Healthcare should not be driven by making money

Costs of commodification

Pharmaceutical companies routinely fail to publish negative results of research into new drugs. James Linney looks at the case of antidepressants

Towards the end of February, The Lancet, the world’s best known medical journal, published a meta-analysis,1 which aimed to answer a question that has caused controversy within psychiatry for the past 50 years. Namely, do antidepressants work?2 In this article I will look briefly at some of the limitations of the study; but primarily I will aim to show how it - and the difficulties in ascertaining the usefulness of antidepressants in general - is a devastating illustration of how the pharmaceutical industry influences and deceives the medical profession. This is something that has consequences for the reliability of all evidence-based medicine.

In reporting this study, the mainstream media were pretty unanimous in their coverage. The stories in The Guardian, The Independent and the BBC were essentially summed by the headline in the Daily Express: “Antidepressants do work and millions more should be on them.”3 This is just another example of why you should never rely on the media to accurately convey the results of any clinical papers. So, before we turn to the role of pharmaceutical companies in antidepressant research, let us see how flaws in the methodology of this study actually undermine any such conclusions.

Firstly, it collected information from 522 studies of 21 different antidepressants, but it limited the responses to people using the medications for just eight weeks. A strange decision, you might think, given that in practice most guidelines suggest using them for at least six months and in most cases they are prescribed for much longer - commonly for many years. One reason for this might have been that there is surprisingly little data on long-term effects of antidepressants. Of what is available, the main body of evidence suggests that, while antidepressants can have some short-term benefits, the longer they are used, the less helpful they are - and may actually be harmful. The evidence for this theory is set out convincingly in Robert Whitaker’s book Anatomy of anepidemic.4 Essentially it boils down to this: long-term antidepressant use alters the constitution of the brain, leaving those treated getting less benefit and more vulnerable to relapse if the medication is stopped. Thus, people may soon need higher and higher doses and end up going from one antidepressant to another, each one giving less benefit.

Secondly, the results of the study are presented in a way that falsely inflates the positive effects of the antidepressants, compared to a placebo (a sugar pill). Without getting too technical, the main results present statistical data using secondary outcomes from the studies, not the primary data (ie, the measures that each trial originally set out to record), meaning that the conclusions are much less reliable. When we look at the primary data, as the authors of this study do in their appendix, the results reveal much more modest benefits.

To summarise, the study can only tell us that, for most people with major depression, antidepressants are marginally better than taking a placebo after eight weeks. This in itself, although not new information, is of some usefulness. However, presented in the study is another and, I would argue, much more significant limitation. In analysing all of the antidepressant trials, the authors rated 82% as having a moderate to high risk of bias, with 78% of the trails being funded by pharmaceutical companies, and many of the remainder not revealing where their funding came from at all. In other words, the vast majority of this research was carried out by companies with the aim of manufacturing a profitable commodity, not of improving the treatment of people suffering from depression. Of course, apologists for capitalism tell us that these two things are not mutually exclusive, but, by taking a look at how antidepressants were developed and the role that the pharmaceutical industry has had in their promotion, we will be able to illustrate some of the harms of commodity-based research.

Exponential increase

I realise that it may come as a surprise to many that the question of antidepressant effectiveness has not already been resolved - especially given the enormous number of people who have been treated with antidepressants since their introduction in the 1960s. Their use has increased exponentially since the drug company, Eli Lilly, first put Prozac (fluoxetine) on the market in 1988. In the UK there has been an annual increase in the number of antidepressants prescribed each year, and more than a doubling over the past decade. In 2016 this equated to 64.7 million prescriptions made by the national health service.5 In America the numbers are even more startling, with the US National Centre for Health Statistics reporting that one in eight people over 12 years old take antidepressants - a rise of 65% over a 15-year period.6

Prozac was initially developed as a potential weight loss treatment - a use for which it proved inadequate. At the time psychiatric drugs were not considered to have a high profit potential. This was partly due to a crisis in psychiatry brought about by the use of anti-anxiety medications called benzodiazepines. Before Prozac, benzodiazepines such as Xanax (alprozam) and Valium (diazepam) were heavily promoted by the pharmaceutical companies as a magic bullet for all mood-related issues. They particularly targeted women for their new ‘happiness pills’ (hence their nickname of ‘mother’s little helper’) and they promised that a few daily doses could cure all anxieties and worries.

Unfortunately, the bubble soon burst when it became evident that benzodiazepines gave short-term sedative effects and anxiety relief, but their long-term use resulted in potent addiction and increased anxiety. From 1968 to 1981 Valium was the world’s best-selling drug, the legacy of which was a huge increase in the number of people presenting to hospitals with acute withdrawal syndrome and worsening anxiety. The reality of the ‘happiness pill’ was a miserable addiction for millions of people - and a whole new story of pharmaceutical deception.

