First written: April 2017. Last updated: December 2017.

Introduction

Tranylcypromine (brand name: “Parnate“) is a nonselective and irreversible monoamine oxidase inhibitor (MAOI) used to treat major depressive disorder. See this previous post of mine about MAOIs (written in 2013). Prophetically enough, I wrote: “So my current plan is to first try Moclobemide and then Selegiline. If I don’t die, I’ll start with tranylcypromine and then live happily ever after.” It was meant as humorous hyperbole, but it’s not that far from the truth. (However, please also see my December 2017 update.)

I’ve been taking tranylcypromine (TCP) for almost 2 years now (with a break of about 3 months) and wanted to summarize my experiences with it because, out of the ≈10 antidepressants I’ve tried over the years, it was by far the most effective one. Overall, I think that tranylcypromine might be the most powerful antidepressant on the market. Tragically, few psychiatrists prescribe tranylcypromine because they labor under the misinformed view that it’s highly dangerous. More on this below. My mind shudders at the amount of suffering that could have been prevented if they were less ignorant.

Probably the strongest Bayesian evidence for tranylcypromine being one of the most powerful antidepressants can be found in this astonishing post by the psychiatrist Scott Alexander. I can also highly recommend Ken Gillman’s website. He is a clinical pharmacologist and probably knows more about MAOIs and their (supposed) dangers than everyone else on the planet.

Tranylcypromine is safe

Although tranylcypromine is an extremely powerful antidepressant, it is rarely prescribed. In fact, it’s not even approved in many (European) countries such as Switzerland, Poland, etc.. Why is it so rarely prescribed? Probably because most psychiatrists have a completely exaggerated view of its dangers and side-effects and believe that its toxicity potential lies somewhere between cyanide and strychnine. Ok, now I’m exaggerating myself.

Still, it’s really quite disturbing just how much misinformation about the dangers of MAOIs is out there; even usually accurate sources like Wikipedia or patient information leaflets contain ludicrous errors. For example, many psychiatrists think (and Wikipedia and others back them up) that it’s extremely dangerous and often lethal to combine MAOIs with all tricyclics including doxepin, amitryptiline, trimipramine (and also trazodone to some extent).

Unfortunately, that’s all nonsense. See Ken Gillman’s excellent post on the topic. And it’s not just harmless misinformation. As I’ll explain below, insomnia is one of the most common side effects of tranylcypromine and drugs like doxepin, trazodone or trimipramine are often effective hypnotics.

Advantages of tranylcypromine

Most of the beneficial effects of tranylcypromine – like greater productivity, motivation, energy, optimism, confidence, “agency”, resilience, self-control – seem to be related; so many of the following factors are all comparably important.

Greater productivity, motivation (and energy)

On TCP, I have a literal urge to “make progress” and be productive. It’s almost as if tranylcypromine transforms me into a machine-like workaholic. On most days, I’m very motivated to work, enjoy doing so and feel cognitively alert and focused. This might sound normal to you but without tranylcypromine, I’m usually tired and unmotivated.¹

Greater productivity – evidence from Rescue Time

This is maybe the most informative section because it’s the least subjective. I started using RescueTime on January 16th, 2015 and logged my time actively (including offline time) until ~February 2016. However, RescueTime logs your activities automatically so even the data after this time are still informative.

With this in mind, here a very revealing summary in which I compare how many productive hours I logged per week on TCP and without TCP. (Note that I multiplied the “very productive” hours by two. In my view, this gives us more meaningful numbers because “very productive” activities include the most difficult yet most important tasks like e.g. doing research (google docs), studying, etc. whereas productive time consist of much easier and less important tasks like e.g. engaging occassionally rather unproductive discussions on Slack.)

Since January 2015, in the months during which I didn’t take Parnate (Jan – May 2015, Sep – Nov 2016, July – Sep 2017), I logged on average 23.2 productive hours per week (again, note that I multiplied “very productive” time by two so the actual time is considerably lower). In contrast, in the months during which I took Parnate, I logged on average 31.1 productive hours (transformed) per week.

