The Serotonin Signal

For decades now, common wisdom has held that the more serotonin floating around in your brain, the less likely you are to experience depression and the sunnier your mood and outlook on life. Doesn’t that explain the value of selective serotonin reuptake inhibitor (SSRI) antidepressants like Prozac, Zoloft and Paxil, which boost serotonin in the brain?

The reasoning seems to make sense. Serotonin is a neurotransmitter—a chemical that carries messages between brain cells—involved in many functions throughout the brain and body. For instance, serotonin plays a role in heart and gut function and helps regulate sleep and wakefulness. Given the efficacy of SSRIs for some with major depressive disorder (MDD), it would seem sensible that low serotonin levels in the brain could create a chemical imbalance and lead to depression.

Last summer, that belief appeared to be upended by a research paper from scientists at University College London. Published in the respected scientific journal Molecular Psychiatry, the report concluded that low serotonin is not a cause of depression, creating more controversy than any neuroscience paper in recent memory.

As the story surged, public outrage grew. Readers began to ask: If low serotonin doesn’t cause depression, then why are so many of us treating it with SSRIs? How could doctors have been so wrong?

As a neuroscientist, I found it frustrating to watch this unfold. Although it seemed shocking and groundbreaking, this paper actually did not tell us anything new about depression. The researchers had not run any experiments or uncovered something novel about the condition. Rather, they had simply summarized a body of scientific research that had already existed on depression for decades.

Let’s back up a bit. In the 1950s, a compound called iproniazid was developed to treat tuberculosis. But when patients received the medication, they displayed an unexpected side effect: improved mood. Iproniazid falls into a drug class known as monoamine oxidase inhibitors (MAOIs), compounds that increase the levels of the neurotransmitters serotonin, dopamine, and norepinephrine. Many years later, it was shown that SSRIs, which only increase serotonin, could produce the same positive effects on mood, and with fewer side effects. This discovery, coupled with some early evidence suggesting decreased serotonin levels in people who died by suicide, led to the approval of SSRIs for treating depression. Simultaneously it spawned a hypothesis: Perhaps low serotonin was the cause of depression.

However, in the years that followed, neuroscientists struggled to consistently find clear evidence that serotonin levels are lower in the brains and bodies of people with depression. There were some hints here and there, but the results were generally inconsistent and showed mixed signals. With time, more and more evidence accumulated suggesting that low serotonin alone cannot explain the whole story of depression. But meanwhile, clinical trials continued to show that SSRIs were quite effective at treating depression.

So, what exactly is going on here? How can SSRIs be helping if low serotonin isn’t the problem? The (perhaps unsatisfying) answer is that in the brain, things are hardly ever as straightforward as this. The brain is a complex orchestration of many systems working in synchrony, all bumping into and influencing one another through intricate connections. Serotonin levels don’t necessarily need to be low for SSRIs to be helpful in treating depression; increasing serotonin could alter the function of some other, downstream system, which then leads to improved mood. It is thus entirely possible that SSRIs exert their antidepressant effects in the brain through a mechanism totally different from serotonin.

Here’s another way to think of it: If you spend six hours squinting at a computer screen at work, you might have a headache when you get home. You go to the medicine cabinet and take some Tylenol to get rid of it. In this case, the root issue is that your facial muscles have been overworked and spasming, but the Tylenol acts on a different system in your body to simply blunt the pain. It does not relax your muscles, but it does stop your experience of pain, and that’s all you really wanted. SSRIs could be acting in a similar way, addressing the issue indirectly.

Fast-forward to the summer of 2022. A group of University College London researchers looked over the decades’ worth of scientific studies I’ve described, finding that depression is not clearly linked to low serotonin. They published their findings in a paper with a strong claim: that the serotonin hypothesis has some serious holes, at best.

In scientific publishing there are a few different types of articles. Most are primary articles, in which scientists present the results of their experiments. A second type is a review article, whose authors summarize a group of primary articles to draw all the results together and make meaning of them. The serotonin paper that sent the media reeling was a third type, a systematic umbrella review. The researchers intended to summarize everything out there, both primary articles and review articles that had already covered related topics.

Here's a summary of some of the umbrella review's major findings:

• Serotonin breaks down into a compound known as 5-HIAA, or 5-hydroxyindoleacetic acid. Measurements of 5-HIAA taken from blood plasma and cerebrospinal fluid show that 5-HIAA levels are not significantly lower in those with depression than in those without.
• Serotonin receptors (specifically 5-HT1A) appear to be expressed at similar levels in the brains of those with and without depression (although the authors note that the quality of the published data here is somewhat poor).
• In the brain, neurons interact with each other at tiny little structures called synapses. At these synapses, one cell releases a neurotransmitter onto another cell, where it binds to specific receptors. In synapses where serotonin is released, a protein called SERT (for serotonin transporter protein) carries serotonin out of the synapse and into cells, thus playing a key role in determining how much serotonin the receiving cells get. If SERT is overactive, too much serotonin is removed from the synapse, resulting in abnormally low serotonin signaling. The review found that SERT levels were lower in those with depression—surprising, since one would expect that lower SERT levels would result in increased serotonin signaling in the brain.
• Genetic studies do not show clear evidence that gene mutations affecting serotonin signaling in the brain are associated with depression.

