This commentary offers practical advice for starting treatment, and adjusting the dose to get a timely response, with irreversible MAOIs (Parnate™, Nardil™, Marplan™ ). It brings together information from my published scientific papers (1-7) and from other previous commentaries on this website (see MAOI sub-menu under ‘Antidepressants’). I am an internationally recognised expert on MAOIs and their interactions.

Whilst preparing to start TCP it is helpful to be confident about what can be used to help while waiting, and what can be given in combination. Contrary to what many doctors have been led to think, it is actually pretty simple — anything, except an SRI. But the details are discussed here

Foods now have lower tyramine levels than in the past, so the diet is quite easy to follow and for many patients involves few, if any, changes. Exhaustive details and discussion about that are in the MAOI sub-menu: the downloadable monograph is the definitive source of data published.


It is preferable to use proper drugs names, not ‘trade’ names — one should know what you are using/taking, and navigating the scientific literature requires using the proper chemical name in order to find information.

So, Parnate™ is tranylcypromine TCP; Nardil is phenelzine PLZ; Marplan is isocarboxazid(e) ISO***

***Alternative spellings are common: in this case there are about equal numbers of references retrieved with each alternative of ISO.

NB. In the text I will just use ‘TCP’ but, unless specifically stated, comments apply to all irreversible MAOIs. Which of these drugs are, or are not, available in which countries is variable, and without any logic whatsoever.

The general principles of clinical pharmacology, in relation to the initiation and adjustment of drug treatment, are not well observed. I have commented previously on how important it is for doctors to clearly document at every visit the exact current treatment and dose duration. Unless that information is readily to hand, along with the degree of improvement at each stage, it is not possible to make logical decisions about future treatment adjustments (there is further discussion on this, and similar issues, in my AD algorithm).

These comments apply particularly to TCP, but many of them can be generalised to non-MAOI drugs. However, the important difference of TCP lies in its unique mechanism of action — as an irreversible inhibitor of MAO — which is different to almost all other drugs, most of which act reversibly at various different sites, receptors and enzymes etc. Once a molecule of an irreversible MAOI drug has interacted with the MAO enzyme, that molecule of MAO is permanently disabled. Enzyme function is then restored only by the production of new enzyme from the genetic code, via synthesis of new proteins that make up the structure of the enzyme.

PET studies indicate that human MAO is replaced at the rate of approximately 3% of the total per day (more slowly with advancing age), so it takes a while for full function to be regained (like 30-40 days (8)), and if 80-85% blockade is a threshold level for full therapeutic effectiveness, as is generally posited, then full effectiveness would be expected to last 3-7 days after the last dose, if full 100% blockade of MAO had previously be achieved (which also indicates that twice weekly dosing should be sufficient — see below). It has also been shown that ~ 40% MAO inhibition is required to reduce MTTP toxicity (9) which indicates some significant changes are probably taking place at a dose of ~ 10 - 20 mg daily, even if 85% blockade is required for the maximal effect. That data coincides closely with my clinical experience.

Starting dose: BP monitoring

TCP has been shown to inhibit >50% of human brain MAO-A after a three-day treatment with 10 mg/day (10). Clinical experience indicates that a small proportion of patients will get effects from that dose (I do not mean optimal or maximal effects, just noticeable ones), even sometimes significant postural hypotension — of the degree depicted in the bar graph in the">MAOI_BP pdf. Illustrating BP changes at days 4 and 11.

It is generally supposed that approximately 85% inhibition is required for full clinical effect to occur.

Using the sound pharmacological principle of ‘start low, go slow’ it is clearly prudent to start with no more than 20 mg per day, to assess the effect for 5 days by monitoring sitting and standing BP to ensure there is not an excessive postural drop at that dose — but note, near-maximum BP drop, on a constant dose, will take 10 days, or more, to develop. This is explained in the downloadable MAOI_BP pdf document here

NB I say sitting and standing BP, rather than lying and standing BP, because most of these assessments will be taking place in the ‘office’ not the hospital. It is therefore easiest to standardise all measurements to sitting and standing.

