Pharmaceutical information and conventional wisdom has held for many years that one cannot become addicted to a medication for depression. If it were possible, this would imply that the reward systems of the brain had responded dramatically and fast to the drug's presence and then demanded a continued supply for more of the drug. That is not the case with antidepressants, but from the lay point of view, if I take a pill, and have to, rather than need to take it again and again or otherwise suffer, then I am addicted, hence the emphasis around the term "addiction". 

For the most part, an addictive drug targets the dopamine-based reward systems of the brain, and when applied to the brain, changes take place in the brain that lead to demand for the substance, sometimes in increasing doses. This is not strictly true of the antidepressants, but Glenmullen explains that there are changes in the brain that mean adaptation to supply or non-supply situations, hence the gamut of medical and psychiatric symptoms which apply in both scenarios.

Historically, this confusion has arisen around the application of the germ theory, medical model in psychiatry.  If I have a disease, I take medication as an ameliorative or restorative treatment, and when the disease entity has retreated, I can stop.  If depression is a disease, the cure is anti-depressants in the same way that antihistamines are the cure for allergenic histamine responses to foreign protein entering the body.  Of course, this is not so, and mood disorders are conditions, as in Major Mood Disorder, rather than major mood illness or sickness, and testing for such conditions is rather more metaphysical and esoteric.

We also know that although antidepressants are found to have an effect on human chemistry almost immediately, as with antipsychotic and other drugs and medications, the benefit takes some time, perhaps two weeks or more, to be felt by the recipient.  Conjecture as to how this may occur includes the role of neurogenesis via brain derived neurotrophic agencies, rather then the activities of the neurotransmitters or their targets necessarily.

Indeed there is emerging evidence to show that knocking out their neuroprotective and neurogenic capacity in this regard, leaving the affinities for serotonin etc intact, renders such medications fairly useless as antidepressants.

The history of antidepressants includes the accidentally discovered and dirty drugs (with regard to their affinities) that were the first to be stumbled upon (tricyclics and MAOI's); the deliberately clean drugs (SSRI's in the Prozac- Paxil genre, reversible MAOI's and so on) and the deliberately dirty (targeting more than one neurotransmitter) drugs that we have today such as Effexor.

For most of us exposed to patients in the sixties, ECT and the anticholinergic tricyclics were all we knew, and patients had dry mouths, shaky hands, poor vision, constipation, urinary retention, sedation, and so on, as we built their drugs up slowly until remission of some symptoms of depression and anxiety.  It was difficult though, and so about 80% of patients just didn't take enough for long enough to get really well, although there was no doubt these drugs worked well, perhaps better than the ones that followed.  However, they could kill, some were cardiotoxic, and some, like the MAOI's, were downright dangerous in the face of dietary mistakes and so on.

We were all then introduced to the Prozac revolution, being sold on the issue that this was a clean drug, free of any real capacity to kill, unlike the previous overdose potential of the accidentally dirty. More convincingly, we were told that one size fitted all, 20mg to start, and after that anyone would benefit.

My colleagues very soon were giving valium with the 20 mg to avoid akathisia and a host of nasty overarousal features.  We learned that fluoxetine was powerfully stimulating, to the extent that for two weeks the patient might feel worse than before treatment started, leaving them desperate to avoid the stimulation side effects.  The rapid introductions of a liquid medication helped us a lot, as we could initially titrate the dose and reduce the weight of symptoms, instead of dissolving the drug in orange juice and varying doses that way.  With Prozac, when we stopped the side effects were long lasting.  When paroxetine (Paxil, Aropax etc) was introduced, we were told it was much quicker to metabolise, and so the side effects lasted barely 24 hours. We didn't cotton on to the fact that this might mean that withdrawal side effects would launch more rapidly and severely.

What this failed to reinforce was that if you missed a day with paroxetine, it felt far worse than if you skipped a fluoxetine, and this wasn't rapid onset of depression again, this was, well, withdrawal, something which our paradigm said should not happen.  Sedation, sexual side effects, loss of sleep and appetite, restlessness and agitation, aggression related to overstimulation, 'zaps' in the head like electricity, cycling moods, a host of side effects attended the drug, changing the dose, and going off it. Nothing was to be seen in the literature, just something we all realised.  And it wasn't in everyone, and neither was improvement necessarily at 20mg, or even 80mg.  Safer, yes, better, no, clear of problems and issues, absolutely not.  Stephen Stahl was quick to point out that these SSRI's were not five similar drugs in one class, but thanks to varying affinities, quite substantially different from one another.  My colleagues Les Koopowitz and Mike Berk wrote of the withdrawal problems in paroxetine in 1995 in Human Pharmacology after observing this for years (it came to their attention in 1992-1993, and Tom Gualtieri had noted problems with Prozac in 1991). They noted that the drugs with the shortest elimination half-lives were the worst, as this made withdrawal symptom onset much more sudden and likely.  This was 10 years before this book was published.

By 2000, Glenmullen had produced his book Prozac Backlash (reviewed in Metapsychology 5: 26). Many others, like Peter Breggin for instance, found themselves on Quackwatch websites, but authors such as Glenmullen, a clinical instructor in psychiatry at Harvard Medical School, were harder to challenge.  Glenmullen won the 2001 Annual Achievement Award from the American Academy for the Advancement of Medicine for his book.

The BBC documentary about paroxetine was sensationalist stuff (2002), but even Thomas Gualtierri in 1991 had warned of the paradoxical effects one might encounter with Prozac, let alone the faster metabolised Paxil, as noted by Koopowitz and Berk, often cited by Glenmullen.

