Searching for a Common Pathway
One of the challenges in determining what causes mental illness lies in defining what constitutes a mental illness in the first place. Dictionaries and reference books differ in their exact wording, but a good all-purpose version might be this: A mental illness involves changes or abnormalities in emotions, cognition, motivation, and/or behaviors resulting in distress or problems functioning in life. Context matters, though. One of the tricky things about defining mental illness is that many—even most—of the symptoms are considered “normal” in at least some circumstances.
We all have emotions, for example, both pleasant and unpleasant ones. We might feel anxious when we face challenging or threatening situations. We might feel depressed when we experience a significant loss, such as the death of a loved one. Even something like paranoia can have an appropriate time and place. Have you ever watched a scary movie—one that truly terrified you? If so, you were likely a bit paranoid afterward. Some people look in their closets before they go to bed after watching such a movie. Or they hear sounds outside and feel terrified, imagining it’s a scenario from the movie. All of this is normal. However, at some point, intense unpleasant feelings and states should diminish, allowing you to go on with your life as before. Therefore, it’s important that any definition of mental illness somehow account for context, duration, and appropriateness.
For an example of what I mean, consider “shyness.” Are people allowed to be shy? Is that normal? Most would say yes. So, at what point does shyness turn into an anxiety disorder such as social phobia? Drawing these lines is a matter of some debate in the field. One of the most notable controversies concerns depression—specifically whether, in some situations, these symptoms are “normal” and not an illness.
The Diagnostic and Statistical Manual of Mental Disorders, or DSM, is the “bible” of psychiatry. It defines all the different diagnoses, their diagnostic criteria, and provides some relevant information and statistics. The current version, updated in 2022, is known as “DSM-5-TR.” In DSM-IV, the diagnostic criteria for depression included a caveat called the bereavement exception.1 It suggested that if someone had symptoms of depression in the context of the loss of a loved one, clinicians should hold off on diagnosing depression. A professional might certainly offer support in the form of talk therapy, but prescribing medications wasn’t necessarily appropriate. The exception had limits—among them, the depression should not last more than two months and should not produce suicidal thoughts or psychotic symptoms. In DSM-5, however, the exception was removed altogether. This served to encourage clinicians to diagnose depression even in the context of stressful life events like the loss of a loved one. Many clinicians and researchers felt that the American Psychiatric Association (which produces the DSM) had gone too far in pathologizing experiences like grief. On the other hand, supporters of the exemption’s removal cited research showing that antidepressants can decrease symptoms of depression even in the context of grief. These advocates felt that not diagnosing the problem and offering medication treatment might be unnecessarily cruel.2
Despite such controversies, there are many situations that seem clear-cut. When someone has crippling hallucinations and delusions, or suffers from overwhelming fear and anxiety every time they leave their home, or can’t get out of bed for weeks at a time due to severe depression, most of us would agree that this constitutes a mental illness. The “unusual” or “inappropriate” nature or degree of their symptoms, the intensity of their distress, and their inability to function all suggest a serious problem that merits the diagnosis of a disorder.
The premise of the DSM, both the current and prior versions, is that there are distinct mental disorders with clear criteria that can be used to separate them from one another. In some cases, these distinctions are obvious. Schizophrenia is very different from an anxiety disorder. Dementia is different from ADHD. These distinctions are supposed to help guide treatment, predict what will happen to people with specific diagnoses (their prognosis), serve as a tool for clinicians and researchers to communicate more effectively with each other, and so on.
The diagnoses in the DSM are given tremendous importance. They are required for clinical care and reimbursement by insurance companies. They are almost always needed to get research funding given that most mental illness studies focus on just one disorder at a time. And they are critically important to the development and dissemination of treatments as well, since to get FDA approval for a drug, pharmaceutical companies must conduct large clinical trials of specific medications for specific disorders and show a benefit. Even interventions such as psychotherapy are normally studied in clinical trials designed around one specific diagnosis. So, in many ways, the mental health field revolves entirely around these diagnostic labels.
The field, however, has been plagued by debates over how to diagnose different mental disorders, especially since (as we discussed in the previous chapter) there are no objective tests to definitively diagnose any mental disorder. Instead, we use checklists of symptoms and criteria. We ask patients and family members what they are feeling, witnessing, and experiencing; we investigate, cross-reference, and explore; and then we make a diagnosis based on the best match or matches.
