a single gene variant (like Huntington’s disease, cystic fibrosis, and sickle cell anemia, for example). More commonly, though, people with certain identifiable genetic variants are at much higher risk for certain diseases. For instance, about 60% of women with the BRCA1 or BRCA2 variant develop breast cancer, compared to 13% of women without those variants. Variants that increase risk in this way are known as “risk genes.”
These strong correlations with specific heritable genetic variants are helpful for delineating the biological pathway of these diseases. For instance, the BRCA gene is considered a tumor-suppressing gene, so it makes sense that differences in that gene can strongly affect cancer risk.
If specific genetic variants could be identified that increase the risk of schizophrenia in a similar way, researchers could learn more about the possible biological pathways of schizophrenia and potentially develop drugs that target that biology. It might also make it possible to develop preventive measures specifically for those at high risk.
This was the dream of the Human Genome Project and its founders. As Torrey tells it, for many of the major figures involved in the creation and direction of the Human Genome Project, identifying the genetic basis of schizophrenia was to be one of the main drivers of their work. Torrey specifically notes that Charles Delisi, James Watson, and Senator Pete Domenici all had relationships that led to a personal obsession with schizophrenia.
It was Delisi who, in 1985, first floated the idea of the Human Genome Project with his colleagues and spent the next few years shepherding the project forward. His wife, Lynn, was one of the researchers focusing on how schizophrenia seemed to run in families.
Watson, already famous as a co-discoverer of the structure of DNA, was the first director of the Human Genome Project. For Watson, the focus on schizophrenia came when his son experienced a crisis at age 15, including a suicide attempt and a runaway scare, along with being diagnosed with schizophrenia and hospitalized for six months.
And according to Torrey, Domenici was known for putting the kibosh on expensive projects in his position on the budget and appropriations committees. But his daughter had been diagnosed with schizophrenia, and he was eager to move forward any research that might help with that specific diagnosis. It was with his help that the project was ultimately funded.
Thus, from the beginning, the Human Genome Project was designed by true believers with the promise of discovering the genetic basis of psychiatric disorders, especially schizophrenia. They spent the 1990s working to sequence the more than 20,000 human genes at an ultimate cost of $2.7 billion. In 2000, with about 90% of the genome mapped, their efforts were heralded as a success in a famous White House press conference (the full sequence was finally finished in 2022).
The Promise—and Failure—of Genetics
After that press conference in 2000, prominent scientists and politicians promised that our understanding of the genetics of mental illness was about to explode, leading to amazing new treatments that would totally change the landscape of psychiatry.
Torrey quotes Francis Collins, director of the Human Genome Project, who stated that by 2020, “gene-based designer drugs” would be available to treat and prevent everything from diabetes to Alzheimer’s disease, “and the diagnosis and treatment of mental illness will be transformed.”
Torrey also quotes Steven Hyman, who was director of the NIMH during the Human Genome Project era: “The search for the genes associated with most mental illnesses will now move forward at a greatly accelerated pace….Highly selective, safe, and effective new therapies will become available for the treatment of the common mental illnesses in the not-too distant future.”
But the promised gains never materialized. The Human Genome Project did help in our understanding of other conditions, especially cancer, as well as the single-gene disorders. But for psychiatric disorders, no genetic test was found; no biological test was found; no psychiatric drug was developed based on a genetic discovery. Not for depression, anxiety, psychosis, bipolar disorder, OCD, ADHD, or any other “mental illness.”
Current research has found that genetic testing could explain less than 1% of whether someone would receive a diagnosis of schizophrenia. Including the entire genome, another study found an explanatory power of 2.28%. Still another found that genetic causes might explain up to 3.4% of whether someone received the diagnosis.
By comparison, a study that included all known genetic factors as well as “environmental factors”—life experiences, social circumstances, family history, pain—found that genetics predicted 0.5% of the risk for schizophrenia, while environmental factors explained about 17%.
That finding was consistent with previous research finding strong correlations between schizophrenia and various life experiences, but no effect of genetics on risk for schizophrenia.
And other researchers have concluded that genetic results provided no useful data, even when considering genes theoretically associated with psychiatric disorders.
This failure of genetic research has paralleled the failure of the rest of psychiatry. Despite a massive increase in the amount of people receiving a diagnosis and taking psychiatric drugs, outcomes have only worsened over the years.