IMCO, BP monitoring is an important procedure to follow because there is a small risk that sudden unexpected postural hypotension might lead to injury – I once had a patient who got out of a hot bath and stood at the top of the staircase listening, because they thought the telephone was ringing (in a world long before mobile phones, oh young ones, when the only phone in the house was in the downstairs hallway!). Fortunately, they did not tumble down the stairs when they fell to the floor. Hot days, hot baths (or here in tropical Queensland, warm swimming pools — eat your heart out if you are in the N hemisphere!), dehydration etc. are what doctors need to remind people to be careful about.

Educating patients (and doctors) about the practical aspects of postural hypotension is wise and one particular paper is especially useful

Sadly, many practitioners do not understand the above, and as soon as a patient feels faint, or falls, they react by ceasing treatment — a completely unnecessary mistake. In this day and age it seems almost mandatory for patients to have their own (upper-arm) BP monitor, and to be schooled in is proper use to monitor their own BP at home and to keep a written record of BP results (we are planning to develop an app for devices to display these results graphically, as in my pdf (anyone who can help with this should This email address is being protected from spambots. You need JavaScript enabled to view it.">contact me).

It is important to appreciate that this hypotensive effect is near-maximal during the 10-14 days after reaching the ‘BP-effective’ dose — that is often around 30 mg per day, and then, if the dose is kept the same, diminishes progressively after that, as the receptors adjust their sensitivity (see pdf file mentioned above re BP). At that point, the dose may be safely increased.

By ‘BP-effective’ dose, I mean the dose that will induce a significant postural BP drop of ~ 10-15 mmHg, as illustrated in the bar graph for days 11 and 14.

Most people on long-term MAOI treatment experience a useful drop in BP and anyone previously on treatment for hypertension will need to have that monitored and reduced, sometimes it will need to be ceased completely. If MAOI treatment is ceased, remember to monitor BP and re-assess need for further anti-hypertensive treatment.

These BP changes require that the measurements have been done in a consistent manner, after being seated at rest for 5 minutes, and that a few pre-treatment measures are done as a ‘baseline’. The pattern shown in the bar graph shows why it is vital that two standing measurements are done. All the published papers on MAOIs and BP are of very limited value because they do not provide data on the relevant dose-change-BP timing, nor two successive standing measurements: I have reviewed">this subject here where I commented ‘these are drugs which were used as anti-hypertensives over 50 years ago! and even now psychiatrists are congratulating themselves for discovering hypotension as a ‘side effect’!’ When used like that, the term ‘side effect’ is devoid of meaning — it is not a ‘side effect’, it is a predictable therapeutic effect. And people wonder why I make disparaging remarks about the knowledge exhibited by psychiatrist.

For patients who are not in hospital I would usually wait a minimum of ten days for the dose to manifest its full effect before any further increase.

So, starting with 20 mg and increasing to 30 mg after 10 days would be the most frequent pattern — however, if after 5 days on 30 mg, the BP drop was less then 10 mmHg, then I would probably proceed to 40 mg, but then allow a full 14 days on 40 mg to judge the effect. If BP drop on 30 mg was more than 15 mmHg then I would wait the full 10-14 days to assess the maximal degree of postural hypotension at that dose. If that level of postural hypotension was still present at 14 days I would wait a full 4 weeks on that dose to assess the degree of illness improvement.

Experience in using these drugs will allow doctors to fine-tune these suggestions to their own practice-population and circumstances, and during this phase education and confidence building about BP and diet is timely.

A proportion of patients will get a good response to 30 mg, so progressing hastily beyond that dose, without fully assessing response, means there is a chance of using an excessive dose for that patient: that is not good clinical pharmacology practice.