In summary, Parnate increased my productivity (as measured per the above method) by 34% since 2015.

However, the data from 2016 onward might be even more meaningful for reasons outlined below. Since January 2016, I logged on average 16.3 productive hours (transformed) per week, when I was not on TCP. In contrast, in the same time, I logged on average 27.7 productive hours (transformed) per week, when I was on TCP.

In summary, tranylcypromine increased my productivity by 70% since 2016.

Detailed graphs

Below a few RescueTime screenshots summarizing my productivity from January to May 2015. I didn’t take TCP during those months.

January – May 2015

As you see, in January I was extremely unproductive and then got more and more productive until March until I became less productive again. I’d say that March 2015 was probably my most productive month since at least the beginning of 2012, potentially even earlier than that.

On tranylcypromine

Now it gets interesting. Sometime in June 2015, I started taking tranylcypromine. Note that during the first few weeks I had to increase the dosage and often had problems with my blood pressure. Additionally, I was taking a vacation, so in June I was still pretty unproductive. I think it’s pretty usual that MAOIs need several weeks until they really start working to their full effect, especially if you need to increase the dosage very slowly – I needed like 8 weeks to finally reach my optimal dosage of 30-35 mg.

June to September 2015

Notice the impressive data for July, and especially August and September. I’d say that every month I set up a new record for my all-time productive month ever.

August 2015

In August, I logged 164h very productive time. That’s like 30h more than in March 2015. Plus, I also logged more productive time than in March 2015.

September 2015

In September, I logged 276 hours of productive + very productive time. That’s like 9.2 hours per day on average, which means 64.4hours per week. Considerably more than even during the already great month before. (But note that I also log stuff like meditation, exercise, chores, etc. as productive time, so it’s really not that great.)

I remember that I was feeling extremely good during July, August and September 2015. I also slept little, so maybe it was all just a short-lived, unsustainable manic phase? To some extent. My mood got worse and my productivity declined substantially over the next 2-3 months because I couldn’t really deal with the pressure of having to do both EA-work and stuying for university simultaneously. So what happened during the next few months?

October 2015 to January 2016

Although my productivity during October, November and December 2015 and January 2016 was lower than during the summer of 2015, it was actually almost as high or even higher than in March 2015! And again: March 2015 was probably my most productive and happiest month since 2012. In January 2016, I mostly focused on writing my master thesis and logged 177h of very productive time, which was again a new record.

February 2016

In February 2016, I started to work full-time for EAF/FRI and my productivity and happiness started to decline because my stress levels increased.

March 2016 – September 2016

Unfortunately, I stopped using Rescue Time in March 2016.

Overall, I was less productive during 2016. But after analyzing the offline data, I concluded that I was basically on average about as productive as during March 2015, if not more so. Note however, that I was able to sustain this level of productivity for many months. I also took a long break from December 2014 to January 2015 which made the productivity of March 2015 possible in the first place. Maybe even more important is to keep in mind that I only had relatively easy university stuff to do in March 2015. I had much less stress and responsibility in March & April 2015 than during 2016 when I really started working full-time for EAF/FRI. I’m almost certain that I wouldn’t have lasted more than a few weeks at this job without tranylcypromine.

September 2016 until now

This was a pretty crazy time for a variety of reasons I can’t go into here. Anyways, if anything, this period gave me even more evidence for the benefits of tranylcypromine.

Greater optimism

Without tranylcypromine, I’m regularly afraid of and worry about the future. I often feel as though nothing makes sense and that all my ideas are not worth much. I experience profound uncertainty and doubt about everything – but more on an emotional than on an actual cognitive level – though the two are of course related and reinforce each other.