In some situations, these diagnostic labels are extremely useful. Remember John who developed major depression? His diagnosis helped inform his treatment, and the treatment worked. John got better—all the way better. After a year of doing well, he was able to stop treatment and remain well. The diagnostic criteria allowed John’s psychiatrist to recognize the disorder, understand different treatment options, choose ones that were likely to work, and then discontinue the treatments after a defined period of time. Unfortunately, it’s not so simple—or successful—for others.
Sorting Out the Similarities
One of the challenges in the mental health field is that no two people with mental disorders are completely alike, even when they get diagnosed with the same disorder. There are two primary reasons for this—heterogeneity and comorbidity.
Heterogeneity refers to the fact that people diagnosed with the same disorder can have different symptoms, severity of symptoms, levels of impact on their ability to function, and courses of illness. Not one of the diagnoses requires that all criteria be met. Instead, it’s a minimum number—for example, a major depression diagnosis requires at least five of the nine criteria. This makes for a lot of variability. One person with major depression can have depressed mood, excessive sleep, problems concentrating, low energy, and be eating much more than usual, resulting in weight gain. Another person with this diagnosis might be unable to sleep more than three hours, have lost their appetite and shed twenty pounds, and along with depressed mood and low energy be thinking about suicide. These patients have very different symptoms that require differing approaches to treatment. One is thinking about harming himself while the other isn’t. One can’t sleep, so might benefit from a sleeping pill, while the other one is sleeping too much. Despite these striking differences, both might benefit from an antidepressant or psychotherapy.
Dr. Alan Schatzberg is a prominent depression researcher and a professor of psychiatry and behavioral sciences at Stanford University who has called for rethinking the diagnostic criteria for major depression.3 Those in this field are frustrated with the lack of understanding of this common illness and continued poor treatment outcomes—as I mentioned earlier, the likelihood of a full and complete remission of major depression symptoms with the first antidepressant a patient tries is only about 30 to 40 percent. Schatzberg notes that some symptoms that commonly occur in people diagnosed with major depression are not included in the core diagnostic criteria. For example, anxiety is a common symptom in many people with depression, but it’s not among the nine in the DSM. The same goes for irritability, which is experienced by about 40 to 50 percent of people with depression.4 Pain is common, too, with physical pain present in about 50 percent of people with major depression compared to only about 15 percent of the general population.5 Are our treatment outcomes so poor because we’re missing or failing to include the treatment of other diagnostic symptoms?
It’s not just depression that causes so much confusion and debate. There’s tremendous heterogeneity in all the psychiatric diagnoses. Sometimes, the differences are stark and dramatic. Some people diagnosed with OCD are still able to work and function normally in life, while others are completely disabled by their symptoms. People diagnosed with autism-spectrum disorder can be wildly different from each other. There are high-functioning billionaire businesspeople with this diagnosis, while others live in group homes unable to care for themselves. So, are these singular diagnoses really the same disorders? Or are they all simply on a spectrum, with some people having severe forms of the illness while others have a mild form? Unfortunately, the complexities don’t end there.
Comorbidity is the other big factor that accounts for differences in people with the same diagnosis. About half of the people diagnosed with any mental disorder have more than one.6 We talked about comorbidity a bit in the last chapter: Remember my discussion of runny nose disorder and sore throat disorder? While some people had one or the other, many had both. A similar example in the mental health field is that of depression and anxiety. The majority of people diagnosed with major depression also have anxiety, and most people diagnosed with anxiety disorders also have major depression. For example, in a survey of over nine thousand US households, 68 percent of people with major depression also met criteria for an anxiety disorder at some point in their lives, and several studies have found that one-half to two-thirds of adults with anxiety disorders also meet criteria for major depression.7 Antidepressants are commonly used to treat both depression and anxiety disorders, while anti-anxiety medications are commonly used to treat people with both anxiety disorders and depression. So, where the diagnoses often overlap and the treatments are sometimes identical, are they really different disorders? Is it possible that they are simply different symptoms of the same disorder? Could anxiety and depression—like runny noses and sore throats—share a common pathway?
Finally, diagnoses can change over time. Symptoms can come and go and morph into very different mental disorders, further complicating treatment and diagnosis, and bedeviling the quest to investigate the nature and cause of these disorders.
