Antidepressants haven’t had a good press of late. The drugs have been labelled as ineffective, harmful, and no better than taking a placebo or dummy pill. Doctors have been accused of over-using antidepressant medication by some, and not prescribing in sufficient doses by others.
So the publication of research in the Lancet recently, a meta analysis of more than 116,000 adults with moderate to severe depression, looking at their response to antidepressants, is a major piece of scientific research aimed at informing the debate. Some 522 double-blind randomly controlled trials of 21 commonly used antidepressants were included, making the study the largest ever meta-analysis in psychiatry.
The Oxford University researchers looked at two principal outcomes: the number of patients who had a reduction in depressive symptoms of 50 per cent or more on a rating scale measured over an eight-week period; and acceptability as measured by the proportion of patients who withdrew from the studies for any reason by week eight.
While all 21 antidepressants were found to be more effective than placebo, the power of the study allowed the researchers to rate the most and least effective types. Interestingly, fluoxetine, the granddaddy of a modern group of drugs called selective serotonin reuptake inhibitors (SSRIs) and marketed under the trade name Prozac, was one of the least effective. However, the drug was one of the most tolerable identified in terms of side-effects.
The research gives patients and doctors some heretofore unavailable information about the relative efficacy and tolerability of various antidepressant medications. By using meta-analysis and including unpublished data, the authors have worked hard to reduce individual study bias. However almost eight in 10 of the trials included in the analysis were funded by the pharmaceutical industry. But the biggest weakness in the study is the eight-week cut off point for the assessment of efficacy and tolerability, which is not a long enough time frame to reflect coal-face practice.
But, in fairness to the authors, they openly acknowledge the limitations and potential biases in their data.
Commenting on her team's results, Andrea Cipriani, the lead author, noted the debate over antidepressants has unfortunately often been ideological. Some doctors and patients have doubts over whether they work at all and point to a significant placebo effect. And some patients do not want to take drugs for a psychological condition.
Talking therapies such as cognitive behavioural therapy (CBT) are as effective as antidepressants in mild to moderate depression. Around 60 per cent of people respond by about two months to medication with a 50 per cent reduction in their symptoms. However, side-effects cause about 80 per cent of people stop antidepressants within a month.
In Ireland, there is insufficient access to timely CBT, especially for patients with medical cards. This has led to a probable overuse of antidepressants by doctors who are unable to access alternative treatments for their patients.
However for severe depression there is no doubt as to the central treatment role of antidepressants. For people who cannot eat or sleep, who have lost a considerable amount of weight and cannot carry on a conversation with friends and family, they are, in my experience, life-savers. At this severity, depression is a very physical disease; people have such low energy levels it makes engaging in psychotherapy almost impossible.
And while no one has come up with evidence to prove the theoretically attractive idea of depression being the result of a chemical imbalance in the brain, feeding a severely depressed body neurotransmitters seems to work.
The ultimate key to recovery though is to engage with CBT or similar talking therapy as soon as possible, develop a regular exercise habit, follow a healthy balanced diet, seek out the company of friends, be mindful and listen to music.
Drugs are never the only treatment for depression. But for some of us they are definitely part of the solution.
mhouston@irishtimes.com
