Our umbrella review that revealed no links between serotonin and depression has caused shock waves among the general public, but been dismissed as old news by psychiatric opinion leaders. This disjunction begs the questions of why the public have been fed this narrative for so long, and what antidepressants are actually doing if they are not reversing a chemical imbalance.

Before I go on, I should stress that I am not against the use of drugs for mental health problems per se. I believe some psychiatric drugs can be useful in some situations, but the way these drugs are presented both to the public and among the psychiatric community is, in my view, fundamentally misleading. This means we have not been using them carefully enough, and crucially, that people have not been able to make properly informed decisions about them. 

Much public information still claims that depression, or mental disorders in general, are caused by a chemical imbalance and that drugs work by putting this right. The American Psychiatric Association currently tells people that: “differences in certain chemicals in the brain may contribute to symptoms of depression”. The Royal Australian & New Zealand College of Psychiatrists tells people: “Medications work by rebalancing the chemicals in the brain. Different types of medication act on different chemical pathways.”

In response to our paper finding that such statements are not supported by evidence, psychiatric experts have desperately tried to put the genie back in the bottle. There are other possible biological mechanisms that could explain how antidepressants exert their effects, they say, but what really matters is that antidepressants ‘work’.

This claim is based on randomised trials that show that antidepressants are marginally better than a placebo at reducing depression scores over a few weeks. However, the difference is so small it is not clear it is even noticeable, and there is evidence it may be explained by artefacts of the design of the studies rather than the effects of the drugs. 

The experts go on to suggest that it does not matter how antidepressants work. After all, we do not understand exactly how every medical drug works, so this should not worry us.

This position reveals a deep-seated assumption about the nature of depression and the action of antidepressants, which helps to explain why the myth of the chemical imbalance has been allowed to survive for so long. These psychiatrists assume that depression must be the result of some specific biological processes that we will eventually be able to identify, and that antidepressants must work by targeting these.

These assumptions are neither supported nor helpful. They are not supported because, although there are numerous hypotheses (or speculations) other than the low serotonin theory, no consistent body of research demonstrates any specific biological mechanism underpinning depression that might explain antidepressant action; they are unhelpful because they lead to overly optimistic views about the actions of antidepressants that cause their benefits to be over-stated and their adverse effects to be dismissed.

Depression is not the same as pain or other bodily symptoms. While biology is involved in all human activity and experience, it is not self-evident that manipulating the brain with drugs is the most useful level at which to deal with emotions. This may be something akin to soldering the hard drive to fix a problem with the software. We normally think of moods and emotions as being personal reactions to the things going on in our lives, which are shaped by our individual history and predispositions (including our genes), and are intimately related to our personal values and inclinations. Therefore we explain emotions in terms of the circumstances that provoke them and the personality of the individual. To over-ride this common-sense understanding and claim that diagnosed depression is something different requires an established body of evidence, not an assortment of possible theories.  

Models of drug action

The idea that psychiatric drugs might work by reversing an underlying brain abnormality is what I have called the ‘disease-centred’ model of drug action. It was first proposed in the 1960s when the serotonin theory of depression and other similar theories were advanced. Before this, drugs were implicitly understood to work differently, in what I have called a ‘drug-centred’ model of drug action.  In the early 20^th century, it was recognised that drugs prescribed to people with mental disorders produce alterations to normal mental processes and states of consciousness, which are superimposed onto the individual’s pre-existing thoughts and feelings. This is much the same as we understand the effects of alcohol and other recreational drugs. We recognise that these can temporarily over-ride unpleasant feelings. Although many psychiatric drugs, including antidepressants, are not enjoyable to take like alcohol, they do produce more or less subtle mental alterations that are relevant to their use.

This is different from how drugs work in the rest of medicine. Although only a minority of medical drugs target the ultimate underlying cause of a disease, they work by targeting the physiological processes that produce the symptoms of a condition in a disease-centred way.

Painkillers, for example, work by targeting the underlying biological mechanisms that produce pain. But opiate painkillers may work in a drug-centred way too, because, unlike other painkillers, they have mind-altering properties. One of their effects is to numb emotions, and people who have taken opiates for pain often say they still have some pain, but they do not care about it anymore. In contrast, paracetamol (so often cited by those defending the idea that it does not matter how antidepressants work) does not have mind-altering properties, and therefore although we may not fully understand its mechanism of action, we can safely presume it works on pain mechanisms, because there is no other way for it to work.