My">AD algorithm stresses the importance of critical assessment of the degree of improvement at each ‘stage’, particularly improvement in ‘core’ symptoms of anergia and anhedonia as reflected in work, social and leisure activities. If there is a good BP drop, and these indices are showing significant improvement within 4 weeks I would wait as much as 3 months to assess the full benefit. If such a degree of improvement is not evident, or there is no drop in BP, I would press on, in similar stages, to a dose of 60 mg daily.

Side effects and problems

A majority of patients who have had multiple different AD drugs rate MAOIs as having less side effects than most, or all, of their previous treatments — not many doctors know that (11, 12).

It is relevant to mention that patients who experience even profound hypotension, sufficient to prevent them standing at all, will usually develop tolerance given time — and it may take 3-4 weeks or more, not just ten days. I have seen quite a number of patients who were unable to stand for significant durations without fainting, who persisted with treatment, and went on to take and benefit from TCP for years without further trouble.

The first simple essential intervention re low BP is to ensure adequate salt intake – this may mean 5 – 10 grams per day (two heaped tea-spoons of sea salt — which has a higher magnesium content), especially in those whose normal diet is low in salt.

In my experience fludrocortisone is not usually required, and if used can often be taped-off within a couple of months. Midodrine is not usually helpful and may have an unpredictable and fluctuating effect.

So, the mantra ‘slow and steady wins the race’ has never been truer.

Insomnia and its treatment

Incidentally, the same long adaption period applies to severe insomnia: this often resolves completely with time — sometimes three months, or even more.

The lady who had a letter published in CNS Spectrums, in response to my 2017 editorial in that journal (6), said she had severe insomnia for around three months. Interestingly, she noted that when she was awake in the night her thoughts were of pleasant past achievements and events. Indeed, her doctor wondered if she was manic. Her hypotension also lasted at least three months.

My comment is that those who are not familiar with the dramatic improvements one sees with TCP might well suspect mania, because they have so little experience of seeing ‘proper cures’. We used to see such rapid and dramatic improvements sufficiently frequently that one of the senior nursing staff in the hospital (the best colleague I ever had the privilege of working with) came up with the flippant line ‘the impossible we do today, miracles take a little longer’. When one has seen severe psychotic depression resolve in 7 days, one appreciates that comment is not without substance.

In the interim, trazodone seems to work well for insomnia and nortriptyline can be useful too (both are potent 5-HT2A antagonists), like-wise doxepin (or quetiapine — if you must, and do not care about the cost to the health-care budget). The fact that most texts, and most psychiatrists, think those drugs cannot be combined safely with TCP is simply a testimony to lack of knowledge and experience.

Melatonin helps some.

The BDZ-type drugs seem less effective, but if used, the medium half-life drugs like lorazepam seem preferable. Long half-life drugs like clonazepam and diazepam are not ideal — the levels can build up over a couple of weeks or more; alprazolam and ultra-short half-life ‘Z’ drugs are less effective. Half-lives of Zs are approximately — zaleplon 1 hour, zolpidem 3 hours, zopiclone 5 hours.

Lastly, L-tryptophan (the amino-acid precursor of serotonin) should not be forgotten. Gradually escalating doses starting with 1-2 grams (not milligrams) at night can be effective. We used to use as much as 4-8 grams, but at such higher doses one sometimes sees hyperreflexia and clonus. There are rare reports of toxicity and confusion (13-17). Some of the cases of confusion reported in association with L-T exemplify a wanton disregard of the principles of good clinical pharmacology mentioned above: e.g. in one of the cases described: dothiepin 300 mg daily ceased abruptly and PLZ started next day, and increased the day after that! and with addition of tryptophan. Well, if you go looking for trouble you will find it.

Change only one drug at a time, and do not suddenly cease drugs known to produce withdrawal effects! Allow at least five half-lives to elapse before assessing the result: you just cannot help some people who are wilfully stupid and disobey the basic rules of sensible clinical pharmacology.

