Mood is a relatively straightforward concept. For the most part, mood varies as to whether it is high, low, or unstable. In contrast, personality is a very complex construct. It describes traits that affect behavior, thought, and emotion. Since personality describes normal variations, as opposed to abnormal states of mind, it is difficult to separate personality disorder (PD), which only some people have, from personality, which everyone has. Another difference is that while depressed or manic mood states can be scaled by clinicians, personality is often measured by self-report systems derived from factor analysis, such as the five-factor model [1], or by an extensive list of traits that can be clinically rated, as in DSM-5 [2]. Finally, mood disorders are often treated with drugs, while personality disorders usually require psychotherapy. For all these reasons, the construct of a mood disorder more readily appeals to clinicians who are looking for targets for treatment, while a personality disorder is seen as a murky and problematic idea.

Diagnostic constructs in psychiatry often reflect currently popular treatment options. Fifty years ago, a wide variety of clinical syndromes, most particularly somatic symptoms, were seen as reflections of abnormal mood or “masked depression” [3]. That diagnosis emerged at the same time as the wide use of tricyclic antidepressants and supported more frequent diagnoses of mood disorder [4]. Physicians naturally favor making diagnoses that lead to a prescription. Even then diagnoses that were indications for psychotherapy, an option that has always been expensive and not readily available, were less popular.

Theoretical ideas about mood disorders have also supported expansion of their scope. Forty years ago, Akiskal and McKinney [5] published a widely cited paper in Science arguing that depression was a single entity that only varies in severity. This construct was influential in the shaping of diagnostic manuals and supported the practice of treating a wider range of patients with antidepressants, a trend further strengthened by the development of selective serotonin reuptake inhibitors. At the same time, psychopathology of all kinds has been seen in the light of variations in mood [6].

While research on depression has been active from the 1950s, systematic empirical studies of personality disorders began to appear only in the 1980s [7]. At the time, mood disorder specialists challenged this research on the grounds that PDs could be better understood as depressive variants. Akiskal et al. [8] dismissed the diagnosis of borderline personality disorder (BPD), suggesting archly that since there was no border on which one could be “borderline,” this term was “an adjective in search of a noun.” Instead, Akiskal recommended that it be treated in much the same way as depression, i.e., with drugs. A counterattack from BPD specialists [9] argued that mood instability is a different phenomenon from sustained low mood. Moreover, evidence failed to show that antidepressants are particularly helpful in BPD [10]. Yet pharmacological treatment for these patients, not to speak of all psychiatric patients, became ubiquitous. To understand this shift in practice, we need to examine changes in the ideology of psychiatry as a medical specialty.

Psychiatry used to be closely identified with psychotherapy. (Even today, the image of a bearded analyst behind a couch continues in New Yorker cartoons.) But beginning in the 1970s, the specialty underwent a paradigm shift [11]. Psychotherapy, in particular psychoanalysis, was seen as unscientific and retrograde. Since then, psychotherapy has been driven to the periphery of the profession. The new paradigm for psychiatry has been based on neuroscience, with treatment redefined as the clinical application of these principles [12]. Psychopathology would now be understood as a problem in neurochemistry or neurocircuitry and treated accordingly, largely with pharmacological interventions.

These conclusions were strongly supported by the pharmaceutical industry and by key opinion leaders drawn from academic psychiatry, who are often supported by the industry [11]. One cannot deny that in choosing interventions for psychiatric patients, money talks. One never sees advertisements in journals supporting psychotherapy. In contrast, each of the latest antidepressants is heavily marketed, even if they differ by only a few atoms from those that have been used for years.

This trend led to the theoretical dominance of neurobiology and a decline in the provision of psychotherapy in psychiatry [13]. It supported diagnoses of mood disorders, which are widely understood to derive from abnormalities of neurotransmission that can be corrected by pharmacotherapy. It undermined interest in personality disorders, seen as poorly defined concepts treated with psychotherapies of doubtful value.

Moreover, patients themselves often prefer to be diagnosed with mood disorders. They may see depression (or bipolarity) as validating—a “chemical imbalance” for which they are not responsible. For some, personality disorder is seen as stigmatizing, implying they have a “bad personality.” It is possible to explain to patients what a personality disorder is and to reassure them that their condition is less chronic than many mood disorders, since research shows that most patients can be expected to get better with time [14]. But while some appreciate this feedback, particularly when antidepressants have not helped, others prefer a diagnosis of mood disorder and request more medication cocktails, showing little interest in talking therapy.

All these factors help to explain why the mood disorder model remains dominant, and some psychiatrists never diagnose a personality disorder. As shown by Zimmerman et al. [15] in a large clinical sample, PDs are highly prevalent but often missed. Of course it is also possible to misdiagnose a mood disorder as a PD, but that is less of an issue in the climate of contemporary psychiatry. Historically, the DSM system tried to encourage clinicians to think about personality by introducing multiaxial diagnosis. But Axis II was a failure, and it only succeeded in marginalizing the concept. In clinical reports, one often sees a statement that Axis II is “deferred,” i.e., to be ignored. In contemporary psychiatry, the roots of psychopathology in personality are downplayed, while many aspects of life are medicalized and understood as epiphenomena of an abnormal mood.

It is often said that PDs cannot be diagnosed in the presence of depression, since abnormal mood distorts personality, and PD features can disappear once mood goes back to normal. While this is sometimes true, when patients are followed over several months, most personality disorder symptoms remain stable even when mood returns to baseline [16]. Yet this idea continues to be taught to students, discouraging them from taking the careful life history required for making a PD diagnosis. It serves as another rationale for ignoring personality disorders, given that patients usually come for treatment when mood is low.

The introduction of lithium for the treatment of bipolar disorder was a heroic chapter in the history of psychiatry. But lithium is a powerful drug that should only be prescribed when definitely required. The introduction of anticonvulsant mood stabilizers, however, made it more possible to consider treating outpatients with milder problems as suffering from variants of bipolar disorder.

The expansion of the bipolar diagnosis has been one of the most influential developments in modern psychiatry [17, 18]. The bipolar spectrum has been extended to patients with a wide range of disorders, including chronic depression, substance abuse, and children with behavior disorders, with the mood instability of BPD seen particularly as lying in a bipolar spectrum [19]. Akiskal [20] continues to see BPD as fictional but now views it as a form of bipolarity rather than depression. Needless to say, Akiskal views psychotherapy as misguided and favors pharmacological treatment for almost all these patients.