Understanding who benefits from antidepressants
- Cathryn Lewis
- Oct 29
- 4 min read
Why is the use of antidepressants so controversial, and how could we prescribe them more effectively? This question is a priority in my research at King’s College London, where I work as professor of statistical genetics. My academic background is in statistics, and I am passionate about using genetics to improve the diagnosis and treatment of mental health conditions. Together with people with lived experience of depression, my research team integrates clinical and biological data to study these important questions.
Antidepressants are some of the most prescribed drugs in the UK. The number of NHS antidepressant prescriptions increased from 36 million in 2008 to 88 million in 2023, at a cost of £230 million. Antidepressants are widely and successfully used to treat depression and other conditions, but their use is controversial. Many people suffer from side effects and need to stop taking them. Others get no benefit from their first prescribed antidepressant, and it can take several rounds of trial-and-error prescribing to find an effective drug.
Antidepressant studies
In clinical trials, participants are randomised (allocated by chance) to receive either an antidepressant or a placebo (a dummy pill). Response rates are higher for people on the drug arm of the trial than the placebo arm, but the differences between groups are low. Placebo rates are high, and there are many reasons for this. Some people’s episodes of depression will naturally reduce over time. And - curious but true - placebos are powerful medicine: being cared for, with the hope of recovery, can have a positive effect on people’s health itself. Across trials, the consistent if modest effect of antidepressants - with more people responding to the antidepressant than to placebo - is reassuring that antidepressants can be helpful. Similar evidence is seen in other studies - when UK Biobank participants were asked whether specific antidepressants had helped them, at least 7 out of 10 people said that the drug had helped.
Clinical trials are stringently designed to ensure that no biases creep in, so that we can have confidence in the results of the trial. To measure how well antidepressants work, patients and clinicians complete questionnaires of symptoms of depression, and a score, summing across the number of symptoms present, is used to assess the severity of depression before and after taking the antidepressant (or the dummy pill). A widely used trial outcome is “remission”, where a participant is no longer depressed by the end of the trial, so their depression symptoms are very low or absent.
In clinical trials, usually only about one-third of people reach remission, which is much lower than the 70% of UK Biobank participants who say that antidepressant drugs helped them. Why are those figures so different?
Remission is a stringent outcome, requiring that someone reaches a depressive symptom score threshold at a specific week of treatment. In contrast, asking people retrospectively about their treatment is a more general question about whether they felt their depression improved while they were taking the drug. Feeling better does not mean you are completely well and free from symptoms of depression.
How can we use antidepressants better?
So, on average, antidepressants work and, on average, people find them helpful. But most patients are not ‘average’. We believe there is scope for improving prescribing of antidepressants so that people are prescribed a drug they find tolerable and which reduces their depression symptoms.
Antidepressant prescribing in the UK is currently a one-size-fits-all all procedure: NICE guidelines indicate an SSRI (selective serotonin reuptake inhibitor; the drug believed to work by increasing serotonin) as the first-choice drug for moderate to severe depression, moving to a second SSRI if there is little response after a few weeks, and then exploring further options as necessary.
But responding to a drug is not a yes/no event. Some people respond very well with a steady reduction in symptoms over the weeks and months they take the antidepressant. Others have no response, or only a modest reduction in their depressive symptoms. Can we tell who is who? Sadly, not yet: there is little information to inform personalised prescribing for antidepressants, where a drug is chosen because it is the best fit for that specific patient, likely to give a response with minimal adverse events.
The real question for antidepressants is not whether they work, but for whom they work. There is much scope for improving the matching between patient and antidepressant, providing patients with a drug where the risk of severe side-effects is low, and the chance of improving their depression is high. But while we think such discriminating factors do exist, further research studies are needed to identify them. Given the heavy burden that depression places on people across the world, shouldn’t this be an important research priority?
Genetics and antidepressants
We know that genetics - our DNA that is present in every cell in our body - plays a role in response to antidepressants. Several key genes, including CYP2D6 and CYP2C19, metabolise antidepressants, controlling how fast the drug is processed and how long it remains in your body. A drug can be a poor match for you - either because you metabolise the drug slowly (poor metaboliser), giving stronger side-effects, or because you metabolise the drug too fast (rapid metaboliser) and the drug disappears before it has had time to work. These pharmacogenomic tests are a good start towards better prescribing of antidepressants, but they are incomplete. My research team aims to discover what other genes contribute to the response to antidepressants, and how these can be used to improve antidepressant prescribing.
In our Wellcome-funded AMBER study, we are applying informatic, genetic, and cellular approaches to unpick why antidepressants work, and for whom they work, focusing on selective serotonin reuptake inhibitors. In informatics, we apply artificial intelligence to patient electronic health records, giving the rich data and large groups of people needed to study antidepressant prescribing. In genetic studies, we look at how biology affects how we respond to antidepressants, measuring the genes that we inherit or changeable biological substances in our blood, such as methylation, proteomics and metabolomics. Finally, laboratory experiments allow us to test the effect of antidepressants in human cells. Together, we hope this research will enable us to better understand antidepressants and enable us to use them more effectively to help treat depression.
Conflict of Interest: Cathryn Lewis sits on the Scientific Advisory Board for Myriad Neuroscience, which develops the GeneSight test.