Researchers write that psychiatric drugs do not improve the course of mental illness or prevent hospitalization or suicide; they write that there is no evidence that psychiatry has improved outcomes over time; and they have found that antidepressant use, antipsychotic use, and even just receiving a psychiatric diagnosis all lead to worse outcomes than for those who don’t get diagnosed or treated—even after controlling for symptom severity.
Breakthroughs Proven Wrong, Again and Again
Torrey begins his discussion by noting that the failure to find any relevant genes for schizophrenia is ironic considering the hyperbolic promises of researchers throughout the years, including constant announcements in the media that researchers have discovered the genetic key to understanding it.
“Schizophrenia alone has probably been the subject of more genetic breakthrough announcements over the last thirty years than any other human disease,” Torrey writes.
But, he adds, the problem is that none of those supposed “breakthroughs” was subsequently replicated. The pattern, according to Torrey, is that one study would find a “breakthrough” genetic correlation, but the next study would prove it wrong. But that study would find a completely different correlation and get heralded as another “breakthrough”—until the next one proved it wrong and found a new “breakthrough,” over and over for decades.
Facetiously, Torrey notes, “This cycle of hypomanic enthusiasm followed by depression after a failure to replicate is similar to the clinical course of bipolar disorder itself.”
Why Does Schizophrenia Run in Families?
One key question for researchers was how to explain the finding that schizophrenia appears to run in families. Early studies of twins and siblings raised separately seemed to support the notion that schizophrenia was heritable. This led many researchers to the genetic hypothesis. But, according to Torrey, more modern research has shown that these studies are methodologically flawed. The fact is, schizophrenia may run in families because child abuse, trauma, and other environmental factors also run in families.
Studies have attempted to disentangle this fact by claiming to include siblings—sometimes twins—“reared apart,” but these studies actually include mostly siblings and twins who were raised together for at least some time, raised by a close family member, or who reconnected before volunteering for the study—meaning that they don’t actually account for family environment at all.
Moreover, according to Torrey, there are biological factors besides genetics that also affect these results—such as the presence of environmental contaminants in the neighborhood (some siblings, even if supposedly raised apart, were raised in homes in the same neighborhood or spent plenty of time in their home of origin). Another factor: fetal environment. Infections—such as Torrey’s favorite parasite, toxoplasma gondii—may be passed down during pregnancy. Thus, even if the siblings were raised separately, they may have shared the same immunological factors before birth.
To bring this point home, Torrey notes that a large twin study in 2018 found that only 15% of the participants shared a diagnosis of schizophrenia with their identical (monozygotic) twin (12 out of 81 pairs in which at least one had the diagnosis), despite sharing 100% of the same genetic material. For fraternal (dizygotic) twins, it was 3%, despite sharing 50% of the same genetic material.
Torrey writes that this concordance rate is similar to that for infectious diseases—with polio having a 36% concordance rate and tuberculosis a 31% concordance rate in twins.
The hope of the Human Genome Project researchers, though, was that genetics studies could get around these methodological problems and directly identify the genes responsible.
The Failure of a Wide Variety of Genetic Studies
In the search for possible genetic roots of a disorder that runs in families, perhaps the most helpful type of study is a linkage analysis. Torrey writes that there have been “at least” 32 linkage analysis studies of schizophrenia and 40 of bipolar disorder. None of them were able to replicate the findings of the others.
Torrey writes, “The coup de grace for schizophrenia linkage analysis studies was administered in 2002 by [Lynn] DeLisi, who had carried out a linkage analysis on 309 families in which at least two siblings had been diagnosed with schizophrenia or a related disorder, including the Galvin family in which 7 of the 12 children had been diagnosed with schizophrenia. DeLisi was unable to replicate the findings of previous linkage studies and her own findings for linkage were weak. She concluded that linkage analysis was not an effective technique for identifying the genetic roots of schizophrenia.”
Another approach was to try to predict candidate genes: guessing which genes might be responsible for a given diagnosis, and then checking to see if they were more common in those with the disorder than without. Since schizophrenia was thought to involve the dopamine system, candidate gene studies often looked at genes that were responsible for regulating or developing that system in the brain.
Over a thousand candidate gene studies were conducted on schizophrenia, according to Torrey, with hundreds more for bipolar disorder. But this approach failed just as completely as the linkage analysis approach.
As Martilias Farrell and colleagues wrote in a 2015 paper: “The current empirical evidence strongly supports the idea that the historical candidate gene literature yielded no robust and replicable insights into the etiology of schizophrenia.”