Like alcohol and recreational drugs, psychiatric drugs produce general mental alterations that occur in everyone regardless of whether they have mental health problems or not. The alterations produced by antidepressant vary according to the nature of the drug (antidepressants come from many different chemical classes – another indication that they are unlikely to be acting on an underlying mechanism), but include lethargy, restlessness, mental clouding, sexual dysfunction, including loss of libido, and numbing of emotions. This suggests they produce a generalised state of reduced sensitivity and feeling. These alterations will obviously influence how people feel and may explain the slight difference between antidepressants and placebo observed in randomised trials.

Influences

In my book, The Myth of the Chemical Cure, I show how this ‘drug-centred’ view of psychiatric drugs was gradually replaced by the disease-centred view during the 1960s and 70s. The older view was erased so completely, that it seemed people simply forgot that psychiatric drugs have mind-changing properties.

This switch did not occur because of scientific evidence. It occurred because psychiatry wanted to present itself as a modern medical enterprise, whose treatments were the same as other medical treatments. From the 1990s, the pharmaceutical industry also started to promote this view, and the two forces combined to insert this idea into the minds of the general public in what has to go down as one of the most successful marketing campaigns in history.

As well as wanting to align with the rest of medicine, in the 1960s the psychiatric profession needed to distance its treatments from the recreational drug scene. Best-selling prescription drugs of the period, amphetamines and barbiturates, were being widely diverted onto the street (the popular ‘purple hearts’ were a mixture of the two). So it was important to emphasise that psychiatric drugs were targeting an underlying disease, and to gloss over how they might be changing people’s ordinary state of mind.

The pharmaceutical industry took up the baton following the benzodiazepine scandal in the late 1980s. At this time it became apparent that benzodiazepines (drugs like Valium- ‘mother’s little helper’) caused physical dependence just like the barbiturates they had replaced. It was also clear they were being doled out by the bucket load to people (mostly women) to medicate away the stresses of life.  

So when the pharmaceutical industry developed its next set of misery pills, it needed to present them not as new ways of ‘drowning one’s sorrows’, but as proper medical treatments that worked by rectifying an underlying physical abnormality. So Pharma launched a massive campaign to persuade people that depression was caused by a lack of serotonin that could be corrected by the new SSRI antidepressants. Psychiatric and medical associations helped out, including the message in their information for patients on official websites. Although marketing has died down with most  antidepressants no longer on patent, the idea that depression is caused by low serotonin it is still widely disseminated on pharmaceutical websites and doctors are still telling people it is the case (two doctors have said this on national TV and radio in the UK in the last few months).

Neither Pharma nor the psychiatric profession has had any interest in bursting the chemical imbalance bubble. It is quite clear from psychiatrists’ responses to our serotonin paper that the profession wishes people to continue under the  misapprehension that mental disorders such as depression have been shown to be biological conditions that can be treated with drugs that target the underlying mechanisms. We haven’t worked out what those mechanisms are yet, they admit, but we have plenty of research that suggests this or that possibility. They do not want to contemplate that there might be other explanations for what drugs like antidepressants are actually doing, and they do not want the public to do so either.

And there is good reason for this. Millions of people are now taking antidepressants and the implications of discarding the disease-centred view of their action are profound. If antidepressants are not reversing an underlying imbalance, but we know that they are modifying the serotonin system in some way (though we are not sure how), we have to conclude they are changing our normal brain chemistry – just like recreational drugs do. Some of the mental alterations that result, such as emotional numbing, may bring short-term relief. But when we look at antidepressants in this light we immediately understand that taking them for a long time is probably not a good idea. Although there is little research on the consequences of long-term use, increasing evidence points to the occurrence of withdrawal effects which can be severe and prolonged, and cases of persistent sexual dysfunction.

Replacing the serotonin theory with vague assurances that more complex biological mechanisms can explain drug action only continues the obfuscation, and enables the marketing of other psychiatric drugs on equally spurious grounds. Johns Hopkins, for example, is telling people that ‘untreated depression causes long-term brain damage’ and that ‘esketamine may counteract the harmful effects of depression.’ Quite apart from the damage to people’s mental health by being told they have, or will soon get brain damage, this message encourages the use of a drug with a flimsy evidence base and a worrying adverse effect profile.

The serotonin hypothesis was inspired by the desire of the psychiatric profession to regard its treatments as proper medical treatments and the need of the pharmaceutical industry to distinguish its new drugs from the benzodiazepines that, by the late 1980s, had brought the medicating of misery into disrepute.  It exemplifies the way that psychiatric drugs have been misunderstood and misrepresented in the interests of profit and professional status. It is time to let people know not only that the serotonin story was a myth, but that antidepressants change the normal state of the body, brain and mind in ways that may occasionally be experienced as useful, but may be harmful too.