In contrast, on tranylcypromine, I tend to be more optimistic about the future which also gives me more motivation. Sometimes, I might even be overly optimistic. This optimism also goes along with having a more entrepreneurial spirit, of having a more “can do” attitude. Generally, on TCP, I regularly get really excited about my work, some book or some other idea. Such feelings of intense excitement usually don’t last forever (which is probably good!) but they often reoccur. Without TCP, I basically never have such feelings about anything – even if I were to stumble on the best idea ever.

Greater confidence

On TCP, I tend to believe more in myself and my ideas. Several people also told me that I seem to speak with more confidence and louder. I also remember being more vocal about voicing disagreements. Otoh, I fear that TCP could also sometimes make me overconfident, especially when I’m feeling very good.

More resilience, stress-resistance; easier to overcome obstacles

I think I can handle stress and setbacks better on TCP. On TCP, I find it easier to push through obstacles and be (much) less bothered by them.

Greater agency and “taking initiative”

On TCP, I’m more “agenty”. That is, it seems to increase my ability “to take initiative and cause a plan to succeed or an idea to get realized (when, by default, it would have withered and and died)”. There are many things that I can’t imagine having done without tranylcypromine because I repeatedly felt that I had to take responsibility, push through and motivate others.

Near hypomanic episodes

On tranylcypromine, I sometimes have (near) hypomanic episodes which may last for several hours though usually not longer.

During these episodes, I feel joy and happiness in intensities which I usually cannot experience in a normal state (I think). Naturally, this is awesome in and of itself. In addition, I’m much more creative during these episodes and can generate novel, out-of-the-box ideas. During those episodes, working on and pushing through extremely difficult and complex tasks is also possible.

Less “background” depression

I guess this is simply the reverse of what I have written above. Without tranylcypromine, I regularly feel a sort of constant background depression. I can feel depressed on TCP, but the feeling is less much less intense and profound. When I’m off tranylcypromine, I can experience episodes of near-catatonia where I find it almost impossible to even talk. During those episodes, my thinking also seems to be slowed down immensely.

Generally, I can have very depressive thoughts on tranylcypromine, but the emotional effect of those cognitions seems “blunted”. TCP acts like an emotional “shield”.

Without Parnate I can look happy on the outside, even feel mostly good, but often there still exists this background feeling of depression, anxiety and worry and it feels like I have to expand constant cognitive effort to prevent the depressive feelings from growing and eventually overwhelming me.

Less extreme depression

TCP also reduces feelings of existential despair. It’s hard to describe but I feel like I can experience more profound, “deep” emotions without TCP. Of course, this is a double-edged sword. One of the reasons why I’m scared of not taking tranylcypromine is that I’m afraid of the occasional bouts of extremely intense depression in which I feel existential despair, loneliness, self-hate, hopelessness, pessimism and anxiety all at once. It’s hard to describe what I mean with existential despair. It’s like being in a metaphysical nightmare in which you realize that the multiverse is an apathetic, cold and dark place full of agony, loneliness, delusion and broken dreams. Don’t get me wrong, even now, while on TCP, I believe that this is basically an accurate description of the world we’re living in but I can know this while simultaneously writing about it with a smile on my face.

Greater self-control

Just to give an example: I remember trying to stop playing video games for months while not taking tranylcypromine. Repeatedly. But I always slipped. All in all, I lost like 400€ on Beeminder – that’s how bad I tried to stop playing video games. But then I started tranylcypromine in June 2015 and a few weeks later (after the real effects started kicking in), I just quit playing video games for literally one year without having to invest more than a tiny bit of willpower.

Importantly, I also think that tranylcypromine makes it easier for me to not consume any drugs like alcohol, cigarettes (or nicotine gums) or weed. I certainly consume considerably less of those substances when I take tranylcypromine and have also much less of an urge to do so.

Disadvantages of tranylcypromine

Listed roughly in order of importance.