Let’s look at an example.
Mike is a forty-three-year-old man with a chronic, disabling mental disorder. But which one? When he was a child, he was diagnosed with ADHD and started taking stimulant medications. They helped somewhat, but school remained difficult. He was often bullied and teased. He reported a lot of anxiety around these social stressors and received psychotherapy for social anxiety disorder. Some clinicians raised the possibility of Asperger’s syndrome, then a diagnosis of its own on the autism spectrum, but they didn’t officially make this diagnosis. By adolescence, he’d developed symptoms of major depression—not surprising given both his academic and social stressors. He was started on an antidepressant, which helped a little. Within a few months, however, Mike began to develop symptoms of mania and was quickly diagnosed with bipolar disorder. He had hallucinations and delusions, and he was given medication targeting both his psychotic and mood symptoms. He was hospitalized several times. Over the next year, when his psychotic symptoms persisted and failed to respond to treatment, his diagnosis was changed to schizoaffective disorder. Also during this time, Mike began to develop obsessions and compulsions, and he was diagnosed with OCD as well. Over the following several years, on top of his continuing psychiatric symptoms, he began to smoke cigarettes and use recreational drugs. Eventually, he became chronically addicted to opioids.
So, what is Mike’s diagnosis? According to DSM-5, he currently can be diagnosed with schizoaffective disorder, opioid use disorder, nicotine use disorder, OCD, and social anxiety disorder. But in the past, he also had ADHD, major depression, bipolar disorder, and possibly even Asperger’s syndrome. You might make a case that the major depression diagnosis was a mistake—many people with bipolar disorder are diagnosed with depression before they have a first manic episode that clarifies the diagnostic picture. The same might be argued about the change of diagnosis from bipolar to schizoaffective disorder. But even if you remove one or two of these, you are left with a lengthy list of what are all different disorders—supposedly with different causes and certainly different treatments. Yet Mike has just one brain. Are we to believe that he’s an extraordinarily unlucky individual who developed some half dozen separate and distinct disorders?
While Mike’s story is extreme, having more than one diagnosis is common, as are changes in symptoms and diagnoses. Having problems with addiction is also common in people with mental disorders. Stories like Mike’s raise serious questions about the validity of our diagnostic labels. If the diagnoses listed in DSM-5 are really separate and distinct disorders, why do so many people have more than one of them? Why do they change over the course of a lifetime? Do some psychiatric disorders lead to others? If so, which ones come first, and what exactly happens to make them cause other disorders? Alternatively, are some just different symptoms or phases of the same underlying problem? Is it like runny nose disorder and sore throat disorder—two seemingly different disorders that respond to different treatments but share the common pathway of inflammation? Is there a common pathway for mental disorders, even ones that appear to be profoundly different from each other?
Taking a Deeper Look
Researchers have been trying for decades to figure out what makes individual disorders different from each other at a biological level. Interestingly, they don’t yet have any clear answers. In fact, as I’m about to share with you, the research to date suggests that different disorders might not actually be all that different from each other, even though symptoms can vary widely.
Let’s look at three of the psychotic disorders—schizophrenia, schizoaffective disorder, and bipolar disorder.
The primary feature of a schizophrenia diagnosis is chronic psychotic symptoms, like hallucinations or paranoia. A diagnosis of bipolar disorder is given to people who primarily have mood symptoms—manic and depressive episodes. However, people with bipolar disorder also commonly have psychotic symptoms when they get manic, and even sometimes when they get depressed, but these psychotic symptoms go away after the mood symptoms improve. Schizoaffective disorder is a diagnosis that includes features of both schizophrenia and bipolar disorder, including chronic psychotic symptoms and prominent mood symptoms. Most people consider these disorders unequivocally “real.” Many in the field hold these disorders apart from things like depression and anxiety, sometimes calling these the “biological” disorders. So, what do we know about them? What makes them different from each other?
A lot of money has been spent researching this question. The NIMH funded a multisite study called the Bipolar Schizophrenia Network on Intermediate Phenotypes (B-SNIP). This study included more than 2,400 people with schizophrenia, schizoaffective disorder, or bipolar disorder; their first-degree relatives; and people without these disorders (normal controls). The researchers examined key biological and behavioral measures, examining brain scans, genetic testing, EEGs, blood parameters, inflammation levels, and performance on a variety of cognitive tests. They found that people with the disorders were different from the normal controls, but they couldn’t tell any of the diagnostic groups apart from each other. In other words, there were abnormalities in the brains and bodies of people with these disorders, but no significant differences at all between those with bipolar disorder, those with schizoaffective disorder, or those with schizophrenia. If they are truly different disorders, how can that be?