As far as tryptophan is concerned it is safe if used sensibly — even in large doses it will not cause serious ST (18): it will induce predictable hyperreflexia and clonus with higher doses, so just back off a little if that happens – no need to panic, these serotonin-mediated effects are dose-responsive and wear off quickly (a few hours).

My observation is that the longer the duration of the illness, and the more severe it is, the greater is the degree of sleep disturbance & ‘REM deficit’ and the longer it takes for the sleep pattern to normalise as recovery proceeds. As far as REM sleep is concerned, the brain seems to be like a money-lender — you can accumulate a big debt, but you have to pay it back eventually. EEG monitoring is an area where further research would be useful.

Adjusting dose: ongoing BP monitoring

Once there is a useful, but not problematic, BP drop on standing (see day 14/18 on graph) it is ideal to stick with that dose for a full 4 weeks to critically assess the speed and degree of improvement, unless the severity of the illness, and/or the life situation, mandate a more aggressive approach.

In this manner, if no substantive response was manifest, I would proceed to 60 mg daily and if there was still no substantive response, would re-assess the case or look for a second opinion, before proceeding with higher doses. In my experience, most patients who are going to respond to TCP do so in the dose range of 30 – 60 mg per day. If only a partial response is evident, at the maximum tolerated dose, one would then consider augmentation with nortriptyline or tryptophan or lithium, or even all three together (for an example of cases histories like this see the article by my old boss in London, Paul Bridges (19)). This illustrates well the type of case I was ‘brought up on’ in our tertiary referral unit. We did psychosurgery too — actually a modified prefrontal tractotomy using radio-active Yttrium seeds — and it produced some spectacular results (but, as one wag said, ‘I’d rather have a full-bottle-in-front-of-me than a full frontal lobotomy’).

It is often satisfactory to time the doses about 4 hours apart, starting with the largest portion of the dose first thing in the morning. If somebody was taking 50 mg one might often give 30 mg as the first dose, and 20 mg between two and four hours later. Doses taken later in the day do seem to be associated with greater problems with insomnia. However, I have had patients who have taken the whole dose at night and claimed to sleep well, so one cannot be too dogmatic. Sometimes taking four equal doses spaced at two-hourly intervals during the morning suits some patients better. Some people find that taking the whole dose first thing in the morning lessens insomnia. The bottom line is you have to be your own ‘guinea-pig’.

Paradoxical hypertension

It is invaluable to have an ongoing record of blood pressure as described above. This is especially so when episodes of paradoxical hypertension occur, which is sometimes seen (in approximately 5% of cases, in my experience). This paradoxical hypertension (not related to tyramine ingestion) characteristically commences an hour or so after the second dose of the day, especially when the dose is a little higher, and lasts for a couple of hours or so before subsiding spontaneously. Although such episodes do cause concern in some people, especially doctors, they are rarely, if ever, sufficiently high to be problematic. In my experience, they rarely go over 180 to 200 mmHg, and then only briefly (see my comment on this here (20)). Again, this phenomenon often dwindles away over time. If treatment is thought to be advisable a beta-blocker like propranolol is appropriate, although it should be noted that propranolol is not suitable for severe hypertension secondary to tyramine ingestion.

And also

This is an appropriate point at which to note a couple of interesting things. I well recall that in many of the cases referred to us in the London unit, which usually came from other teaching hospitals, it would be said that a full course of various kinds of treatment had been given. It was of course my job to check all these things, and when I went through the notes of the referring hospital I would quite often find that such statements were inaccurate and that the treatment had been prematurely truncated, or not given in the dose stated. Many referrers were convinced that if their favourite treatment did not work, there was no point in continuing their efforts. This was especially so when the patients had failed to respond to ECT. Many doctors would effectively give up on treatment after that. We sometimes use to proceed to psychosurgery — even though people had not had a full course of an MAOI, or other suitable medication combinations — simply because we knew full well that if we sent them back to the referrer with instructions to institute such a regime, that it simply would not happen. This was the genesis of my habit of always recording the total duration of treatment, and duration of treatment on the maximum tolerated dose. If this information is not readily available it is inevitable that major mistakes in the treatment decisions will occur.

