Skip to main content
special series

The right treatment, Julie Marriott says, ‘totally changed my life.’Michelle Siu/The Globe and Mail

This is part of a series on improving mental health research, diagnosis and treatment. Join the conversation on Twitter with the hashtag #OpenMinds

Lunacy. Madness. Demonic possession. Black bile. Such archaic notions of mental illness have given way to clinical terms. Now we have schizophrenia, bipolar disorder, social phobia, depression. But as scientific as they sound, each of these disorders, by medical definition, is nothing more than a cluster of symptoms with any number of potential causes.

A diagnosis such as major depressive disorder is about as telling as fever. All kinds of things can cause a fever: bacterial infection, meningitis, flu. Similarly, depression may be triggered by anything from hormonal imbalances to the activation of specific genes, or a history of child abuse. When a patient has a fever, a doctor will prescribe an appropriate treatment after trying to diagnose the cause. In most cases, however, psychiatrists have no surefire way of pinpointing the roots of a patient's despair. Treating mental illness is a shot in the dark.

But what if doctors could order lab tests and scan patients for dozens of known causes of mental illness? What if they could offer a precise diagnosis – such as "chromosome 3p25-26 depression" – using a classification system largely based on the biological signatures of these disorders? Imagine if a doctor could give a patient this advice: "Go directly to brain stimulation treatments – do not try medications, do not go for psychotherapy. They won't work for you."

Psychiatry may be on the verge of such a breakthrough, one that could shake the foundations of the diagnostic system. A growing number of specialists, with a Canadian team at the forefront, are joining forces with researchers who study genetics, the hormonal, metabolic and immune systems, and how the brain works. They're putting aside a century's worth of theories, and delving into the biology of mental disorders on a scale never before seen. The aim is not just to broaden our understanding of mental illness, but to overhaul how we diagnose and treat it.

An overhaul can't come soon enough. One in five Canadians will suffer from mental illness in their lifetime. Many will suffer for years, cycling through one ineffective treatment after another.

Julia Marriott, of Ancaster, Ont., knows how that feels. She had 15 years of psychotherapy and tried more than a dozen different antidepressants, but nothing gave any lasting relief. She chokes up when she talks about hiding her mental illness from her daughter, who was 8 when Ms. Marriott's depression took hold.

Most nights, she says, "I would just go to bed and hope I didn't wake in the morning." In all, trial-and-error treatments consumed two decades of her life, says Ms. Marriott, now 66. "I'm not big on self-pity," she adds. "But it was awful."

Diagnostic models and a focus on symptoms

The ability to predict which treatments will help individual patients is the holy grail of psychiatry, but the quest has been challenged by the field's silo mentality. For more than a century, psychiatry has ping-ponged between biological explanations and theories about the unconscious forces that drive our emotions and behaviours.

As early as the 1860s, some psychiatrists theorized that mental disorders were illnesses of the brain. But brain dissections were too crude to reveal consistent abnormalities linked to mental illness. Theories got far-fetched. In the 1940s, Austrian psychiatrist Wilhelm Reich became famous for his eureka moment that the mentally ill were deficient in "orgone energy." The "cure" involved sitting in a closet-like "orgone energy accumulator."

By comparison, Sigmund Freud's psychodynamic approach was genius. Freud, a neurologist by training, was the first to propose concepts such as repression and denial. He theorized that any mental illness could be treated by resolving unconscious conflicts among the ego (the inner realist), the superego (the moralist) and the id (primal instinct). Decades after his death in 1939, Freud's theories dominated the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM).

Eventually, it was posited that Freud's theories mainly helped the "worried well," says Dr. Jeffrey Lieberman, recent past president of the APA and author of the newly published Shrinks: The Untold Story of Psychiatry. In 1980, psychiatrists in charge of the DSM's third edition rejected all unproven causes of mental illness. Instead, they drew from the latest clinical data to define and classify mental disorders based on symptoms alone – a practice that continues.

Since then, however, psychiatry has not kept pace with advances in other areas of medicine, according to Dr. Thomas Insel, head of the U.S. National Institute of Mental Health. Unlike medical definitions of heart disease, lymphoma or AIDS, psychiatric diagnoses are based on a consensus about symptoms, "not any objective laboratory measure," he wrote in a searing blog post in 2013. "Patients with mental disorders deserve better."

