Serotonin is just the first page of the antidepressant story
Going beyond the sensationalist headlines to understand what the recent attention-getting paper means (or rather, doesn't mean)
Yes, the serotonin hypothesis doesn’t explain the full story of major depressive disorder, but that is NOT breaking news and does not mean our most frequently prescribed antidepressants are a sham.
The serotonin hypothesis is dead ☠️, and everything we thought we knew about treating depression was wrong. That’s what any reasonable person might conclude from all the press on this topic over the last few weeks. Well, I guess it’s time for us to hang up our white coats 👨🏽⚕️ and find new careers…NOT.
Yes, a recent study published in Molecular Psychiatry raised valid questions about the serotonin hypothesis of depression. However, while these conclusions seemed revolutionary to much of the general public and generated lots of headlines in the press, it may be a surprise that many of us in the field have known about this (and have been operating under this assumption) for quite some time. Note: entire textbooks have been written on the pathophysiology (underlying disease biology) of major depressive disorder and the role of serotonin (knock yourselves out with those); we do not aim to replace or recreate that commentary here.
Instead, we’ll break down the conclusions from the article, explain how clinicians think about these conclusions, and discuss the implications for how we deliver care to patients suffering from depressive disorders moving forward.
A note: while we focus on depressive disorders and antidepressants for the purposes of this Stack (as this is what the paper focused on), the term ‘antidepressant’ is somewhat of a misnomer. Because while the broad category of antidepressant medications 💊 are first-line medications for depressive disorders, these medications are also the most effective medication treatments for a wide range of anxiety and traumatic disorders (and some of the conclusions we will make later hold true for antidepressant mechanisms of action in treating those illnesses too).
Breaking down the study…
The authors of the study linked above utilized a sound, methodical approach (umbrella review) to aggregate results from systematic reviews, meta-analyses, and large analyses focused on examining the link between serotonin (and related neurotransmitters) and depressive disorders (of which Major Depressive Disorder is one type). To rephrase this: they combined previously conducted reviews into one review (to rule them all).
In completing this review, the researchers concluded that there exists no convincing link (either associative or causative) between low serotonin levels and depressive disorders. Notably, the authors point out in the discussion that limitations to their approach included some of the included datasets ranking low in quality based on the AMSTAR-2 (a standardized tool used to rate the rigor of systematic reviews).
Additionally, a good number of the studies included either did not or could not control for participants who had previously or were actively prescribed antidepressant medications, which could clearly impact the underlying analyses of either serotonin receptor activity, serotonin transporter protein activity, or 5-HIAA concentration. The authors’ concluding point that: ‘it is time to acknowledge that the serotonin theory of depression is not empirically substantiated’ was the single most amplified line from the study in pop culture coverage; we’re here to tell you that this is NOT breaking news 🗞 and will dive into this momentarily.
We do have one specific bone 🦴 to pick with the paper before we get there. Specifically, the line:
The idea that depression is the result of a chemical imbalance also influences decisions about whether to take or continue antidepressant medication and may discourage people from discontinuing treatment, potentially leading to lifelong dependence on these drugs.
A patient taking an antidepressant may take the medication for the rest of their life. That decision is shared between the patient and their treating clinician after weighing the risks, benefits, and alternatives (something we call informed consent). However, calling this ‘dependence’ is problematic for two primary reasons:
Dependence in a medical context means something particular: a state in which abruptly stopping medication or use of a substance causes withdrawal symptoms. This is a physiologic state, not a psychological one, and the serotonin hypothesis being true or not has nothing to do with a patient experiencing dependence (which is possible with antidepressants, as it is with MANY medications such as beta-blockers and steroids).
Using dependence colloquially perpetuates the negative connotation and stigma that already hurt patients seeking care for their depression or anxiety. If the authors want to call a patient being treated with sertraline for major depressive disorder dependent on their treatment, we will call all our readers taking a beta-blocker 💊 for their hypertension or insulin 💉 for their type II diabetes dependent on their medication as well 😜.
The history…(read more here)
The serotonin hypothesis (also known as the monoamine hypothesis) of depression was borne out of the development and effectiveness of some of the first antidepressant medications (classes of medications we use less often today, called tricyclics [TCAs] and monoamine oxidase inhibitors [MAOIs]). Until then, there had been no firm consensus on the pathophysiology of depressive disorders. However, these new treatments (TCAs and MAOIs) effectively reduced patients’ depressive symptoms 😃. It was reasonable to surmise that the mechanism of action of these medications (increasing serotonin levels in the brain) was somehow related to the underlying cause of the disease, and thus the serotonin hypothesis was born.
However, as many clinicians and all psychiatrists know, this theory never entirely held water 🛁. If this theory held, antidepressant medications would catalyze an improvement in symptoms within hours, as we know they directly increase intrasynaptic (between neuron) serotonin levels in this timeframe.
Note: the Molecular Psychiatry review included many studies which looked at plasma (blood) serotonin levels, and we know that there is not a strong correlation between plasma and CSF (brain) serotonin concentrations. One more reason to take the study results with a grain of salt.
Based on the thousands of patients we have treated with these medications, we can tell you with certainty that it takes orders of magnitude more time 🕰 (at least 4, and typically 6+ weeks) for patients to experience a therapeutic effect. So straight away, literally from our first week as resident physicians, we had been taught to (and saw with our own eyes a reason to) question the serotonin hypothesis.