And in 2017, Emma C. Johnson and colleagues wrote, “Taken as a group, schizophrenia candidate genes are no more associated with schizophrenia than random sets of control genes.”
(Torrey also notes that candidate gene studies of depression completely failed in the same way.)
Genome-wide association studies (GWAS) were viewed as the next breakthrough. This type of study casts a wide net at more than a million mutations called single-nucleotide polymorphisms (SNPs). These SNPs are like markers that flag nearby genes as potentially increasing risk for a disease.
Torrey writes, “It was hoped that [GWAS] would lead to the discovery of a few genes of large effect which are involved in the cause of schizophrenia. Unfortunately that is not what researchers found. Rather than finding a few genes of large effect which could be linked causally to schizophrenia, they found hundreds of genes of very small effect which have not been linked causally to this disorder.”
Even if the hundreds of genes identified are truly related to schizophrenia, even in aggregate they explain such a microscopic amount of the risk that there is no clinical utility in listing them.
Worse, the problem of causality is paramount here. As one example of how these studies can confound correlation and causation, Torrey writes that people with schizophrenia are far more likely to be tobacco smokers, so some genes found in a GWAS to be associated with schizophrenia might actually just be genes linked to smoking.
Differences in ancestry can also produce spurious results. For instance, due to racial biases in diagnostic categories, far more Black people are diagnosed with schizophrenia. Thus, GWAS studies might inadvertently identify genes related to African ancestry as increasing the risk for schizophrenia.
Finally, Torrey adds that the genes identified by GWAS as increasing risk for schizophrenia overlap among all psychiatric diagnoses, including depression, ADHD, and autism. Thus, these genes are not specific, but rather seem associated with emotional distress in general and altered cognition of many kinds.
The closest thing a GWAS has found to a consistent genetic link to schizophrenia is the Major Histocompatibility Complex (MHC), a correlation that was replicated in several studies. This area of chromosome 6 has been studied since the 1970s as a potential factor in schizophrenia. But this is certainly not a specific finding: the MHC has been linked to nearly every autoimmune disorder and a wide variety of infectious diseases. To some extent, this supports Torrey’s belief that schizophrenia can be caused by toxoplasma gondii. But, like the other GWAS findings, this correlation explains very little of the risk and is too broad to be of any clinical utility.
Torrey notes that Lynn Delisi and colleagues studied the entire genome of nine families that had multiple members with the schizophrenia diagnosis.
He writes, “Surely, it was thought, such heavily burdened families would lead to the identification of causal genes of large effect. Alas, each of the 9 families had different genetic findings, none of which were widely shared among the other families or among other individuals with schizophrenia. Nor could any of the genetic findings be definitively linked to the cause of the disease and none of the variants identified overlapped with the SNP loci identified in the GWAS study of schizophrenia.”
To sum up the failure of GWAS, Torrey quotes researchers McClellan and King, who published a thorough overview of the topic in 2010:
“The general failure to confirm common risk variants is not due to a failure to carry out GWAS properly. The problem is underlying biology, not the operationalization of study design. The common disease–common variant model has been the primary focus of human genomics over the last decade. Numerous international collaborative efforts representing hundreds of important human diseases and traits have been carried out with large well-characterized cohorts of cases and controls. If common alleles influenced common diseases, many would have been found by now. The issue is not how to develop still larger studies, or how to parse the data still further, but rather whether the common disease–common variant hypothesis has now been tested and found not to apply to most complex human diseases.”
After the failure of GWAS, researchers turned to the next type of genetic study: identifying copy number variants (CNVs). Some part of the human genome consists of repeated genetic patterns; copy number variation is a term for the difference in the number of times certain parts of the pattern repeat. Essentially, CNVs are either extra copies (duplication) or missing copies (deletion) of genetic code. CNVs are common and usually they don’t cause any problems. However, certain CNVs have been identified as causing genetic disorders. For instance, Huntington’s disease is directly caused by a large number of repeats of the CAG trinucleotide of the HTT gene, and the more times the code repeats, the earlier the onset of Huntington’s.
The CNV that causes DiGeorge syndrome is associated with an increased risk for schizophrenia, and it is estimated that about 1% of people with schizophrenia have that CNV, compared with about 0.025% of the general population. But again, this does little to explain a cause for the other 99% of people with schizophrenia—or why most people with DiGeorge syndrome don’t develop schizophrenia.