Recent studies have reinforced the idea that the diagnostic system falls short. In a study published in February, researchers at Stanford University School of Medicine found consistent brain changes in thousands of mentally ill patients, whether diagnosed with schizophrenia, bipolar disorder, depression, addiction or anxiety. All showed similar grey-matter losses in brain areas associated with high-level functions such as concentration and decision-making, noted the study, published in JAMA Psychiatry. In a 2013 study, researchers at Massachusetts General Hospital detected shared genetic glitches in the mentally ill across diagnostic categories.

A steady stream of findings like these could leave psychiatry's classification system in shambles. After all, if schizophrenia and bipolar disorder look the same in brain scans and molecular tests, are they, in fact, distinct illnesses? Could they be different manifestations of the same genetic condition, or subtypes of an as-yet-unnamed brain disorder? To find answers, psychiatrists need to look at the bewildering science of mental illness in new ways.

Dusting for depression's fingerprints

Canada, it turns out, is leading the way, through a multiyear study called the Canadian Biomarker Integration Network in Depression (CAN-BIND). It brings together clinical psychiatrists, neuropsychiatrists, molecular scientists, neuroimaging specialists and experts in bio-informatics, who use computer algorithms to analyze complex data such as genetic code.

Part of the mission is to identify as-yet-unnamed subtypes of depression. But the ultimate goal is to shorten the path from diagnosis to the right treatment. "This is not just a study," says Dr. Sagar Parikh, a University of Toronto psychiatrist who is working on CAN-BIND. "This is a program to transform depression treatment."

CAN-BIND is following a model used in breast-cancer research. In the mid-1980s, researchers divided cancer patients into groups: those who got better with treatment and those who didn't. Scientists analyzed thousands of biological traits to find markers that set patients apart, using computers to crunch the data.

In patients who got sicker, researchers found high levels of HER2, a protein that stimulates tumour growth. The finding led to new drugs to block the action of this protein. Since then, life expectancy for patients with early-stage HER2-positive breast cancer has increased 30 per cent.

In much the same way, CAN-BIND is dividing patients with depression into two groups – responders and non-responders to a selected treatment. Depending on the study phase, patients receive antidepressants, or psychotherapy, or repetitive transcranial magnetic stimulation (a non-invasive treatment that uses magnetic pulses to activate specific parts of the brain). Researchers are combing through patients' biological and psychological makeup, acting on the hunch that different types of depression may respond to different treatments – and leave distinct fingerprints.

The CAN-BIND model is like a game of Clue, Dr. Parikh says. The "murderers," "weapons" and "crime scenes" in Clue – three variables involved in solving the mystery – correspond to the study's three research areas.

The first area involves a psychiatric evaluation that takes into account factors such as substance abuse, early childhood trauma and recent life stress; any of these may affect biological systems such as brain function. The next area uses brain imaging to find abnormalities. The third covers blood tests, which may detect proteins produced by specific genes, disruptions in metabolic or hormonal function, or signs of inflammation. (Some researchers believe that inflammation due to an overactive immune system may trigger mental illness.)

Results from the battery of tests are fed into software sophisticated enough to find patterns among thousands of patient variables. The idea is to uncover clues that can be used to predict whether a specific treatment will work for future patients. Hypothetically, Dr. Parikh says, "the best predictor of a treatment working might [prove to] be a combination of a sleep disturbance, together with an underactive part of the brain, combined with one protein that is off."

Similar studies are under way in the United States, but CAN-BIND is the first to pull together this many variables in a collaborative effort of nearly a dozen universities and research centres. The same model can be adapted to study other mental illnesses, researchers say.

The "big data" approach is a radical departure from the usual hypothesis-driven studies, which typically focus on a single research question. Dr. Parikh acknowledges that CAN-BIND is a "fishing expedition."

Dr. Lieberman, the former APA president, cautions against pinning too many hopes on studies like CAN-BIND. As with any fishing expedition, he points out, "you could end up not having caught anything."

