It Can't Be All That Bad
Making sense of negative articles about psychology and psychiatry
Posted Sep 19, 2019
We accept the fact that because the human brain is by far the most complex entity in the workd our grasp of how it works and what to do when things go wrong still has many gaps, puzzles, and frustratingly inconsistent findings.
But it is still disconcerting for people who practice or make use of these fields to read in quick succession three headlines that seem to call into question everything we think we know and do. They are:
- “Study finds psychiatric diagnosis to be scientifically meaningless"
- “A Waste of 1,000 Research Papers: Decades of early research on the genetics of depression were built on nonexistent foundations”
- “Study casts doubt on evidence for ‘gold standard’ psychological treatments”
If we take these headlines at face value, psychologists and psychiatrists do not have a valid way of making diagnoses. One way to validate any diagnosis is to show that it has a replicable genetic cause, but at least in the case of depression, we are being told that reported genetic findings aren’t true. Perhaps it doesn’t even matter that we don’t know how to make diagnoses because, according to the third of these articles, we don’t have treatments that work to help people anyway.
We often read appeals to legislators and health insurance company executives that we need to expand access to mental health care. One approach to stemming the alarming increase in U.S. suicide rates is said to be the provision of psychiatric care to more people who need it. Although based on spurious reasoning, the refrain that we need more affordable mental health care emerges every time there is a mass shooting.
These three articles, however, call into question whether expanded mental health care as we know it today even works. It is all enough to make psychologists, psychiatrists, and indeed anyone seeking help from them depressed. Of course, there have always been many voices that demean the value of traditional psychology and psychiatry, often promoting a myriad of alternative methods for dealing with mental illness (assuming that these voices even accept that there is mental illness, of course). But this recent spate of articles challenging the foundations of psychological and psychiatric care come from peer-reviewed journals and cannot be easily dismissed.
Let’s take each one of the three in turn and see what they say.
Is Psychiatric Diagnosis Meaningless?
The study in which the claim that psychiatric diagnosis is meaningless appeared was published in the journal Psychiatry Research by a group of English researchers. They performed a “thematic analysis” of the standard diagnostic system used by psychiatrists and psychologists, known as the DSM-5. They found that there is so much overlap of symptoms among different disorders in DSM-5 that the diagnoses listed there are rendered meaningless. They also state that the DSM diagnoses minimize the role of trauma and adverse events in mental illness and that the individual patient’s actual situation is lost in the array of symptom criteria and diagnostic rules that is the basis for the DSM-5.
The current diagnostic system grew out of the DSM-III, published in 1980, which for the first time included criteria for making each mental health diagnosis. The system was supposed to be theoretically neutral, so that psychoanalysts and psychopharmacologists, then awash in conflict, could both accept it. DSM-III was seen as a major breakthrough in psychiatric diagnosis; before that clinicians could use any criteria they wished to make a diagnosis and so it was impossible to know if one psychologist’s definition of depression was at all similar to another’s.
According to one group of prominent psychiatric researchers, “ Criteria for validating diagnoses have been used for the past 40 years, among them tests of diagnostic boundaries to see if they clearly separate disorders from each other, have a common clinical course, have a common response to treatment, cluster in families, and have common laboratory test findings (which more recently may include temperament/personality traits, neurocircuitry imaging, pathophysiology, and genetic markers)."
To these criteria, DSM-III and its successors added field trials in which large groups of patients were independently interviewed by two clinicians to see how frequently there would be agreement on the diagnosis. The level of agreement is measured using a sophisticated statistical test called the kappa score, with high kappa scores indicating greater degrees of agreement. The results of these field trials showed that most, but not all, of the psychiatric diagnoses tested were “reliable,” and more tinkering was then done with the criteria to try to make them more reliable.
No one argues that DSM-5 is perfect and almost everyone agrees that there is a lot of symptom overlap. Patients with almost every psychiatric illness can have low mood, for example, even though low mood is the hallmark of the diagnosis called major depression. Although there is a diagnosis called panic disorder that is based on patients having repeated panic attacks, panic attacks occur in a wide variety of illnesses. And the hallucinations and delusions that are central to a diagnosis of schizophrenia also occur in some people with bipolar disorder, severe forms of depression, and substance abuse.
Does this mean that psychiatric diagnosis is meaningless? To the extent that a clinician needs to know what they are treating, it is probably better that a diagnosis is based on a reliable set of criteria than on whatever the individual clinician decides to make up. When someone goes to see any doctor for any illness, they, of course, want to be treated as an individual and to have their particular issues and life circumstances taken into careful consideration. But usually, people also want an answer to the question “what do I have?” Psychologists and psychiatrists use an imperfect system based on extensive research to answer that question as best as current science permits. As long as we are appropriately humble about the many shortcomings of that system, it is still a valuable tool.
Do We Have a Clue What Causes Mental Illness?