Insomnia

By far the most annoying and detrimental side-effect of tranylcypromine is severe insomnia. Symptoms include:

• Not falling asleep for several hours. Fortunately, I almost never had this side effect in the last two months or so.
• Waking up after 2-6 hours of sleep and then not being able to fall asleep again for at least one, more often two to four hours. This occurs on about 540% of the days, even after months of taking TCP. That’s very annoying, needless to say. I tried meditating to fall asleep (helps a bit but not really). The best solution is probaly to just get up and get some work down and fall asleep again. Another possible solution is to sleep for 5h every second day, i.e. get up if you wake up at 5 and immediately start your day (with taking TCP, caffeine, etc.). Hopefully, you’ll be so sleep-deprived that you’ll be able to sleep for 7-9 hours straight. Sometimes works.
• Generally, I only sleep like 7h on average on tranylcypromine. Usually, I sleep about 9h which seems optimal for my mood and brain.
• Trazodone can help somewhat, but not that much. The biggest problem is that it causes relatively strong hangover at doses where it’s really effective. However, another promising solution might be low-dose doxepin. Or maybe another tricyclic (though you’ve gotta be a bit careful with those).

I don’t really think that it’s possible to completely overcome insomnia caused by Parnate. Insomnia is really annoying because it can destroy your productivity for the next day. Good sleep is also necessary for having good health, memory consolidation, heightened mood, creativity, concentration, etc.

Of course, lying awake in bed, ruminating and just wanting to fall asleep again can also be extremely frustrating.

Afternoon fatigue, sudden “crashes”

This is a phenomenon that many people experience on Parnate. When you have a “crash”, your mood and alertness decrease severely during 30-60min. It’s much more intense than the usual sleepiness most people experience at some points during the day. One basically feels very tired and unmotivated. During the most extreme crashes, my pulse can go below 50 bpm and my blood pressure rise above 150/100. Uncomfortable feelings ensue.

Naps can help. Unfortunately, even naps often don’t end up restoring one’s mood and alertness completely back to baseline. One still feels very sleepy and unmotivated after the nap for like 2-5 hours. In the literature, this, or a related phenomenon, is also described as “afternoon fatigue”. Sometimes, the crash is less intense (or the crash is followed by a feeling of depressed mood and tiredness of lower intensity but longer duration) and you are just feeling irritable, tired and depressed for hours. Sometimes, especially in the beginning or when taking too large a dose of Parnate, or when not being able to lie down and nap, the symptoms can last for ~10 hours.

I think, however, I discovered a solution to this problem. First, I need to take many small dosages of TCP throughout the day. For example, 10mg and then 4 times 5mg every 1.5 – 2 hours later. This is quite uncommon I think, but every since I take my tranylcypromine this way, I have much less often crashes. The second solution is to maximally nap for 20 minutes! Overall, I’m much less concerned with crashes now and they are much more predictable than they have been in the past. I think the key is to experiment a lot. Experiment with total dosage, the size of morning dosage, the duration between dosages, etc. etc.

Unstable rhythm

In the past, it seemed impossible to have a stable rhythm. And TCP does indeed make it harder to have a stable rhythm or routine. But it’s not impossible. I think one just has to be very strict about going to sleep at the same time (making an alarm to take melatonin helps me). And to never get up too late. Better get up at 5, take your first TCP dosage and start your day (and maybe take several naps throughout the day if need be) than to lie awake in bed for 3-4 hours and then fall asleep again for 3 hours and then you got up at 11 or 12 and your rhythm is ruined.

(Hypo-)Mania

It’s not clear whether this is actually a disadvantage. But one side effect of hypo-mania is that I get more irrational, overconfident and insensitive.

Being less compassionate and “deep”

I think I become more of a “robot” on TCP. I’m a bit more superficial, less compassionate, less likely to feel profound feelings. Of course, as I’ve written above, being unable to feel deeply is often an advantage because it makes you more resilient, less likely to have existential angst or depression, etc. But I also value being a compassionate person with deep feelings, especially towards my friends and my girlfriend.

