What do we want from a depression diagnosis?

Originally published in, Kendler, K. S., Parnas, J., Zachar, P. (Eds.). (2020).& Levels of analysis in psychopathology: cross-disciplinary perspectives. Cambridge University Press.

On National Public Radio on January 21, 2014, the singer Jennifer Holliday was interviewed by Robin Young about her career, including her experience with depression. She has also struggled with her weight, and the following exchange ensued: https://www.wbur.org/hereandnow/2014/01/21/jennifer-holliday-returns

Young: Did that issues [sic] around the weight compound the depression or did the depression compound the issues around the weight or . . .

Holliday: I think that the weight came upon as the depression and then added to it. Now, I had clinical depression, so that's actually a medical condition. I now live with and suffer with multiple sclerosis.

Young: Oh, no. Yeah.

Holliday: The side effect of that is also depression. So a lot of things have various factors involved with depression.

Ms. Holliday perceives that there is something important about the designation of her depression as “clinical”, and in this brief response to Professor Sato’s erudite review of the historical, cultural, and scientific contributions to the evolution of the construct of depression (Chapter 35, this volume) I will explore that perception. Why do we care so much about the official, DSM-encoded, clinical definition of depression? There are, of course, many good reasons to care, as Sato’s chapter reminds us. But there are also important domains in which it doesn’t matter very much, and life might be easier for everyone concerned if theoretical and empirical questions about the nature of psychiatric disorder were isolated from the quotidian concerns of clinical practice.

As Sato documents, the long historical discussion about depression as a disorder has recently reached an apotheosis in a narrow but heated disagreement about the so-called grieving exception. The latest version of the DSM allows that under some circumstances the typical, normal, one might say necessary responses to the death of a loved one may qualify as a disorder. In theory, it makes sense. If it means anything to say that grieving is a function of human cognition, then it ought to be possible for that function to fail. Presumably psychotic delusions would not represent the normal function of grief, however understandable they might be in the recently bereaved. Yet at the same time, it would be difficult not to worry about the medicalization of ordinary human grief that is enabled by the exemption’s expiration. History has taught us all too well how these matters proceed, as what begins with exceptional treatment of frank disorder ends with ubiquitous methylphenidate in the morning and zolpidem at bedtime. My goal is to sort out these competing theoretical demands. The answer, I think, is a limited psychiatric libertarianism.

I begin with another example of reaction to depression from the sufferer’s point of view, this one close to home. When my late mother was dying from the lung cancer that had invaded her bones, she was in great physical pain, but she was also anxious (not a sufficient word for what she was feeling), depressed, and unable to sleep. She was, one might say, mourning her own death, and the most reliable comfort she could find was from Xanax® (alprazolam). She took it regularly. I don’t know where it came from; presumably her primary care physician prescribed it. Was she suffering from “clinical” depression or anxiety, from something that had not been exempted from the current version of the DSM? Who cares? My mother certainly didn’t and neither did I. She was suffering, Xanax® relieved it, and that was that.

It seems safe to say that fear of impending death is a normal function of the human mind, and the universality of this fear is the very reason empathy exists. “Some things may never happen:” said Philip Larkin, “this one will”. But by the argument about grieving, if fear of death is a function it can also fail to function properly, and one can imagine reasoned discussion and empirical investigation of the circumstances in which fear of death might be considered a disorder. Are you going to have a heart attack today? Most of the time, however, whether one’s fears about death are functional or dysfunctional doesn’t matter. Dying people suffer, and we grant them the right to choose how they relieve their suffering. They might take Xanax®, they might pray, they might see a grief counselor, or read a book about accepting death, and any or all of these things may or may not work for them. How well they work on average and what their risks might be is a topic for empirical research, and as professionals we might advise patients about their relative costs and benefits, but to even think about diagnosing normal and abnormal behavior under sentence of death is, in any imaginable situation, just too much.

Less of this little analysis than one might think depends on mortality as a cause of human distress. Jennifer Holliday had experienced great success following her role in Dreamgirls, but afterwards her weight shot up to almost 400 pounds. Her record company released her and she couldn’t get a job performing. She was diagnosed with multiple sclerosis. She was—never mind DSM—depressed, and who could blame her? But did she “meet criteria” for “clinical” depression? There are many reasons not to care. Like my mother she was suffering, and like my mother she had a variety of options. She could take medication, see an analyst or a cognitive behavior therapist, read a book, visit a website, whatever. Such options may be more or less effective for people in general, and more or less effective given Ms. Holliday’s individual constitution and needs, but when all is said and done it was up to her. It would be rare for a diagnosis or the absence of one to carry enough information to be of serious assistance in helping Ms. Hudson decide what to do. We in the clinical professions don’t like to say it, but treatment with selective serotonin reuptake inhibitors (SSRIs), as opposed to cognitive behavior therapy or reading a book, doesn’t work better in depression cases that happen to meet DSM criteria. SSRIs affect depression in the same sense that Xanax® affects fear and loathing, albeit probably less successfully in the short run. Drugs have effects, under some circumstances those effects are desirable, and all drugs come with attendant risks. It is no different for psychoanalysis, substituting relationships for drugs.