MAOIs frequently work when ECT has failed. It is unfortunate that so few doctors appear to understand that.

Indeed, the enquiries I have had through the website so-far this year (six months) have included half a dozen such patients, a couple of them of them being medical practitioners.

Other considerations

PLZ & ISO ‘dose equivalence’

TCP 20 is approximately PLZ 30, usual dose range for PLZ 45 up to 120 mg

Isocarboxazid seems a little more potent than PLZ (they are both hydrazines), dose range being 20 – 60 mg or thereabouts. ISO may have less of a sedative effect

The only recent review of ISO I know of is Larsen’s (21).


The hydrazine MAOIs (PLZ and ISO, but not TCP) inevitably reduce pyridoxine and it is probably wise to do a pre-treatment pyridoxine level and routine follow-up levels after 1, 6 and 12 months. It seems logical to give pre-emptive B6 supplementation at 100 – 200 mg per day (22-25).


It has now been convincingly demonstrated using PET that sublingual absorption of selegiline (aka deprenyl) inhibits brain MAO-A (26), as well as MAO-B — selegiline is only slightly selective for B over A, so quite a small increase in ‘normal’ levels will produce MAO-A inhibition. It would be expected that TCP would like-wise be absorbed sublingually, perhaps thereby reducing some side effects.

In a unique paper by my old friend (I lectured in Toulouse many years ago) Jean-Michel (27), it was neatly demonstrated that selegiline works well given twice weekly — TCP would be expected to be the same: sadly, no-one has tested that very important hypothesis — a testimony to the paucity of perspicacity in research.


1.         Gillman, PK, A systematic review of the serotonergic effects of mirtazapine: implications for its dual action status. Hum Psychopharmacol, 2006. 21(2): p. 117-25.

2.         Gillman, PK, Tricyclic antidepressant pharmacology and therapeutic drug interactions updated. Br J Pharmacol, 2007. 151(6): p. 737-48.

3.         Gillman, PK, Triptans, Serotonin Agonists, and Serotonin Syndrome (Serotonin Toxicity): A Review. Headache, 2009. 50(2): p. 264-272.

4.         Gillman, PK, CNS toxicity involving methylene blue: the exemplar for understanding and predicting drug interactions that precipitate serotonin toxicity. J Psychopharmacol (Oxf), 2011. 25(3): p. 429-3.

5.         Gillman, PK, Advances pertaining to the pharmacology and interactions of irreversible nonselective monoamine oxidase inhibitors. J Clin Psychopharmacol, 2011. 31(1): p. 66-74.

6.         Gillman, PK, “Much ado about nothing”: Monoamine oxidase inhibitors, drug interactions and dietary tyramine. CNS Spectr, 2017: p.

7.         Gillman, PK, Monoamine oxidase inhibitors: A review concerning dietary tyramine and drug interactions. PsychoTropical Commentaries, 2017. 17(1): p. 1-104 Available at:

Accessed May 2017.

8.         Fowler, JS, Volkow, ND, Logan, J, Wang, GJ, et al., Slow recovery of human brain MAO B after L-deprenyl (Selegeline) withdrawal. Synapse, 1994. 18(2): p. 86-93.

9.         Jossan, S, Sakurai, E, and Oreland, L, Is the monoamine oxidase-B activity rate limiting for mptp neurotoxicity. Biogenic Amines, 1987. 4(4-6): p. 371-379.

10.       Fowler, JS, Volkow, ND, Wang, GJ, Pappas, N, et al., Inhibition of monoamine oxidase B in the brains of smokers. Nature, 1996. 379(6567): p. 733-6.