TCP dampens the effects of LSD

There are no dangerous, direct interactions between LSD and tranylcypromine but Parnate dampens the effects of LSD. So, say, 300 mcg might feel like 50-100 mcg. ²

However, one can never be certain that it’s really pure LSD on one’s blotter rather than, say, 25C-NBOMe, which may be dangerous to combine with TCP.

Weight gain

I suspect that I weigh about 3kg more on tranylcypromine. Not sure why. Maybe increased appetite. On the other hand, I think most people actually don’t gain weight so that’s not a common side effect at all.

Other disadvantages

• The possibility of having a hypertensive crisis. I sometimes had high blood pressure (e.g. 160 systolic), though not high enough to classify as a hypertensive crisis. From extrapolating, a hypertensive crisis will be extremely unpleasant. That said, keep in mind that getting a hypertensive crisis is very unlikely and even if you have one: It’s very unlikely you will die from it. I’d guess that the probability of suicide because you don’t treat your depression – or take ineffective antidepressants – dwarfs the probability of dying of a Parnate-induced hypertensive crisis by at least one order of magnitude.
• Worrying about interactions with other drugs or pharmaceuticals.
• Not being able to eat everything.
• Having to go to the psychiatrist every 2 months.

Dec. 2017 Update: Thoughts about TCP – excerpts from an email conversation

In December 2017, a reader wrote to me (some details edited out to make it anonymous):

[When I started Parnate,] I was transformed into a highly motivated, ultra productive person – exactly what I wanted. I finally became who I wanted to be. Somewhat corresponding to your timeline, after 3 months […] I felt it suddenly decline in effectiveness (without any external life changes). I stopped it, and tried different methods to reverse this tolerance, without a clear answer. I’m now going to restart it again.

[..] I know you explained your decline in productivity post Oct 2015 as due to life circumstances and work situation, but do you think there’s any chance this has to do with declining efficacy and drug tolerance? Or are you still confident that TCP works well even after 2+ yrs of use? How do you know?

Also…why did you stop for 3 months in 2016 and 3 months in 2017? Did you feel very amotivated/down off of it?

I’m thinking I should restart it because it’s the only thing that’s worked for me in terms of motivation/productivity, and accept the fact that it may never recapture the initial honeymoon of the first 3 months, but be content with the improvement nonetheless…

Below some excerpts from my response:

thank you for your kind words!

Interesting that your experiences seem so similar to mine.

[Beware: Rambling ensues.]

Yeah, my current thinking is that tolerance/poop-out probably had something to do with the declining effectiveness of tranylcypromine over the last two years. However, note that in January 2016, I was about as happy and productive as in the first few “golden” months on Parnate, so tolerance/poop-out cannot be the complete story. There is one thing that the three summer months of 2015 and the month of January 2016 have in common: I was not overly stressed out and didn’t need to “juggle too many balls”. In the summer months of 2015 I could only focus on my work and in January 2016, I could only focus on university.

However, later in 2016, I finished my degree and although I now could solely focus on my work, TCP still seems to have lost some of its magic. However, one also has to note that I gradually assumed more and more responsibility and difficult tasks over 2015 and 2016, so that the work I did in the summer months of 2015 was a lot easier than late in 2016. (Otoh, my skills also increased.)

I stopped for 3 months in 2016 and for 3 months in 2017 because I never was able to overcome the insomnia, and, over time, it seems to have taken some physical toll on me. I could cope less and less with the lack of sleep. I was also worrying more about the detrimental health effects of the lack of sleep. (However, note that I still slept about 7 hours per day (including naps), which is not too bad but about 1.5 hours of how much my “body needs” for optimal functioning, I’d guess.) I can easily imagine that the accumulating sleep debt contributed to the loss of effectiveness of Parnate.

One more comment: Since the Big Bad Thingy happened in late 2016, the hours I lie awake in bed are filled with obsessing over the past, regretting my mistakes and shortcomings, feelings of having been betrayed, mourning the foulness of human nature, etc. Whereas in 2015 and early 2016, whenever I could not fall asleep, I was much more often thinking about pleasant or productive things (e.g. related to work).