This umbrella review was far from perfect. In fact, in an upcoming critique, a group of 35 scientists from 14 institutions are sending a clear message: “A leaky umbrella has little value.” The researchers point out methodological flaws in how the umbrella review was conducted, noting that several studies which do show evidence for low serotonin in depression were inexplicably left out; other measurements cited, the critics say, are poor indicators of serotonin level in the brain, and thus of limited value.

From an objective perspective, the evidence for low serotonin in MDD is mixed, with some articles reporting supportive evidence and others reporting no such findings. The fact that science has struggled to pin down consistent, reproducible signals of low serotonin in MDD certainly suggests that it’s not a robust indicator for depression. It is very possible that with more advanced tools and techniques, neuroscientists will more thoroughly unravel this mystery and find, for example, that some MDD patients show lower levels of serotonin while others do not.

Alas, low serotonin is not the overwhelming cause of depression for many who suffer from it! Breaking news to some, but not particularly surprising to those who were already familiar with the literature. In fact, I prepared a video for my social media channels on the neuroscience of depression in March 2022 and found the same thing: “Looking at the brain, there’s surprisingly not super-obvious evidence that serotonin is the problem.”

Despite all this, the story came as a huge shock to those outside the field and stirred public outrage. The reason? An unfortunate disconnect between science and the public. Without question, this is a complex neuroscience topic, and something that only researchers who study depression would know much about. But why didn’t scientists update the public that this whole “chemical imbalance” thing involving serotonin wasn't as solid as many believed? And why didn’t the London scientists, in particular, strive to create public context for their work?

Two issues stand out. First, there are very few channels through which scientists can message the public. We publish our papers in paywalled journals full of jargon, creating unnecessary distance from the public. I therefore believe in the importance of using public media channels (like social media) to create a dialogue. Further, scientists are usually very careful in discussing topics like this for fear of misleading the public. Until the evidence is fully there and scientists reach a consensus, we don’t want to make any strong claims in our public messaging. In the case of serotonin and depression, the evidence had been building, but no sort of consensus had been demonstrated until the article in July. (And to be fair, this may not be the end of the road for the serotonin hypothesis.)

Of course, media outlets do not exercise this same caution or feel this same hesitation. The primary objective of modern media outlets is to catch your eye and your click, and often the most startling story achieves this best. When that is your goal, it’s no wonder that scientific accuracy gets discarded.

For example, when Fox News’s Tucker Carlson reported on this study, he stated, “Last week we learned that actually SSRIs don’t cure a chemical imbalance in your brain.” Wait a minute—what? This paper did not include any information about SSRIs; it only discussed the underlying neurobiology of depression. Regardless, Carlson continued, “So, first we were told that SSRIs would save lives. Now we learn they don’t actually work as intended. . . . And yet, here is the best part, people are ignoring this news and the drugs are still being prescribed.”

While the logic seems correct, the argument is skewed. The paper provided no evidence showing that SSRIs are not effective for depression. Carlson simply extrapolated that, incorrectly. He also failed to mention that SSRIs could be acting on systems downstream of serotonin in the brain.

In fact, this is what some scientists believe: The brain has hundreds of billions of synapses, and they often change with experience—a phenomenon known as synaptic plasticity or neuroplasticity. There is a growing belief that SSRIs act by increasing neuroplasticity rather than by simply increasing levels of serotonin. It has been proposed that over time and with repeated use, SSRIs allow synapses to become more dynamic, enabling a sort of flexibility that lets one reassess certain thoughts or perceptions that may be affecting mood.

For instance, imagine you believe that you’ve been underperforming at work lately. You feel ineffective and worry that everyone is judging you; perhaps your boss is even considering firing you! These beliefs have become hard-set in your mind, dragging down your mood. By increasing neuroplasticity, SSRIs may allow these hard-set beliefs to loosen up and take new shape as the physical structures in your brain change and become more adaptive. You may start to see things from a new perspective: “Actually, I’m performing better than last year, and people aren’t judging me, they’re just interacting less because I’ve been socially withdrawn.” This gradual change in neuroplasticity could also be the reason why SSRIs take a few weeks before the effects kick in. The neuroplasticity takes some time to get going before the reshaping of negative percepts can occur. Finally, this could be why SSRIs pair nicely with talk therapy: The added flexibility enables more productive therapy sessions and greater openness to reshaping viewpoints. Note that this is currently only a theory, not something scientists are certain about yet.

As the story of the apparent serotonin controversy spread in July 2022, it hurt to watch science get misrepresented, so I posted a video on my social media channels. But my concerns linger. There is no way to tell how many people abruptly stopped taking SSRIs because of this and may have been hurt, especially because SSRIs can have severe withdrawal effects when halted abruptly. It’s impossible to predict what scientific topic will be misrepresented next and the harmful impact this may have.

So, what is the solution? Perhaps scientists need to be more thoughtful about what they publish and consider the impact it may have on society. Or maybe media outlets need to be more accurate in their reporting. But is it really the responsibility of either side alone? In an ideal world, the dialogue between scientists and journalists would grow, allowing each side to draw on the strengths of the other. Journalists can do their excellent work of finding, framing, and building stories, while scientists can provide a sort of fact-check to keep the reporting accurate. Maybe we need more scientists speaking up on social media, or fewer barriers keeping the public from understanding science, through fewer paywalls and less jargon. While the controversies of July 2022 were hard to watch, they provide a valuable opportunity to reflect on where the problems lie and how they might be fixed.