On the one hand, when we consider more information, maybe these findings aren’t so surprising after all. First, although the diagnosis of schizophrenia isn’t supposed to include prominent mood symptoms, the reality is that one of the common features of schizophrenia is a group of symptoms called negative symptoms. These include blunting of facial expressions, severely reduced speech and thought, losing interest in life (apathy), getting no pleasure from life or activities (anhedonia), reduced drive to interact with others, loss of motivation, and inattention to hygiene. You might notice significant overlap with the symptoms of depression. Interestingly, DSM-5 specifically cautions clinicians against making the diagnosis of major depression in people with schizophrenia, even though many of these negative symptoms are the same symptoms found in depression. Instead, clinicians are encouraged to diagnose a schizophrenia spectrum disorder. The implication is that even though the symptoms might overlap, we shouldn’t call them the same thing. Why not? Is there science to support that recommendation? In reality, the DSM-5 acknowledges in its introductory remarks that we don’t know what causes any of the psychiatric diagnoses. Therefore, if people are having the same symptoms, how can we say they are not caused by the same process?
The treatments for these disorders overlap as well—more than you might think. Mood stabilizers, such as lithium, Depakote, and Lamictal, are commonly used in bipolar disorder and have approval from the FDA for such use. However, about 34 percent of people diagnosed with schizophrenia are also prescribed these same mood stabilizers, even though by definition, those with this diagnosis aren’t supposed to have significant mood symptoms.8 Antidepressants are also commonly used in both bipolar disorder and schizophrenia. Studies show that almost all bipolar patients receive an antidepressant at some point in their illness for their depressive episodes, and about 40 percent of patients diagnosed with schizophrenia do, too.9
And then there are the antipsychotic medications. These are used for schizophrenia, bipolar disorder, and schizoaffective disorder, and are prescribed to treat all the symptoms of these disorders, not just the psychotic ones. The FDA has even approved many of these medications both as “antipsychotics” and as “mood stabilizers” for the treatment of bipolar disorder.
On the other hand, while all of this suggests quite a bit of overlap between bipolar disorder, schizoaffective disorder, and schizophrenia, it is also true that the symptoms of bipolar disorder and schizophrenia can be dramatically different. Many people with bipolar disorder never have psychotic symptoms. Many are never hospitalized, and many function quite well in life. Meanwhile, almost all people with schizophrenia will experience severe impairment in functioning, with the majority qualifying as disabled.10 That’s not to say that there aren’t high-functioning schizophrenics, or that bipolar disorder can’t be disabling. In fact, one study that followed 146 people with bipolar disorder for almost thirteen years found that the people were symptomatically ill about 47 percent of the time despite treatment.11 It’s difficult to keep a job when you are ill almost half the time. But there are definite differences in the usual presentation of these diagnoses. Could it be that people with schizophrenia have a more severe form of the same illness, or one less responsive to our current treatments, while people with bipolar disorder may have a milder illness and/or symptoms that respond better to our treatments, resulting in episodes of recovery?
Dr. Bruce Cuthbert, the acting director of the NIMH at the time of the B-SNIP study, suggested, “Just as fever or infection can have many different causes, multiple psychosis-causing disease processes—operating via different biological pathways—can lead to similar symptoms, confounding the search for better care.”12 However, the study failed to find any hallmark biological markers to distinguish the diagnoses. What Cuthbert didn’t mention is that we know that fever is itself a symptom, with one clearly defined biological pathway—inflammation that triggers the hypothalamus to increase body temperature. However, there are many things that can trigger the inflammation, such as infections or allergic reactions. Diverse infections can have the same symptoms, through common pathways, even when the infectious agents (bacterial or viral) are different.
It seems very plausible that the symptoms of bipolar disorder, schizoaffective disorder, and schizophrenia all share a common pathway as well.
