What they don’t tell you about antidepressants

For generations, people have known that fish and walnuts are good for the brain. They didn’t need to know why, only that they worked. When queried, they might have pointed out the textural similarity between fish flesh and the grey matter of the brain, or the bi-lobed crenellations of walnut structure. Nowadays we know that it’s the omega-3 oils, and the protective effect on brain neurons, but those findings lagged behind folk knowledge. Which is just to say, people will use whatever seems to work and science has to play catch-up.


So here’s the thing about antidepressants:

These drugs have been developed in the modern age, based on scientific knowledge and testing. They act to increase levels of neurotransmitters in the synapses of the brain – serotonin, noradrenaline, dopamine – which are known to induce feelings of ease and satisfaction, or alertness, or pleasure. That seems to make sense, and these drugs have been shown to be effective, so one tends to assume that the medical profession understand how they work, and that their recommendations on safe dosage, term of use and discontinuation can be trusted.

The reason for this post is because I’ve been taking a low dose of one of these drugs for 8 months and last week I decided I was feeling consistently OK and I wanted to come off it. This led me to go online to check safe ways to quit. What I discovered was conflicting and somewhat disturbing:

  1. Official advice from drug companies and the medical community included the following:
  • For drugs with a short half-life, discontinue by tapering off, reducing by 25-50% of the original dose every 2-4 weeks to avoid ‘discontinuation syndrome’ (that is, withdrawal symptoms). Normally takes 4-6 weeks.
  • Discontinuation syndrome (characterised by flu-like symptoms, insomnia, nausea, poor balance, sensory disturbances, hyperarousal or anxiety) is not dangerous and usually subsides over 1-2 weeks.
  • If you’re on the lowest therapeutic dose (like me), you don’t need to taper at all and can just stop taking the drug.
  1. A vast quantity of unofficial advice (‘folk knowledge’) from the blogosphere indicating:
  • Some people experience withdrawal symptoms for months or years after they stop taking the drugs. Since this is not ‘supposed to’ happen and because symptoms can be similar to mood disorders, doctors may take this to be a relapse of the original condition.
  • Some people, particularly those who’ve been taking the drugs for several years, have been unable to come off the drugs due to horrible withdrawal symptoms.
  • To be on the safe side, it’s best to taper very slowly, at a rate of 10% every 2-4 weeks – and that is 10% off the current dose, not the original dose. This requires careful pill-cutting, opening capsules to remove beads, or creation of solutions to administer increasingly tiny doses, and takes months.
  • The invention of ‘taper strips’, which pre-package pills in reducing doses to suit your needs.
  1. Recent scientific papers and articles which are essentially admitting (according to my inferences and in my own words):
  • The notion that anti-depressants act by correcting a problem with brain chemistry is not correct. If that were the case, peak benefit should be obtained as soon as the drug reaches steady-state concentration in the body (typically within about a week). In practice, peak mood improvement typically takes 6 weeks or more. In other words, they still don’t understand exactly how the drugs work.
  • Drug testing has typically been undertaken over short periods of 6-8 weeks. There is limited data on safety of drug use over the long term or on safe discontinuation.
  • Old advice on tapering was based on a linear reduction in drug concentration (same percentage off original dose at each step). New research indicates that a hyperbolic taper (reduction in smaller and smaller increments over a long period of time) results in a linear reduction in neurotransmitter concentrations in the brain. (So the folk knowledge may have got that right). https://www.nytimes.com/2019/03/05/health/depression-withdrawal-drugs.html
  • When antidepressants are used over the long term, the brain adapts to the constant high levels of neurotransmitters by down-regulating (reducing the number of receptors). When people stop taking the drug it may take a long time for these to switch back on or re-grow (or in some people they may never return to original levels). In other words, it is possible to become permanently dependant on these drugs.

While that all sounds rather scary, I should say that roughly half the people coming off antidepressants report no withdrawal symptoms at all, and of those that do, half do not rate them as serious. That still leaves a sizable minority who experience major issues, though. I think it’s all down to differences in dosage, length of time on the drug, and personal variations in metabolism and brain chemistry.

