Fallacies of Psychiatry

The Biological Fallacy

Frequently, when I express some skepticism about the diagnostic validity of such constructs as adult attention-deficit/hyperactivity disorder (ADHD) or borderline personality, I receive variations on the following rejoinder: "But they cause changes X, Y, and Z in the brain, as shown in neuroimaging. How can you ignore that? Isn't that proof that they are 'real,' that they are biological diseases?"

I suppose that would be the case if we also accept that I have just caused in you, reader, the "Medscape disease" of reading Nassir Ghaemi. Because I have just caused changes in your brain which, if you were now hooked up to a functional MRI machine with radioligand binding, would show changes in your brain blood flow and maybe even in your dopamine system activity.

Because of our eclectic absence of conceptual clarity in psychiatry, we confuse the term "biological" with "disease." All things biological are not disease, even though we can define disease in such a way that all diseases are biological.

This matter is obvious once pointed out. A few assumptions that seem either patently true or very likely: All human psychological experience is mediated by the brain, and each person only has one brain; therefore, the brain will always be biologically changing as we have psychological experiences. Reading a blog post about the brain is a psychological experience. Having delusions from schizophrenia is a psychological experience. The first brain change does not reflect disease; the second does.

So showing MRI changes with adult ADHD or borderline personality does nothing to demonstrate that those conditions are diseases. If you watch television and play video games inordinately, you will have changes in your brain, and you might also develop clinical symptoms of ADHD. If you are repeatedly sexually abused, you will have changes in the brain, and you might also develop clinical symptoms of borderline personality. But those changes in the brain do not have the same causal role as the neuronal atrophy that happens with trisomy 21, or with schizophrenia, or bipolar illness -- all of which are completely, or almost completely, genetic in origin.

In the case of diseases, biological changes are etiologic; they cause the clinical symptoms. In the case of borderline personality or inattention symptoms in someone without bipolar illness, biological changes are the effect, not the cause, of the other etiologies for the clinical syndrome (eg, sexual abuse).

Biology does not mean disease, because it often reflects pathogenesis, not etiology. All things are mediated by the brain; the brain is the final common pathway of all experience. But changes in the brain are not the ultimate cause of those experiences; they are always the proximate cause, but not the ultimate cause.

This leaves us in the unusual position where so many psychiatrists are more biological than the most extreme cardiologists. We psychiatrists want to emphasize biology in ADHD so that we can feel good about handing out amphetamines (which are neurobiologically harmful); we want to emphasize biology in borderline personality so we can feel good about diagnosing it, and not diagnosing other bona fide diseases, such as bipolar illness, and then refusing to hand out mood stabilizers. For our varied ideologies, we recruit biology as a rationale. Even psychoanalysis -- long the bastion of anti-biological thinking -- has turned to "neuropsychoanalysis" to justify itself.

But because the brain is the final common pathway, all things are biological, including this article. This is a truism; it doesn't prove that something is a disease. For that claim, the further step must be taken of extending the realm of the biological from pathogenesis, which is trivial, to etiology, which is diagnostically meaningful. This has been done in the many definitive genetic studies of schizophrenia and bipolar illness, and it has been disproven in the extensive evidence of psychological and social, but not biological, etiologies for the syndrome of borderline personality-like symptoms (or post-traumatic stress disorder [PTSD], formerly known as "hysteria"). And for adult ADHD, etiologic evidence is simply weak. (Readers should not cite the few genetic studies that make major claims to a highly genetic cause. Those studies did not correct at all for other conditions on which ADHD could be "comorbid," such as bipolar illness; they are thus invalid, as I've reviewed elsewhere.)

So let's put aside biology, unless we are willing to distinguish etiology and pathogenesis.