Sorting Out the Overlaps
I have now suggested that bipolar disorder, schizophrenia, and schizoaffective disorder are possibly the same illness, but on a spectrum of symptoms and with different responses to existing treatments. Earlier in the chapter I suggested that major depression and anxiety disorders might be similarly related and share a common pathway. For many people in the field, neither of these assertions is difficult to grasp or believe. Mental health professionals have struggled with these distinctions for decades and know all too well about the overlap in these disorders and their treatments.
However, the overlap doesn’t stop with these conditions.
Symptoms overlap between all kinds of mental diagnoses, not just those you would expect to be related. As I’ve mentioned, many different disorders, both mental and medical, can lead to psychotic symptoms. In fact, about 10 percent of patients diagnosed with major depression will have psychotic symptoms.13 Anxiety symptoms are also common in multiple diagnoses. The overall prevalence of anxiety disorders in the general population is quite high to begin with—in any given year, about 19 percent will experience an anxiety disorder. When looking at lifetime prevalence, that number rises to 33 percent, meaning that one out of three people will meet the criteria for an anxiety disorder at some point in their life.14 The rates in people with depression, bipolar disorder, schizophrenia, and schizoaffective disorder are much higher—about double. Sometimes, we just rationalize these symptoms away: “Wouldn’t you be anxious if you had schizophrenia?” As appealing and intuitive as this sounds, it’s not so simple. There is a strong bidirectional relationship between schizophrenia and anxiety disorders. In other words, people who first manifest an anxiety disorder have anywhere from an eight-to thirteen-fold increased risk of developing schizophrenia or schizoaffective disorder.15 Those are not trivial increases. But why should this be?
In 2005, Dr. Ronald Kessler and colleagues reported the results of the US National Comorbidity Survey Replication, a household survey that included a diagnostic interview of more than nine thousand representative people across the United States.16 Overall, 26 percent of people surveyed met criteria for a mental disorder in the last twelve months—that’s one in four Americans! Of those disorders, 22 percent were serious, 37 percent were moderate, and 40 percent were mild. Anxiety disorders were most common, followed by mood disorders, then impulse control disorders, which include diagnoses like ADHD. Of note, 55 percent of people had only one diagnosis, 22 percent had two diagnoses, and the rest had three or more psychiatric diagnoses. That means almost half the people met criteria for more than one disorder.
Diagnostic overlap is easier to dismiss when we are talking about anxiety disorders, perhaps because anxiety is a mental state we all experience. So let’s look at autism spectrum disorder. Most people don’t think of autism as a purely “mental” illness, but more as a developmental or neurological disorder that starts early in life. Yet 70 percent of people with autism have at least one other mental disorder and almost 50 percent have two or more.17 It’s also interesting to note that embedded in the criteria of autism spectrum disorder are many of the symptoms of obsessive-compulsive disorder (OCD).
And what happens to people with autism over longer terms? Are they at higher risk for developing additional mental disorders? Again, the answer is often yes. A prominent feature of autism is impairment in social skills, so it stands to reason that a diagnosis of social anxiety disorder could follow if interactions caused anxiety. In such a case, many would assume the autism spectrum disorder came first, and that the social anxiety was an understandable consequence of the autism. However, it’s now well documented that autism itself puts people at higher risk for developing every other type of mental disorder.18 This includes mood disorders, psychotic disorders, behavioral disorders, eating disorders, and substance use disorders. How can this be? Is it just that autism is stressful? We know that stress can put people at risk for all sorts of mental disorders, and having autism is undoubtedly stressful. But as you will learn, the explanation is much more complex than that.
This phenomenon isn’t limited to anxiety disorders or autism spectrum disorder, either. Looking at eating disorders, bulimia nervosa occurs in about 1 percent of the population, anorexia nervosa in about 0.6 percent, and binge eating disorder (the newest disorder in the category) in about 3 percent.19 Many people consider these societal disorders rather than biological brain disorders. Yet overall, 56 percent of people with anorexia, 79 percent of people with binge eating disorder, and 95 percent of people with bulimia have at least one other mental disorder.20 So here we go again—which one comes first? Do eating disorders cause other mental disorders, or do other disorders cause eating disorders? Both: There is a bidirectional relationship between eating disorders and other mental disorders. Which other disorders, you ask? All of them. The same is true of addiction. Again, it’s a bidirectional relationship. People with any substance use disorder are at higher risk of developing a mental disorder, and people with mental disorders are at much higher risk of using and abusing addictive substances. Why is that?