I should also say that I’m not against the use of anti-depressants. In fact what they do is really quite amazing. I found this fascinating paper online, entitled “How antidepressant drugs act: A primer on neuroplasticity as the eventual mediator of antidepressant efficacy”. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3025168/

What it says, to paraphrase, is that prolonged or severe stress changes the wiring of our brains, increasing activity in the amygdala and impairing function of the hippocampus, prefrontal cortex and downstream structures. This re-wiring is what creates symptoms of anxiety or depression, not a chemical imbalance. (Makes sense to me! See my previous blog posts)

The paper hypothesises that what antidepressants do is produce chemical conditions that encourage synaptic strengthening, dendritic growth and branching, and new synapse formation in the hippocampus. In this way they induce neuroplasticity, which can reverse the neurological changes in response to stress. Since the hippocampus is primarily related to learning and memory, strengthening of its function enables us to properly process memories and learn from them. Neuroplasticity may enable the hardwiring of healthier cognitions, affective responses and behavioural expressions through the learning of healthier cognitive strategies and coping methods.

So that’s pretty cool.

However, now the drug has worked its magic on me, I’m left with the thorny problem of how to come off it without ending up in a worse place than I was to start with.

The next part is my own interpretation of the available information on the particular drug I’m on (mirtazapine 15mg), in an attempt to work out the fastest (safe) way of coming off it. This may be of interest to anyone who is in a similar predicament – but only in a general sense, as a guide to my thought process. I’m not guaranteeing that any of the charts I’ve scrounged off the web, or simply made up, are correct, and the conclusion I came to is based on one particular drug and one particular person (me!) and is unlikely to directly suit anyone else. Or in other words – please don’t try this at home!

So, first up, these pills I take reach peak concentration in about 2 hours and have a half-life of 20-40 hours (average for women 37 hours). In chart form, you get a change in concentration over time like this:

Scan_20190630 (3)

At first I thought that, because it takes a few days for the drug to clear out of my system, by taking the pills every day I might be getting ever-increasing amount of drug in my body. But apparently that’s not how it works. As long as there’s no problem with our metabolism, elimination works faster at higher concentrations, and drug concentration will reach a steady state. For mine, that should happen in about 6 days.

2_steady-state as multiples of half-life

Depending on metabolism, some people may have issues in achieving steady-state, but that usually becomes obvious by either horrible side-effects or just lack of any therapeutic effect. For myself, I’m pretty sure I’m in the normal metabolism range, and the drug seemed to work, so that was OK.

3_drug concentration and metabolism

So far so good. For the next part, I couldn’t find the right chart, so I’m making it up. I’ve drawn what I’m thinking was the intention behind the old (and wrong) tapering advice. I think the intent was for people on higher dosages to step down over periods corresponding to the times to achieve steady state, rather than going cold turkey and coming straight down from a high dosage to zero in one go.


One problem with the above approach is that it assumes there is a direct and linear relationship between drug concentrations in the blood and withdrawal symptoms. It does not recognise that withdrawal symptoms arise from steep drops in neurotransmitter concentrations in the brain, which apparently is not linearly related to the drop in dosage – hence the need for smaller dosage reductions over a longer period.

Judging by the observation that antidepressants typically take a few weeks to show an effect, perhaps this is the time it takes for neurotransmitter levels to reach steady state? Or else, there may be some other unknown effect at work, which is delaying the effect of the drug on mood – and which might also be at play on withdrawal.

Plus, we also need to recognise the down-regulation effect as the brain adjusts itself to the new levels of neurotransmitters. So rather than have a simple recovery curve-to-straight-line like this:

phases of treatment

I suggest it’s more likely to look something like this, with an improvement of mood at first, followed by the brain’s recognition that mood level is becoming too high and the implementation of a down-regulation to bring it back down.

Scan_20190630 (2)

I think this is why they aren’t supposed to give these drugs to people with bipolar or to kids. Either because of over-dosing or because of the differences in children’s brains, they are more likely to overshoot into a state of ‘activation’*.

*Symptoms of SSRI-induced activation syndrome: irritability, agitation, somatic manifestations of anxiety, panic attacks, restlessness, hostility, aggression, insomnia, disinhibition, emotional lability, impulsivity, social withdrawal, akathisia**, odd behaviour, hypomania, paranoia and other psychotic symptoms. **Akathisia: a subjective unpleasant or distressing restlessness and need to move, often accompanied by an inability to sit or stand still. Usually the legs are most prominently affected. People may fidget, rock back and forth, or pace. Prevalence rates of activation are as high as 20% in children with mental retardation, autism, Tourette’s Syndrome, panic disorder and pervasive developmental disorder.