One woman's victory

Despite great leaps in neuroscience and genetics, psychiatrists still don't know why one-third of patients with depression – or half a million Canadians each year – don't get better with standard treatments. Ms. Marriott fought depression with everything she had. After years of psychotherapy and antidepressants, she tried light therapy, vigorous exercise, mindfulness courses, fish oil – "anything that might work." But she could not escape the crushing feeling that everything was "black, negative and pointless" – except during episodes of mild mania. Occasionally, she would get the sudden urge to redecorate: "I would give away a perfectly good couch and then buy another one."

Ms. Marriott's official diagnosis is "major depressive disorder with a hypo-mania component." She grew up watching her mother, who had bipolar disorder, spend most days in bed. One wonders whether their shared genes had something to do with Ms. Marriott's unsuccessful treatments. So far, there are no diagnostic tests to answer questions like this. Eventually, however, Ms. Marriott did find an effective treatment. In 2012, she became a patient in a study of repetitive transcranial magnetic stimulation; each treatment lasts about three minutes and feels "just like a woodpecker is pounding on your upper forehead."

Since her last round of brain stimulation in December, 2013, Ms. Marriott has been depression-free. She says she feels like her "pre-age-40 self" – interested in seeing friends and eager to travel to places like Mexico and Botswana. Once more, she is capable of feeling "excited, happy, touched and sad – all those normal emotions." She emphasizes the sense of security she feels in knowing that, if she starts to relapse, she can go for another round of therapy. Getting the right treatment, she says, "has totally changed my life."

Biology on the fritz or something more?

Early findings from the CAN-BIND study will be released later this year. In the meantime, preliminary results from a multicentre U.S. study suggest that brain imaging has the potential to predict whether a depressed patient will respond to a specific treatment. Patients underwent positron emission tomography (PET) scans, which use a radioactive sugar to create images of brain activity. Researchers found that depressed patients who responded to psychotherapy had sluggish activity in the insula, a brain region involved in emotion and self-awareness, unlike those who did well on antidepressants.

Brain imaging would be an expensive treatment-selection tool. But if new studies make a strong case that brain scans lead to more successful treatment, they may not be out of reach for average patients down the road, says Dr. Jeff Daskalakis, chief of the mood and anxiety department at the Centre for Addiction and Mental Health in Toronto.

"It costs a lot of money to miss a diagnosis," notes Dr. Daskalakis, who is working on the CAN-BIND study. In Canada, the cost of mental-health services combined with lost productivity and income due to untreated mental disorders is estimated at nearly $30-billion a year.

Still, researchers emphasize it could be years, if not decades, before brain imaging or blood tests become reliable, let alone practical, tools. And that's assuming their studies net big fish.

For now, we are left with the same big questions that have baffled physicians and philosophers for centuries: Is mental illness simply a matter of biology on the fritz – a physiological problem that can be solved as soon as scientists crack the code? Or is the anguish of each patient also a unique expression of the sense of isolation and dread that may strike any of us at our core?

In mental illness, unlike other diseases, life events are refracted through our subjective perception in ways that can damage our mental and physical well-being. In his book, Dr. Lieberman uses himself as Exhibit A. After surviving a home invasion at gunpoint in his early 20s, his youthful mind chalked it up as "a thrilling adventure." Years later, he suffered from post-traumatic stress disorder, after an air conditioner slipped out of his grasp and fell to the street below. For months, he was tormented by the thought that he could have caused someone's death. He lost his appetite, had trouble sleeping, and played the incident "over and over in my mind like a video loop." But he was the same person who had escaped from the home invasion without psychological scars. He explains, "You can have something that is purely experiential and yet it produces enduring symptoms."

Even if scientists come up with blood tests to screen for mental illness, the lived experience of a mental disorder will remain highly personal. For these reasons, mental disorders, in turn, will remain "existential diseases" that require compassionate care as well as effective medical treatments, says Dr. Lieberman.

The new approach to studying mental illness may be compatible with this philosophy. The strength of a project like CAN-BIND, says Dr. Parikh, is that it integrates many specialties and ways of looking at the problem. "That's the real beauty of it." Researchers are no longer determined to prove that a single treatment will help every patient. Instead, he says, the question has become: "What is the best fit?"

Adriana Barton is a reporter in Vancouver for Globe Life, specializing in health and science.

The Centre for Addiction and Mental Health has purchased advertisements to accompany this series. The organization had no involvement in the creation or production of this, or any other, story in the series.

Interact with The Globe