Over the last two decades, our understanding of the pathophysiology of depressive disorders and how antidepressant medications treat symptoms have evolved significantly. Again, entire textbooks 📚 and papers written by folks much smarter than us go into excruciating detail summing up our current state of understanding here. To summarize, we now know that while antidepressants do increase intrasynaptic serotonin levels, this is not how they treat depressive disorder symptoms (and therefore, the pop culture narrative of low serotonin -> depressive disorders is just that - a pop culture narrative and nothing more). We now know that the acute effect of increasing serotonin levels in the brain is followed by the longer-term dual effects of:
Reducing the sensitivity of serotonin receptors in the brain, and
Increasing the rate with which serotonergic neurons in the brain fire
This still doesn’t fully explain the delayed onset of antidepressant effect and also doesn’t explain why medications that don’t primarily impact serotonin are also effective in treating depressive disorder, which brings us to the neuroplasticity hypothesis. Increased short-term serotonin concentration and mid-term change in receptor sensitivity and neuron firing rate lead to several beneficial downstream intracellular changes. These changes include:
Changes in gene expression (including for two crucial compounds: cAMP-response element binding protein [CREB] and brain-derived neurotrophic factor [BDNF]
Modulation of glutamate (namely, reducing its release in stressful situations)
So, antidepressants actually treat depressive symptoms through a remarkably complex mechanism of action far beyond their immediate impact of increasing serotonin levels.
What does it all mean…
So, yes, the serotonin hypothesis itself doesn’t tell the whole story of depression and its treatments. But it is the first page of the right story. And just because it isn’t the whole story doesn’t mean the medications we use to treat depression today don’t work. Because contrary to what many might believe to be the order of operations - that we developed the serotonin hypothesis and then created medications based on it - you now know the reverse was true: we discovered a class of medications that worked and then developed a hypothesis of the disease around it. Our predecessors didn’t realize that what they were seeing (increased intrasynaptic serotonin levels) was just the first step in a long cascade of effects that was ultimately responsible for treating the suffering of patients they were seeing.
So let’s not throw the baby 👶🏽 out with the bath water, or in this case, our antidepressants out with the serotonin hypothesis.
On a serious note from us as practicing physicians, please never stop or change your medications, particularly psychiatric medications, without discussing the risks and benefits and creating a plan with your treating clinician.
And don’t be fooled 😮 by the unsophisticated headlines and echo-chamber commentaries that make you believe that the (generally accurate) conclusion of that paper is earth-shattering science. There has also been some commentary around the leanings and affiliations of the authors themselves; we don’t feel the need to wade into that, as we can debunk the headlines based on the science itself.
We’d also like you to remember that psychedelic substances (such as MDMA, LSD, and psilocybin) also primarily modulate serotonin and trigger the same cascade of downstream impacts, which are likely important for their mechanism of action to reduce depressive symptoms. There is a group of folks out there who denounce antidepressants as a big pharma scam while shouting from the rooftops about the potential of psychedelics 🍄. Unfortunately for these folks, that is an incredibly hypocritical stance as antidepressants and psychedelics utilize nearly identical mechanisms of action. Instead, we believe in (and prescribe daily to our patients) antidepressants and look forward to a new class of treatments (psychedelics) becoming safely available for us to prescribe to our patients in the not-so-distant future.
And what comes next…
So, how do we move forward from here? We know these medications work (based on a substantial body of evidence). But how do we find out why they work? Over the last 1-2 decades, we’ve developed exciting and fascinating ways to image the brain, moving from static images (like CT scans and MRIs) to functional images (like fMRIs and PET scans). Through these types of advanced imaging and more sophisticated molecular-level research, we are just beginning to learn more about how brain circuits work - i.e., the interplay between different brain regions (facilitated by MANY neurotransmitters, not just serotonin).
We need more funding and less stigma to get this research done; while slowly closing the gap, research funding for psychiatric illnesses still pales compared to other chronic diseases (such as cancer). That’s why we got into this field 🧠 and decided to start this Substack. We strongly believe that psychiatry is one of the MOST exciting medical fields, with many opportunities for clinicians, researchers, and innovators. What we don’t know far outweighs what we know. The impact that intelligent, passionate, and motivated folks can have in advancing the science and caring for patients is nearly unlimited in magnitude 🚀.
What gives with the dramatic (sad) gap between mental health and oncology research funding? What will it take for mental health funding to rise to the same level as cancer research, and how soon (if ever) can we get there? [yes, we know that was really 3 questions in one]
Why is medical misinformation SO rampant in the lay press, and what can we do as clinicians and healthcare operators/builders to get more accurate information out there?
✌🏽 A + A
To read more about our vision for the Stack, check out our intro post here.
Feedback? Find us on Twitter: Amit & Arpan
Thanks for reading Two Docs and a Stack! Subscribe for free to receive new posts and support our work.
Thanks for writing this, extremely clear and digestable--personally was listening to a popular podcast called the All-In Podcast and they were talking about exactly this issue but without the important clinical intuition demonstrated here!
Not an expert, but my hypothesis re: 1) apart from the obvious one of stigma is lack of great measurability. Cancer has much more defined diagnostic criteria linked to physical measurements that are discrete, as opposed to it feels like items like the GAD-7 are self-reported and not measuring the fundamental physical components that cause anxiety/depression. This also makes it more difficult to extrapolate potential cost-savings without a high degree of variance -> less incentive for grant funding/research. On example in mental health, Alzheimers, stands out which has received a ton of funding (albeit recent Nature article debunking this)-- I believe it's because there was something measurable (e.g. amyloid plaques) that one could point to and try to develop therapeutics for.
To your second question about medical misinformation (+ disinformation) running rampant, I feel like the pandemic has really highlighted the need for a more robust, unified federal public health system.
Health communication is its own art for a reason; building trust with the public by communicating in a straightforward manner, while still giving the tools to follow up on nuance for those who want to go deeper, would go a long way toward combating sensationalized headlines. Axios’s “Smart Brevity” style is a great example of trying to appeal to multiple audiences when reporting on complex issues!