Of course, if we had objective markers of illness in psychology and psychiatry it would make diagnosis a lot easier. An X-ray makes the suspected diagnosis of pneumonia secure and the use of a blood pressure cuff reveals whether or not a person has hypertension. To be sure, even those diagnoses have substantial heterogeneity. The cough, fever, and chest pain characteristic of pneumonia occur in many other medical conditions; many different organisms cause pneumonia, and no two people with pneumonia are exactly the same. But that chest X-ray does provide a very useful objective, biological marker that assists in making the diagnosis of pneumonia.
For the most part, no such objective markers exist for mental health conditions. There are no X-rays for depression, blood tests for schizophrenia, or gadgets to diagnose panic disorder. All clinicians have to go on are the symptoms that patients report, hence the need to develop symptom lists to make diagnoses as in the DSM system. This lack of objective markers has spurred the neuroscience field into doing extensive research trying to find some biological markers that would help validate psychiatric diagnoses. One such potential marker is an abnormal gene.
We know that many psychiatric illnesses run in families. That could be because the experience of living with a parent or sibling who has depression or bipolar illness influences an individual’s development and is the cause of these familial associations. But it could also be because a genetic mutation is passed from parents with mental illnesses to their children.
One way to find if there are genetic mutations for any illnesses is to select “candidate genes,” genes that do something believed to be relevant to the illness. If we were to look for genes for diabetes, for example, using the candidate gene approach we might select genes known to be involved in the way the body processes glucose to see if there is a mutation in one of them not present in people without diabetes. For an illness like depression, one strategy has been to select genes involved in neurotransmitters that antidepressant medications work on, like serotonin.
Several reports of genetic mutations that seem related to patients with depression have been published, including one involving a gene called slc6a4 that is involved in the transport of serotonin into neurons in the brain. The article in The Atlantic claiming that thousands of research papers on the genetics of depression are a total waste is based on a study published in the American Journal of Psychiatry. A group of scientists reviewed the papers reporting on candidate genes for depression, re-analyzed all the data, and found no support for any of them. “Any evidence that the results might not be reliable was simply not what many people wanted to hear,” writes Ed Yong, a science journalist at The Atlantic. According to him, the positive findings for candidate genes in depression are all “flukes” and the enthusiasm for them sustained by confirmation bias — scientists believing what they wanted to believe.
The failure to confirm previous findings is always discouraging. Also, as detailed in Neuroscience at the Intersection of Mind and Brain, overemphasis of structural genetic causes for mental illnesses can obscure the now abundant evidence that they are all caused, at least in part, by adverse early life experiences and traumatic life events. At the same time, the candidate gene approach to finding genetic underpinnings to disease, used extensively at the beginning of this century, is already outdated. More recent work using more sophisticated technologies still indicates important genetic markers are likely present for illnesses like schizophrenia, bipolar disorder, and autism. We certainly need to refocus attention on traumatic causes of psychiatric illness, but this does not mean we should ignore the probability that genetic research into the causes of these conditions will also eventually bear fruit.
Do Any Treatments Actually Work?
The final bit of disturbing news comes from a paper published in the Journal of Abnormal Psychology in which researchers reanalyzed data that have been used by the American Psychological Association to give specific psychological treatments the designation “empirically supported treatment (EST)." They concluded that the data are not strong enough for a variety of reasons to support most of these treatments as being EST’s. This, of course, doesn’t mean that the therapies in question don’t work. These authors are pointing out, however, that saying they are strongly supported by data is for most of them inaccurate.
Of course, there are some psychotherapies that have a very secure database. These include cognitive behavioral therapy (CBT) for most anxiety disorders, obsessive compulsive disorder, PTSD, and depression. And the new paper says nothing about medication treatments, which have their own issues when it comes to agreeing on the strength of the data. Does it suggest, however, that people should stop seeing their therapists or look to treatments that are not designated as EST’s by the American Psychological Association?
Of course, no one would recommend an abrupt cessation of therapy, especially if the therapy is helping. It is important for patients and potential patients to know, however, that not all psychotherapies are equal in terms of their evidence foundation and that it is probably safest to undertake mainstream approaches like CBT, for which there are firm foundations.
Mental health research pushes ahead and if we took a survey and asked all the people who have had or are now having treatment for a mental illness, many of them would undoubtedly say they found it helpful. There are rigorous studies showing that some psychotherapies and medications are effective for some mental health conditions. A great deal of very impressive research has gone into devising the DSM diagnostic system, trying to find the underpinnings of psychiatric illnesses, and developing effective treatments for them. These three articles need not strike fear and terror into providers or consumers of mental health care. We should still advocate for better access to the evidence-based care we have available.
At the same time, these three articles indicate that we need to have a lot of humility when we approach diagnosing and treating psychiatric illness. A lot of it is based on experience, common sense, and informed hunches. People get better, but we are still in the dark about far too much.