I could go on this way, diagnosis by diagnosis, but I won’t—an important 2019 study clarifies the bigger picture. In this study, researchers used a Danish health registry to analyze psychiatric diagnoses in almost six million people over seventeen years.21 What they found was that having any mental disorder dramatically increased the chances of that person later developing another mental disorder. There were strong bidirectional relationships for everything! Even disorders that most people think are completely unrelated—schizophrenia and eating disorders, intellectual disability and schizophrenia. Mix and match them however you like. The odds ratios across the board in this study were generally between two and thirty. This means that if you were diagnosed with any mental disorder, you were two to thirty times more likely to later be diagnosed with another mental disorder. Which one? Any of them! While some of the very high odds ratios are due to overlap in symptoms between different disorders, the point is that the odds ratios were elevated for all disorders in all directions.
What’s more, this bidirectional relationship also applied to mental disorders and so-called “organic” mental disorders. “Organic mental disorder” is the term used to refer to symptoms of a mental disorder that are thought to be caused by a medical condition or a medication. We discussed this briefly earlier: For example, if someone with cancer loses their appetite and gets depressed, they often are not diagnosed with major depression. The assumption is that these symptoms are due to the cancer, and not a true “mental” disorder. Yet the evidence of this study now shows that if people develop “mental” symptoms attributed to a medical problem, they are then much more likely to develop a mental disorder in the future—and vice versa. This finding certainly raises the question of whether separating “organic” mental disorders from the rest really makes sense.
All told, this study raised several important questions. Bidirectional relationships, especially ones that are particularly strong in both directions, suggest a common pathway does exist. While the symptoms may differ, perhaps our diagnoses are far more similar than we’ve long thought.
The Danish study wasn’t the first piece of scholarship to suggest all mental disorders might share one common pathway. In 2012, Dr. Benjamin Lahey and colleagues studied the symptoms and prognosis of eleven different mental disorders in thirty thousand people.22 They looked at “internalizing” versus “externalizing” disorders. Internalizing disorders are ones in which distress is thought to be directed inward, such as depression and anxiety disorders. Externalizing ones are when distress is outwardly directed, such as substance use disorders or antisocial behavior. They found tremendous overlap in these different disorders and raised the possibility of a “general factor” leading to all of them.
In 2018, doctors Avshalom Caspi and Terrie Moffitt took this research further by including all mental disorders in a review article, “All for One and One for All: Mental Disorders in One Dimension.”23 They reviewed a tremendous amount of research, including epidemiological studies, brain imaging studies, and studies of known risk factors for mental disorders, such as genetics and childhood trauma. The data was exhaustive, encompassing research on people of different ages, including children, adolescents, and adults, and from many different parts of the world. After examining all this data, they found strong correlations among all mental disorders. When they looked at risk factors for mental disorders, they found that not one risk factor conferred risk only for a specific disorder—instead, each and every risk factor conferred risk for many. For example, one study they examined looked at the genetics of psychiatric disorders.24 The study evaluated more than three million siblings, hoping to identify which genes conferred risk for depression, anxiety, ADHD, alcoholism, drug abuse, schizophrenia, and schizoaffective disorder. Given that these are all different disorders, one would expect them to have different associated genes. However, the researchers found that most of the genetic variations conferred risk for a broad range of disorders. There were no genes that were specific to only one disorder. Even childhood abuse confers risk for most mental disorders, including PTSD, depression, anxiety, substance use disorders, eating disorders, bipolar disorder, and schizophrenia.
Given the never-ending overlap in correlations among all mental disorders and all of their risk factors, Caspi and Moffitt used a complex mathematical model to analyze these correlations in the hope of making sense of them. This model offered a shocking conclusion. It suggested that there appears to be one common pathway to all mental illnesses. Caspi and Moffitt called it the p-factor, in which the p stands for general psychopathology. They argued that this factor appears to predict a person’s liability to develop a mental disorder, to have more than one disorder, to have a chronic disorder, and it can even predict the severity of symptoms. This p-factor is common to hundreds of different psychiatric symptoms and every psychiatric diagnosis. Subsequent research using different sets of people and different methods confirmed the existence of this p-factor.25 However, this research was not designed to tell us what the p-factor is. It only suggests that it exists—that there is an unidentified variable that plays a role in all mental disorders.
Our job is to figure out what it could be.