Prozac, was one amongst several, very similar ‘selective serotonin reuptake inhibitors’ (SSRIs) that were developed and promoted as the next big miracle psychoactive drugs in the late 1980s and early 1990s. Serotonin is a neurotransmitter that, it was theorised (mostly in pharmaceutical companies’ promotional booklets), is deficient within certain areas of the brain in people with depression. The drug companies pushed this theory in a big way - here was potentially a measurable chemical deficiency that could be treated and responses measured. And so the ‘chemical imbalance’ myth was born and it was used (and still is) to justify the ever increasing use of SSRIs.

Depression, the narrative went, was now an illness - like, say, diabetes, where sufferers simply needed to take long-term or even life-long medications to keep their brain ‘equalised’. The only drawback was that there is no evidence, and never was any, to back this theory up. But this was not too much of a concern to the drug companies, which have always spent far more on their sales and promotions than they have on research. So it was not long before SSRIs were some of the most profitable drugs ever developed.


Let us now take a look at some of the strategies drug companies have used in influencing clinical research and subsequently selling these drugs, which are still the most prescribed antidepressants on the planet.

Firstly, before anything else, the drug companies needed to get their drugs licensed and they do not want to let anything as petty as seriously harmful side effects get in the way. Yet, soon after they had hit the market, some clinicians started to notice a link between SSRIs and increasing thoughts of suicide in those taking them. This link is documented very thoroughly in David Healy’s book, Let them eat Prozac.7 Healy, a professor in psychiatry at the University of Cardiff, was one of the first to notice that some patients on SSRIs soon developed severe agitation and suicidal thought - completely out of character and proportion to their previous symptoms. Drug companies denied that this was the case, their defence being that any suicidal thoughts were simply a part of the disease progression.

It has subsequently been revealed though that from a very early stage they were well aware of the potential harmful effects of SSRIs, but they simply did not publish any of the data. In 2005 a leaked Eli Lilly memo, written in 1989, notes a finding of increased suicidality in their early Prozac trials. Similarly GlaxoSmithKline developed another SSRI called Paxil (paroxetine) that it marketed to children, and it was later also found to have withheld safety data from clinicians. After decades of campaigning, court cases and harm to those taking them, drug companies were forced to put a warning on SSRI medications, stating that they can increase suicide in “people under 25 years of age”.

As well as simply burying trials that reveal harmful side effects, drug companies do not in any case routinely publish trials that fail to produce positive results (ie, favourable for them). Surreally, there is no obligation to publish studies that give negative results. So drug companies often simply carry out multiple similar trials until they eventually get favourable results - and then publish only these. This practice is actually encouraged by medical journals, which much prefer to publish positive result studies than negative ones. As highlighted by Ben Goldacre’s very useful book, Bad pharma, in 2008 a group of researchers wanted to check which trials had been published for antidepressants that came onto the market between 1987 and 2004. They found that of 74 studies there was an even split between positive and negative results. As Goldacre states,

… Thirty-seven of the positive trials - all but one - were published in full, often with much fanfare. But the trials with negative results had a very different fate: only three were published. Twenty-two were simply lost to history, never appearing anywhere.8

These are just a couple of ways that pharmaceutical companies have manipulated research into antidepressants. They give us a hint as to why, more than 40 years after they first started being prescribed, and although they are now being given to millions more people every year, we still do not have a complete picture as to how useful they are - or how harmful they may be in the long term.

My point is not that antidepressants do not work - the limited evidence we have suggests that they can be a useful tool in treating some patients with major depression. But the legacy of decades of pharmaceutical companies promoting these medications, whilst simultaneously manipulating and withholding research, is that they are massively over-prescribed, unnecessarily putting people at risk of harm and distracting us from the real causes of, and appropriate treatments for, depression.

The story of the development and promotion of antidepressant medications contains within it a thread that runs through all scientific research: namely, the destructive consequences of commodification.


1. A meta-analysis attempts to gather data from many different studies with common primary measures in an attempt to get a more statistically accurate result.

2. A Cipriani, TA Furukawa et al ‘Comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: a systematic review and network meta-analysis’ The Lancet February 21.

3. www.express.co.uk/life-style/health/922423/depression-news-antidepressants-drugs-work-study-oxford-university-UK-mental-health.

4. R Whitaker Anatomy of an epidemic New York 2015.

5. www.theguardian.com/society/2017/jun/29/nhs-prescribed-record-number-of-antidepressants-last-year.

6. www.cdc.gov/nchs/products/databriefs/db283.htm

7. D Healy Let them eat Prozac New York 2004.

8. B Goldacre Bad pharma London 2012, p20.