But yeah, tranylcypromine in 2015 might have been strong enough to overcome all these bad things. But note that e.g. from March to April 2017 I tended to feel very good, so tranylcypromine still seems to have worked for me. I was also very productive from about February to July 2017. After I stopped taking Parnate in July 2017, I’ve become waaaay less productive. I tried starting it again in late October 2017, but stopped taking it a few days ago, mostly because I simply could not cope with the insomnia anymore. I’ve also become more worried about potential detrimental effects on my health, especially my cardiovascular health, due to the lack of sleep and potentially tranylcypromine itself. However, I could not find any substantial evidence, not even individual anecdotes, that Parnate has detrimental long-term effects on one’s cardiovascular health (of course, aside from hypertensive crisises due to consuming tyramine which is completely preventable.)

Hope that gives you some useful information.

Another quick observation: On the internet, I found many people for whom Parnate seemed to work over a decade only to then suddenly lose some of its effectiveness. On a neurochemical level, this seems very confusing to me. I mean for literally all other drugs (amphetamine, opioids, benzodiazepines, etc.), tolerance and loss of positive effects occur after months, or, in most cases, even *weeks* of usage. Not after years or decades…

Detour: Price gouging

This is not very important, but very annoying. Here some quotes from Ken Gillman first:

Many people will be aware that out-of-patent drugs, that should be cheap (like 10 – 20c per tab) are now frequently being sold at prices way above (like 100 times higher) what can be justified by decent business or ethical considerations. I will not dwell on it here, just Google the phrase ‘outrageous price increase drug’.

[…]

The two drugs that I am discussing here provide a number of revealing contrasts from all possible points of view, the first being in the price-hike field. The price of Parnate (tranylcypromine) has for many years been very modest since it has been out-of-patent for ages. But more recently the price has been much increased in most jurisdictions. The UK National Health Service are currently paying £248.14 for 28 x 10 mg tablets (around 700c per tab.), putting the cost of a typical month’s treatment at a couple of thousand US dollars or more. Insane, obscene.

The other drug, mirtazapine, remains dirt cheap, £1.80 for 30 x 30 mg tablets, around 10c a tab, which is all it is worth. That is about 500 times cheaper! Perhaps when they learn whatever tricks the ‘Parnate people’ seem to be getting away with, it will rocket up in price also! […]

A similar phenomenon (though less extreme) exists in Germany. There are two producers of tranylcypromine in Germany: Aristo (brand name: Jatrosom) and Neuraxpharm (generic). 100 tablets a 20 mg of Jatrosom cost 227,91 €. 2,27€ per tablet. Pretty expensive for a drug that’s over 50 years old and probably costs a few cents to produce (per tablet). You might think that the generic must be cheaper. And you’d be right. 100 tablets of 20mg generic tranylcypromine cost… 227,87€. A price difference of 0.04 cents per tablet! Amazing. Of course, the two companies Aristo and Neuraxpharm have to have some kind of deal and keep prices artificially high. This is usually illegal but I don’t know if Big Pharma was able to somehow buy some nice regulations in their favor. It’s also not as if anyone cares since there are so few people who take tranylcypromine.

Footnotes

1. I remember, however, that in 2011 I was reading LessWrong like a machine for 11 hours per day on average for months without needing many breaks or seeing friends (not that I really had any back then). So it’s possible for me to have lots of motivation even without tranylcypromine. On the other hand, reading is also very easy and of course, it’s to be expected that someone like me would get extremely excited when they discover LessWrong and the whole memeplex of rationality, AI, superintelligence, transhumanism, Bayesianism, decision theory, etc. Anyway, even back then I had great trouble when it came to actually “doing” stuff in the real world and not just reading.
2. A reader comments that this is because one’s HT2A/2C and HT7 receptors are downregulated from the TCP causing higher endogenous serotonin concentrations.