(Just thought I’d add that in, because ‘activation’ sounds an awful lot like the issue I was having – similar to but not quite fitting anxiety or PTSD – which peaked about 3 months into my drug treatment. Just saying.)

Anyhow, to get back to the point, what I would expect to happen on discontinuing the drug would be the following:

  • Synaptic neurotransmitter levels fall
  • For drugs with a short half-life, levels fall faster than the brain can adjust. This triggers withdrawal symptoms such as dizziness, nausea, anxiety
  • Over 2 weeks or so, the brain recognises that neurotransmitter levels are lower than usual and tries to adjust by releasing more
  • For long-term users, the adjustment may not be adequate, owing to down-regulation of receptors. Withdrawal symptoms may continue.
  • Following discontinuation, if adjustment has not kept up with rate of taper, mood will drop below the normal threshold, mimicking a relapse of the original condition.

I’ve tried to illustrate this graphically like this:

Scan_20190630 (4)

What we can do to maximise success would be:

  • Don’t go cold turkey, even coming off the lowest therapeutic dose
  • To minimise withdrawal symptoms, adopt small reductions in dosage and be sure symptoms have gone before each step down
  • For long-term users and those coming down from high dosages, adopt an even longer taper, not only to ease withdrawal symptoms but to avoid apparent relapse in mood disorder

Having said all that, I’m not planning to follow the cautious 10% hyperbolic taper. I really can’t imagine myself chopping pills into smaller and smaller doses over the course of months – and I doubt I’m going to need to. Considering I’ve only been on a low dose and for less than a year, I’m thinking (hoping) I’m a low risk for an apparent relapse, and can go faster than that. So my plan is to taper by about 25% steps, based on making up dosages by combinations of half and whole pills – because the pills are tiny and half is the smallest I can easily go.

This is what I’m planning to try:


The ‘folk wisdom’ is not to alternate half and whole pills like this, and I understand that will result in increased fluctuations in dose. I believe I can mitigate that somewhat by adjusting the times when I take the pills. At the moment I take one pill in the evening. When I’m alternating ½ and whole pills, after the ½ pill I’ll be running too low towards evening so I’ll need to bring the whole pill forward, say to mid-morning or lunchtime. Things get might get tricky when I’m on the 3-day alternations but I’ll play it by ear.

I’m also thinking to offset the final 67% reduction in dose at the end with supplements that might naturally boost levels of serotonin. St John’s Wort acts like an anti-depressant in inhibiting serotonin reuptake (but might take too long to take effect). I can’t take it earlier as it can’t be mixed with anti-depressants. 5-HTP and L-tryptophan are precursors to serotonin in the body and might work faster. The latter is a natural amino acid in food and may be safer than the former. However, I’ve never tried any of those before, so I have no idea whether that will help. Thinking there’s no harm in giving one of them a go.

Well, that’s the plan (…depending on how it goes, I may or may not let you know whether it is successful…). I’d be interested to hear others’ experiences with coming off antidepressants, too.

Wish me luck!



Musings on (my) Mental Health

Looking at what I’ve written here, I’m dumbfounded that I’m actually going to publicly post something so revealing of myself.


I don’t know, maybe because this has been on my mind, and yesterday I finally figured something out. Maybe because keeping silent only increases the stigma, and something here might help someone else. Maybe because I am who I am and I do what I do.

So, here goes…


(illustration of amygdala by Bona Kim)

So… there’s a little almond-shaped area in the brain, located in the anterior temporal lobe just below the pre-frontal cortex and in front of the hippocampus, called the amygdala. This area is responsible for making instant assessments on levels of threat in the environment and raising or lowering our state of arousal accordingly. For me – and for anyone with an anxiety or stress disorder – the amygdala has become over-sensitised. Sometimes small stressors trigger an extreme response and sometimes the response doesn’t shut down properly.

It took me a while to work out that I was ill, probably because I’m on the autism spectrum and my amygdala works a little differently anyway. Neuroscience has yet to provide all the answers, but they’ve noted differences in amygdala size and numbers of neuron connections in some of us. I’m taking this to explain why I’m naturally under-reactive to most things, and over-reactive to a few. Hence my difficulty in distinguishing the boundary between “normal (for me)” and “something wrong”.