Libertarian thinking of this kind can untie a lot of difficult knots in psychiatry. I’ll work one more example: alcohol. It is almost universally accepted by now that alcohol abuse and dependence are disorders, and in the terms I have suggested here, of course they are. If alcohol consumption can be normal, it can also be abnormal. But the question of who has this disorder, and what the implications of having it ought to be, is perpetually fraught at both the general level and for individual cases. I enjoy a glass of wine in the evening, but I often do without one and hopefully don’t qualify as disordered. My parents, in the fashion of the Madmen generation of which they were a part, favored a couple of stiff cocktails every single night, and they did not tolerate abstention cheerfully. Did they meet criteria, given their overall high level of functioning and the boozy standards of the time? You know my answer: who cares? People drink. Some people drink more than others. People who do drink derive some benefits and suffer some consequences. People who, in their own opinion or the opinion of others, drink too much and want to drink less can find a variety of methods to help them, as usual ranging from the pharmaceutical to the psychotherapeutic to the cultural to the religious. For most clinical purposes, exceptions to follow, the question of who has a disorder and who doesn’t can be safely ignored.

There are, needless to say, good reasons why questions of disorder matter. The first, which I won’t dwell on with this audience, is that the nature of psychiatric disorder is just an interesting and difficult theoretical problem, practical irrelevance notwithstanding. Beyond that, the most obvious reason to care about diagnosis is money. I don’t know how my mother’s Xanax® was paid for, but if Medicaid was involved, questions of whether the prescription was justified by a clinical diagnosis would have played an important role. A second class of reasons is legal, involving compulsion to treatment or exculpation from criminal behavior. If a person becomes suicidal at the end of life or a family wants to compel a member into alcohol treatment, clinical diagnosis provides at least an officially prescribed route through the nearly impossible tangle of ethical conundrums that are involved. Practical considerations such as these are undeniably important, especially to clinicians who earn a living providing treatment or patients who need to pay for it, but it is ironic that so many of the reasons we care about diagnosis are ultimately mercenary, and so removed from the core mission of comprehending, empathizing with, and alleviating human suffering.

But that is too easy, because there is one other good reason to care about diagnosis, and it is relevant to both our deepest theoretical questions and the real concerns of doctors and patients. It is, in fact, what Jennifer Holliday was referring to. Her response to Young’s question about the role of weight in her depression was actually an objection. After acknowledging that her weight and her depression may have been mutually causal, she changed course, reminding Young that she had clinical depression, which is “actually a medical condition.” What did she mean by this? She meant, I think, that as a medical condition her depression has the same status as her multiple sclerosis, a neurological condition that descended on her, which as such was not entangled in the web of intentional causes and effects that made up her psychological life and her experience of the many personal difficulties she was having at the time. She was almost certainly encouraged to think this way by her clinician, probably in the customary lay language of chemical imbalances in her brain.

What we all—clinicians and patients alike—want from diagnosis is relief from the Sisyphean burden of understanding the relationship between our bodies and our intentions. Why we suffer in the aftermath of great success, why it is so hard to lose weight, why our drinking habits are so hard to change, are questions of such enormous philosophical, clinical, and personal difficulty that they eventually exhaust us, once again doctors and patients together. Melancholia, obesity, and alcohol use are not—sorry—simply genetic or neurological conditions inflicted on us like multiple sclerosis, awaiting medical treatment if anyone could only figure out how. Neither, of course, are they simply lifestyles, matters of immediate and straightforward will. Perhaps nothing is. What, exactly, such conditions are, how in Holliday’s words, “the weight came upon the depression and then added to it,” is a question that has not been answered: not philosophically, not medically, not psychologically, not personally. If it sometimes offers us relief from our collective clinical and personal uncertainty to act as though human struggles are just diseases, and our efforts to overcome them just treatments, so be it. I wouldn’t want to deny anyone the professional security or personal comfort that comes from believing that it is so, but at the same time we must bear in mind that it is also an oversimplification.

Honesty about our clinical and personal oversimplifications compels an explanatory humility that will be well served by a default libertarian philosophy of how we regulate our clinical interactions. Empirical investigation can certainly contribute to our understanding of how people grieve, how we might alleviate the unbearable pain grief can bring, and even to the diagnosis and classification of grieving gone awry. The controversy caused by the expiration of the grieving exception, I would suggest, is not rooted in anti empiricism, but instead in skepticism about the clinical professions’ willingness to abstain from clinical overreach once a disorder has been legitimized. The empirical scientific foundations of our notions of disorder remain incomplete, and even as they are completed their application to individual cases remains extraordinarily difficult. Until there exists hard scientific knowledge to support or undermine the notion that some form of grieving in a particular person at a specific time is pathological, it is best to leave the question out of our discussions with patients, and instead guide them in seeking relief from their problems according to the unique configuration of their own desires.