A newer type of genetics study focuses solely on “rare coding variants” that are detectable using exome sequencing. The problem, according to Torrey, is that the variants detected using exome sequencing are indeed extremely rare. Thus, even if there is a slight correlation in which people with these variants are more likely to develop schizophrenia, it has very little clinical utility. The vast majority of those with the diagnosis do not have these variants. Of course, this also means that these variants are not a very good indicator for understanding the theoretical underlying biology.
According to Torrey, “it is unclear what role, if any, the rare variants play in causing schizophrenia.”
Indeed, researchers who completed an exome study for schizophrenia wrote that “The main conclusion of this investigation is a negative one. The diagnostic yield for exome sequencing of known neuropsychiatric genes in this sample is about 1%.”
Searching for the Psychiatric Yeti
To sum it all up, Torrey quotes a number of eminent researchers on the failure to identify a genetic basis for schizophrenia:
- Gershon, 2011: “Where is the missing heritability? […] Among scientists in the field, there is a sense of disappointment in the air.”
- Crow, 2011: “There comes a point at which the genetic skeptic can be pardoned for the suggestion that if the genes are so small and so multiple, what they are hardly matters, the dividing line between polygenes and no genes is of little practical consequence. Have we reached this point?”
- Uher and Rutter, 2012: “It can be summarized that molecular genetic studies of psychiatric disorders have done a lot to find very little. In fact, in the era of genome-wide association studies, psychiatric disorders have distinguished themselves from most types of physical illness by the absence of strong genetic associations.”
- Latham, 2011: “The most likely explanation for why genes for common diseases have not been found is that, with few exceptions, they do not exist.”
Perhaps the most telling quote comes from Steven Hyman, director of the NIMH during the Human Genome Project era:
“I made one enormous error. I thought that family and genetic studies, advances in neuroscience, and the newly emerging discipline of molecular biology would soon elucidate pathogenesis and result in improved therapeutics. How wrong I turned out to be.”
Ultimately, Torrey writes that the search for schizophrenia genes can be compared to a wild goose chase—although, he adds, “it is even worse than that since a wild goose chase has, at least theoretically, a wild goose that might be caught. If indeed genes causing schizophrenia do not exist, then the thirty year search has been more like a search for a psychiatric yeti.”
Torrey’s final argument against schizophrenia as a heritable disease is an evolutionary one. According to a study in 2010, people with psychiatric diagnoses are far less likely to have children than people without a diagnosis. Those with schizophrenia are at the lowest extreme. For every 100 men without a diagnosis who have a child, only 10 men with schizophrenia have a child; for every 100 women without a diagnosis who have a child, only 18 women with schizophrenia have a child.
Moreover, Torrey notes, hundreds of thousands of people with schizophrenia were forced to undergo sterilization as part of the eugenics movement in the first half of the 1900s, including 18,000 people in the United States. In Nazi Germany, in addition to sterilizing about 132,000 people with schizophrenia, about an equal number were murdered.
Yet, during that same time (1850-1950), the number of new incidences of schizophrenia only increased, including by sevenfold in the United States.
Torrey writes, “Schizophrenia as we know it clinically has been well described for over 200 years during which time the reproduction rate of those affected has been exceedingly low. If schizophrenia was truly a genetic disease it should have died out by now. Thus, the fact that it still exists is strong evidence that it is not a genetic disease.”
Despite the investment of billions of dollars into research on the supposed genetics of schizophrenia over the course of decades, nothing useful has been found. Continuing down this road, Torrey argues, is a waste of time and money.
He writes, “NIMH invested extensive resources in this research with little to show for it and at the expense of alternative research projects. Since schizophrenia does not appear to be a genetic disorder, NIMH’s research portfolio should be reviewed.”
Torrey suggests that the NIMH should focus on basic research into the biological causes of schizophrenia, including his pet theory, that schizophrenia is caused by immunological problems (lead candidate: toxoplasma gondii). He also suggests that the NIMH should fund more clinical trials in an effort to develop new drugs to treat the disorder.
However, the NIMH’s tight focus on funding genetic research has also prevented the exploration of the known psychological causes of schizophrenia, such as the impact of trauma, isolation, and poverty. It has also prevented the proliferation of non-biological understandings of psychosis, such as the Hearing Voices movement, and non-medical treatments, like Open Dialogue and Soteria.
MIA Reports are made possible by donations from MIA readers like you. To make a donation, visit: https://www.madinamerica.com/donate/