11.       Parker, G, Mitchell, P, Wilhelm, K, Menkes, D, et al., Are the newer antidepressant drugs as effective as established physical treatments? Results from an Australasian clinical panel review. Aust NZ J Psychiatry, 1999. 33(6): p. 874-81.

12.       Parker, G, Roy, K, Wilhelm, K, and Mitchell, P, Assessing the comparative effectiveness of antidepressant therapies: a prospective clinical practice study. J Clin Psychiatry, 2001. 62(2): p. 117-25.

13.       Levy, AB, Bucher, P, and Votolato, N, Myoclonus, hyperreflexia and diaphoresis in patients on phenelzine-tryptophan combination treatment. Can. J. Psychiatry., 1985. 30(6): p. 434-6.

14.       Pare, CMB, Potentiation of monoamine oxidase inhibitors by tryptophan. Lancet, 1963. 2([letter]): p. 527-528.

15.       Glassman, AH and Platman, SR, Potentiation of a monoamine oxidase inhibitor by tryptophan. J. Psychiatr. Res., 1969. 7: p. 83-88.

16.       Baloh, RW, Dietz, J, and Spooner, JW, Myoclonus and ocular oscillations induced by L-tryptophan. Ann. Neurol., 1982. 11: p. 95-97.

17.       Pope, HG, Jonas, JM, Hudson, JI, and Kafka, MP, Toxic reactions to the combination of monoamine oxidase inhibitors and tryptophan. Am J Psychiatry, 1985. 142: p. 491-492.

18.       Gillman, PK, A review of serotonin toxicity data: implications for the mechanisms of antidepressant drug action. Biol Psychiatry, 2006. 59(11): p. 1046-51.

19.       Hale, AS, Procter, AW, and Bridges, PK, Clomipramine, tryptophan and lithium in combination for resistant endogenous depression: seven case studies. Br J Psychiatry, 1987. 151: p. 213-7.

20.       Gillman, PK, Ti: myths about monoamine oxidase inhibitors perpetuated. Aust NZ J Psychiatry, 2009. 43(11): p. 1084-5.

21.       Larsen, JK, Krogh-Nielsen, L, and Brøsen, K, The Monoamine Oxidase Inhibitor Isocarboxazid is a Relevant Treatment Option in Treatment-Resistant Depression-Experience-Based Strategies in Danish Psychiatry. Health Care: Current Reviews, 2016: p. 1-4.

22.       Larsen, JK, Bendsen, BB, and Bech, P, Vitamin B6 treatment of oedema induced by mirtazapine and isocarboxazid. Acta Psychiatr. Scand., 2011. 124(1): p. 76-7; discussion 77.

23.       Visser, ME, Texeira-Swiegelaar, C, and Maartens, G, The short-term effects of anti-tuberculosis therapy on plasma pyridoxine levels in patients with pulmonary tuberculosis. Int. J. Tuberc. Lung Dis., 2004. 8(2): p. 260-2.

24.       Steichen, O, Martinez-Almoyna, L, and De Broucker, T, [Isoniazid induced neuropathy: consider prevention]. Rev Mal Respir, 2006. 23(2 Pt 1): p. 157-60.

25.       Heller, CA and Friedman, PA, Pyridoxine deficiency and peripheral neuropathy associated with long-term phenelzine therapy. Am J Med, 1983. 75(5): p. 887-8.

26.       Fowler, JS, Logan, J, Volkow, ND, Shumay, E, et al., Evidence that formulations of the selective MAO-B inhibitor, selegiline, which bypass first-pass metabolism, also inhibit MAO-A in the human brain. Neuropsychopharmacology, 2015. 40(3): p. 650-7.

27.       Andreu, N, Damase-Michel, C, Senard, JM, Rascol, O, et al., A dose-ranging study of selegiline in patients with Parkinson's disease: effect of platelet monoamine oxidase activity. Mov. Disord., 1997. 12(3): p. 293-6.