Anyway, having eventually established that something was actually wrong, here is my approach to putting it right.

A lot of this I’ve discovered recently and I’m still working through putting it into practice, so it’s early days, but I thought it worth writing down, if only to keep myself on track.

  1. Addressing the Cause

It’s worth reflecting on the reasons why people are vulnerable to developing mental illness. I recently read an interesting book on evolutionary psychiatry –  it’s called “Good Reasons for Bad Feelings” by Randolph M. Nesse.

Good Reasons

In the prologue is stated something seemingly obvious but also (I think) quite profound:

Natural selection shaped emotions such as anxiety, low mood and grief because they are useful …our suffering benefits our genes. There are also good evolutionary reasons why we have desires we cannot fulfil, impulses we cannot control, and relationships full of conflict …evolution explains the origins of our amazing capacity for love and goodness and why they carry the price of grief, guilt, and, thank goodness, caring inordinately what others think about us.

The take-out for those of us with anxiety or stress disorders – the thing to remember in the darker moments – is that our amygdala is our friend; it’s doing its best to keep us safe. If something in our environment has caused it to over-react, we should be looking first at changing our environment.

So when I set about trying to cure myself, my first move was to address all the big problems in my life, trying to reduce major stressors and overcome any barriers to my mental health. The surprise was in realising that my over-sensitive amygdala had already solved the biggest problem of all.

Nearly a year ago, in sudden overnight haste, with no prior planning or fore-thought, I left my husband. From the point of view of my long-term health and wellbeing, it was probably the best thing I ever did. But would I have left him if I’d been in my right mind at the time? Probably not. I didn’t know it but I was already ill. Bless my hypersensitive amygdala, it caused me to freak out intensely enough and for long enough to leave. It saved me.

Unfortunately, once mental illness develops, one can’t just shake it off and ‘get over it’. Just like a physical illness, it needs treatment and takes time to heal.

Probably the most common treatment method is Cognitive Behavioural Therapy (CBT), but when I first read about it I was dismissive. CBT seemed to be based on “thinking yourself happy”, and therefore of no use in addressing the real life problems underlying one’s mental state. However, after going through the process of pruning the problems from my life, after doing everything I could to set myself back on track, I was frustrated to find that I was still having acute stress reactions. Over nothing.

This is when it hit home to me that, although they call it mental illness, it’s essentially a physiological problem in the brain. This is why someone with no apparent worries can still be stuck with anxiety, or why someone with minimal stress in their life can still have a stress disorder. Even after getting our lives back on track, there’s a malfunctioning process in the brain which still needs to be fixed.

  1. Setting Up for Success

I think of this illness as being like heart disease. If you have a natural vulnerability to heart disease what’s the sensible thing to do? Even if there’s a potential treatment, you’d want to look after your heart in the meantime, so you’d do what you needed to do: eat healthily, limit stress, increase exercise.

Didn’t I say that mental illness is a physiological problem? The brain is a critical organ just like the heart, so the same principle applies.  I’m breaking down what I need to do to manage my mental health five ways:

  • Maintaining motivation. Regardless of inherent vulnerabilities, my natural state is a state of good health and I don’t have to accept less.
  • Exercise. It’s not about fitness or weight loss. It’s about easing a prolonged state of arousal by going outside and walking… and walking some more, until the nervous energy burns out. Adrenaline is not our friend.
  • Watching the diet. Two things seem to be key:

(a) Cutting the sugar. Yes, I know the brain runs on glucose, but with blood sugar, steadier is better. A sugar high is always followed by a sugar low. (Not that I’ve managed to quit the white poison yet – this is a work in progress).

(b) Brewed coffee is my nemesis. Every time I have an episode I swear I’ll quit, but then comes the downcycle and I’m as sluggish as a… slug?… and I start on the coffee again because I absolutely need it. And I can handle it fine… until I can’t. (This is another work in progress).

  • Taking medication. As long as it’s understood that there is no medication that will cure an over-sensitised amygdala. None, nada. What medication can do is ease the post-meltdown blues enough to get me to work in the morning.
  • Quitting the guilt-trips. Every mother will know exactly what I mean, but the ‘aeroplane oxygen mask’ rule holds true. We need to look after our own health to be fit to look after our kids. It feels selfish, but it really isn’t. And I try not to be embarrassed at treating myself carefully when I ‘don’t seem ill’; that just means I’m doing a good job.
  1. Uncovering the Cure

For a while, it seemed to me that my nervous system was broken and there was nothing I could do about it, but I’m reliably informed that’s not the case. We just have to de-sensitise the amygdala.

