There’s no such thing as an anti-depressant: change my mind

My first point of contention is that the construct of depression as a neuropsychiatric disorder has significant complexity, beyond the typical Diagnostic and Statistical Manual of Mental Disorders (DSM-V) diagnosis of major depression. Since at least the second century – and likely before – the concept of ‘melancholia’ has existed as part of our lexicon. Through various incarnations of the DSM, the modern view of major depression has evolved, with ‘low mood’ sitting at its core. Although I wouldn’t question that basic tenant, there is substantial heterogeneity within the major depressive disorder (MDD) population, and very high rates of what is often called ‘comorbidity’. Despite such heterogeneity, attempts to sub-categorise MDD have proven difficult, and from a clinical perspective, pharmacotherapeutics have also not displayed much in the way of differential effects according to any sub-classification. More of that later…

‘Comorbidity’ remains one of the main challenges with the DSM construct, with around 60% of patients with a diagnosis having ‘comorbidity’. It is difficult to be clear though whether such concomitant diagnoses are a comorbidity: they may precede the onset of the ‘primary diagnosis’ or may indeed be facets of the heterogeneity that exists within the spectrum of a disorder such as MDD. Because we measure these disorders through symptoms, which can vary within hours let alone days and weeks, heterogeneity is complex within, let alone between, patients. Furthermore, a symptom-based diagnostic relying on core symptoms does not account for underlying differences in causation and neurobiology: we have wide and deep evidence demonstrating heterogeneity in both but continue to classify MDD based on core clinical symptoms, despite the realisation that such symptoms have different derivations. RDoC – the Research Domain Criteria Initiative from the US National Institute of Mental Health (NIMH) was established to examine the interface between behavioural dimensions (self-reports, behaviour, and physiology) and neural systems (genes, molecules, cells) at the level of circuits. In doing so, it seeks to delineate circuitry that underlies symptomatology: critically this means deconstructing DSM conditions back to symptoms, rather than longer lists of symptoms that may be present in an individual.

RDoC is an important initiative; one that remains at present an academic, not a clinical, approach. It recognises though that pharmacologic interventions in psychiatry need to relate back to underlying circuit dysfunction, and that targeting of circuits may have benefit in treating symptoms of mental illness, rather than any DSM condition overall. This could pave the way for a ‘precision psychiatry’ approach, where identification of symptoms or behavioural changes could drive use of pharmacology aimed at those specific issues, with the propensity to combine approaches where the symptom pattern is more complex.

So, this brings us back to the question of what it is to be an ‘anti-depressant’. We have plenty of evidence that the most widely used pharmacotherapies in MDD – the SSRIs – influence the overall cluster of MDD symptoms, as defined in outcome measures such as the MADRS or HAM-D. So, does this make them ‘anti-depressants’? As you might expect, the answer to that question is complex. SSRIs have a clinically meaningful effect on overall depressions scores in patients with moderate-to-severe depression. Underneath the overall effect, they appear to have broadly similar levels of effect on the different symptoms that make up the construct (low mood, loss of interest, guilt and delusions, psychological anxiety, somatic anxiety, early morning insomnia, sleep onset insomnia, midnocturnal insomnia, energy/fatigability, hypochondriasis, psychomotor agitation, suicidality, psychomotor slowing, and reduced libido).

This does not address a fundamental issue however: do SSRIs have a core effect on a key symptom such as low mood, with the effects on other symptoms a consequence of improving the low mood? There are some that have posited that this is the case, particularly for sleep disturbances. A large real-world study conducted in the UK found however that the more immediate effects of SSRIs (the first weeks of use) were mainly on anxiety, not low mood, with the latter improving more slowly. So, is our view of the hierarchy of symptoms in MDD really accurate? Further, is it plausible that this class has a ubiquitous effect on all these symptoms directly? That feels far-fetched, and we lack good evidence to back that hypothesis.

My final point on the SSRIs is that they have also (variably) demonstrated effects in anxiety disorders, obsessive compulsive disorder, post-traumatic stress disorder, and others that have not made it into their labels. To what extent can we suggest SSRIs are ‘anti depressants’ when they have such broad effects across at least 4 different groups of disorders (depressive, anxiety, trauma- and stressor-related, obsessive-compulsive, and related disorders)?

The title of my article was deliberately provocative. I wouldn’t argue that the SSRIs don’t influence the symptoms of depression. Rather I would argue that the future of pharmacologic intervention in mental disorders is to target underlying neurobiology more specifically and examine which interventions have benefits for different groupings of symptoms and behaviours, across the boundaries of more classical diagnostic categories. Psychiatrists invest themselves in treating patients, not diagnoses: providing tools that can be tailored better to individual needs is our goal. It is also a strong reminder that the term ‘efficacy’ has many dimensions in psychiatry (we haven’t gone beyond symptoms into functionality and HRQoL: more domains of efficacy that are critical), further reinforcing the need to consider the detail of the effects of any pharmacologic interventions.

For these reasons, I don’t believe in the term ‘anti-depressant’. We have – and are developing – interventions to improve outcomes for patients with MDD but the effects of these interventions are complex, often multi-faceted, and may well have value beyond depression, and equally may be ineffective for patients whose low mood is driven by different biology. Time to look beyond the obvious and examine the next generations of neuropsychiatric drugs in the detail that patients with mental disorders deserve.