And so the book enters the fray with the FDA and others, driven by a need to buttress drug company and federal agency poverty of information on what might occur in some, sometimes most, of those patients who try to get off their antidepressants.

At first a warning: this book teaches by repetition, based, I think, on his worry, sometimes expressed, that people will dive in to specific chapters without reading others, such as first going to the children's section, rather than reading what came prior to this.  Case histories and tick-box forms with rating scales are the books finest features, painstakingly describing a thorough and commonsense approach to assisting patients escape from the trap of medication.  Overall he is trying to reach the consumers themselves, and hence his tendency to repeat and restate his case in a serial fashion.

The first and sharpest lesson is that symptoms that mimic depression but which occur in the hours, minutes and perhaps days that follow a reduction or change in dosing. This must be considered withdrawal, not the re-presenting of the original illness or an exacerbation: that can take weeks and even months to emerge. Moreover, such withdrawal can occur even when the patient has only been on the medication for a month or so.  The faster the elimination half-life, the more likely is the withdrawal to rapidly occur after altering the dose, and more severely in some cases, with a confusion of medical symptoms interspersed with psychiatric.  Effexor may be one of the worst, with a case described of severe withdrawal after just a single missed dose. Some patients are known to have opened capsules of slow release antidepressants, counting out 280 granules, and taking off just two at a time, in order to avoid withdrawal.  Sertraline, escitalopram, fluoxetene, these may be less violent in terms of potential for withdrawal effects.

Given that children metabolize faster than adults, the chances of withdrawal appear to effectively double in this patient group.

Glenmullen thus embarks on a teaching exercise for professionals and patients alike that is a solid How-To book on taking oneself or others off of the offending drug. Overall, his tone is cautionary, but effective, warning the world at large that there is a science to this, and patient and doctor alike have to be alert and aware, or lives can be dominated by unnecessary symptoms, with undetermined causes if the appropriate stance is not taken.  In order to do this, he repeats and repeats, restates and reworks, and each citation given is repeated often in most chapters, as he comes back to the same issues, again and again.  However, this is not boring, just perhaps a little tedious, but he is serious about this information and wants it conveyed in each and every chapter nuance.

At about US$25 it is pretty much recommended as absolutely necessary for every GP I know, and for many psychiatrists at resident or registrar level, as well as every psychologist and helping profession member on the planet: its just too important a set of data to ignore.

This doesn't mean everyone will suffer withdrawal.  It does however makes the point that those who do are as diverse as those who don't, and so each package has to be tapered according.  This results in a method similar to that now being revealed by the NIHM CATIE study into novel antipsychotic medication, which is finding similarly that we are far from being predictably equal in terms of response to medication. Individual treatment plans have to consider individual withdrawal scenarios, using the many forms and tables he puts forward here.

Reading between the lines, it is clear that Glenmullen spends a lot of time with his patients, and that each client is in receipt of an individually tailored plan for both treatment and termination.

In his afterword, he notes that this book should never have had to be written.  Most drugs were released with comments that withdrawal was in the order of .01%, not the over 60% which has been found subsequently.  So bad is this for drug companies that they have tried to reframe this phenomenon as "antidepressant discontinuation syndrome" which reformulation Glenmullen attributes to Lilly Pharmaceuticals.  Glenmullen focuses on what he calls deceptive processes and semantic gamesmanship, and elaborates on what these have been like within the Paxil and Prozac stables.  The longsuffering FDA also comes under scrutiny, and as we all now know, it has been found sorely lacking in credibility and neutrality. It is well that he writes with anger only at the end of the book, as if this was in the beginning, one would wonder how objective he was being, given the sorry state of affairs he elaborates on at the end.  To his credit, he notes that he has avoided the murky politics of why doctors and patients do not know more about this subject until the end of the book. He makes some suggestions as to how the situation can be addressed.

Reading this book, and its more saddening end section, the snake oil salesman of the past come to mind.  We all work on fairly intimate terms with pharmaceutical companies and their representatives, and some of us benefit financially from honoraria.  I do notice however that many companies that I work with are much more cautious in their sponsorship and marketing, keen to avoid the slur that they are influence peddling and lobbying their products into use. Large scale studies such as the antipsychotic CATIE that look at antidepressants, their efficacy and outcomes on a large scale may clear the scientific air as to what these drugs really accomplish.

The difficulty for all of us that work with what we cannot see, or measure directly will continue to hamper our efforts to provide accurate information on medication and its efficacy and utility.  Beating placebo effects, namely rising in a statistical way beyond the drug and examiner interactions with human clients is always a challenge.  Glenmullen however is making the point that the prior ignorance or downright avoidance of the symptoms of withdrawal when medication is changed or discontinued was and is closely related to the vested interests of pharmaceutical companies and the doctors and researchers that make their living off of them.  Stereotypical statements about the impossibility of dependence had a long half-life of their own, and have taken years to be challenged.  Books like this however do largely stand out on their own even though there is peer-reviewed literature to support them: Glenmullen's is an unpopular position take even though he seldom advocates for an end to such medication attempts, let's say as in the partisan scientology approach to life.

He does however make a stronger stand when children are involved, as in his previous book, and directs parents and doctors to take a closer look at safer alternatives.

Overall, he is advocating for an honest and commonsense approach to antidepressants and their priests, and the highest form of science, human observation with an open mind, but with purpose.

© 2006 Roy Sugarman

Roy Sugarman PhD, Acting Director of Psychology Royal Rehabilitation Centre Sydney; Conjoint Senior Lecturer in Psychiatry University of New South Wales, Australia