This process has been demonstrated with people with specific phobias. They undertake exposure therapy, which involves repeatedly exposing themselves to the object or situation that triggers them until they become de-sensitised. Apparently, it works every time.

The complicating factor in my case was that I could not pin down a discrete object or situation that was triggering me. With a bit of thought I could identify some connecting factors to my meltdowns, but they were internal – a configuration of emotions. I was stymied for a while by the near-infinite number of situations with potential to trigger those emotions.

Now I’ve realised identification of triggers is only the starting point. Because the amygdala is essentially activated by a fear response, I need to dig a little deeper into working out which subconscious fears underlie my emotional responses. Exposure therapy for me will be to recognise and address those underlying fears.


So yes, I was surprised to discover that CBT might be useful after all. The useful part is the theory related to identifying one’s Core Beliefs, and how adherence to those beliefs can create cognitive distortions. In other words, sometime in the past, I developed a belief about myself and/or the world which led me to fear certain situations. The brain, being the highly evolved, tricky thing it is, ‘creates’ logical explanations for one’s behaviour, to give oneself the illusion of self-determination rather than recognising that our actions stem from old fear conditioning. These are the cognitive distortions.

It takes a certain amount of self-examination to work past the ‘logical’ explanations which we think underlie our behaviour, to the fear response below that, and then back to the original Core Belief which produced that fear response. But that’s what I need to do: to get right back to the Core Belief and work on changing it, because that (hopefully) will eliminate the fear response entirely, together with the need to employ cognitive distortions to maintain my illusion of self-determination.

Yes, I know it sounds a bit hippy, or New Age or whatever. I didn’t think I was into all that stuff, either. Peace, sister 🙂

hippies 2

  1. Working the Cure

If you want to understand how it works, here is an example (dealing with one of my common and more obvious triggers):

The triggering emotional configuration:

  • upset/anger over injustice, intense frustration and a sense of powerlessness.

Common element in triggering events:

  • Someone in a position of authority has made a decision which affects me and about which I was not consulted or my opinion not taken into account.

Actual events that triggered me:

  • My employer requested a doctor’s report on aforementioned health condition. I was fine with that idea for a few days, before having an epic meltdown over the thought that I would have no control over what the doctor wrote, and my employer might use it to make decisions about my employment.
  • My employer decided to cancel my corporate card (which was mostly my fault for not completing training on time). Minor meltdown at having no recourse.
  • The rental agency for my house agreed to extend my agreement another year (excellent news!) but with a 17% rent increase. No meltdown, but I got a bit keyed up about the size of the increase.

‘Logical’ explanations for my reaction (Cognitive Distortions):

  • This is unfair (fallacy of fairness)
  • This is going to cause me so much trouble (filtering/catastrophizing)

(Now it starts getting painful – press the sore point.) Underlying fears:

  • Loss of control over aspects of my life –> imposed changes, uncertainty

(Ah. Hello again, Resistance to Change. I’m not ready to attempt addressing this one – let’s move on to Core Beliefs.)

Underlying Core Belief (tease it out like a splinter):

  • I can’t express myself effectively and my opinions are not worthy of consideration; I have no power or means to influence those who are making decisions affecting me.

What actually happened:

  • On querying my employer I was reassured that the doctor’s report was just for information and I would be consulted on any decisions affecting my employment. My doctor discussed with me what to put in the report.
  • After sending an email explaining why I missed the training deadline and appealing the decision to cancel my corporate card, I was offered another opportunity to do the training.
  • I checked typical rent amounts in my suburb and the rate of inflation, and sent an analysis to the rental agency. They accepted a lower rental increase.


  • My Core Belief has been demonstrated to be false. I’m perfectly capable of expressing my opinion, and that opinion has been deemed worthy of consideration on several occasions. If decisions affecting me are unfair, I’m able to mount an appeal with a reasonable chance of success.
  • I’m a lot more resourceful, capable and effective than I give myself credit for.
  • New Core Belief: There’s no need to panic, I can handle this 🙂

See, that’s how it’s done.