Overview: MAOI and TCA interactions

by | Last updated Jan 31, 2021 | Published on Jun 29, 2019 | Anti-Depressants, MAOIs

This is a brief outline of the safety of combinations of MAOIs and TCAs to assist those who are not already familiar with these issues to grasp the overall picture. Contrary to most of the (mis)information in standard texts there is an enormous amount of published evidence demonstrating such combinations are perfectly safe. How the contrary view got into the literature is an object lesson in misinformation, overreliance on poor quality case reports, biasing influences, and poor scholarship, which I have discussed elsewhere.

Below is a footnote about the lengthy report submitted to the American College of Neuropsychopharmacology by White and Simpson in 1980. Although this is now only of historical interest, it is a useful source of references. Its timing was unlucky, because it was just after 1980 that much of the data that forms the substance of my papers, explaining the interaction as a manifestation of serotonin toxicity, became known.

The list of references below is almost certainly the most comprehensive such list yet to be compiled. If anyone is aware of significant other references that I had missed please inform me, and I will add them — I daresay there is mention of such things in various older textbooks to which I do not have access; although some of that information is of limited value, it is worth documenting it for the record. I have been unable to source a copy of the 1964 UK ‘Witts Committee’ report.

We must remember that what is published in the scientific literature is but the tip of the iceberg in terms of the totality of clinical practice experience. What is ‘published’ can give a misleading impression of clinical reality. RCTs especially can be unrepresentative and misleading: we are unlikely to advance the treatment of depression by fiddling with statistics. Many eminent writers, including Ronald Kuhn, Sargant, and Bradford Hill, have made such observations.

Also, the categorisation used in relation to these drugs is changing. I have repeatedly discussed elsewhere the move towards neuroscience-based classification of drugs. The group of drugs referred to as tricyclic antidepressants (TCAs) which were brought into use in the early 1960s were grouped together on the basis that they all had a three-ring central structure. However more recent pharmacological advances in research have made it clear that their pharmacological properties and actions differ substantially from one another [1]. Pharmacologically, they are a heterogenous group. Statements about these drugs that start ‘the TCAs’ are almost certain to be wrong. Insufficient knowledge of their pharmacology was the root of the earlier misunderstanding, as exemplified in the report of White and Simpson below.

TCAs are structurally related to the first-generation antihistamine drugs, which are all ‘tricyclics’, and also to the first generation of neuroleptic drugs, represented by chlorpromazine. Indeed, all of those drugs share many properties, especially, for instance, their potency as sedative antihistamines (H1 antagonists).

Some of the drugs that were marketed as antihistamines in the 1960s would now be classified as antidepressants, and several of the ‘antidepressants’ would now be classified as specific and potent antihistamines (for example trimipramine and doxepin — indeed doxepin has recently been re-marketed as a sleeping tablet, as suggested in my TCA review paper some years ago [1]). Both trimipramine and doxepin are still used, in my opinion misguidedly, as antidepressants (in some countries). A full appreciation of this new knowledge has still not percolated widely into standard texts, nor everyday clinical practice, nor into the product information approved by the FDA.

The relevance of this is that the past reports of dangerous interactions between MAOIs and ‘TCAs’ were not informed by an understanding of their neuropharmacology and they confused different adverse effects and toxicities [2].

Indeed, almost all the past reports were poor quality case reports of single cases, usually with incomplete data and poorly informed discussion — I have previously discussed at length the frequently misleading nature of case reports [3-5]. Such reports produce a great deal of misinformation and wasted clinical time and effort, especially for less experienced practitioners. Better quality data has emerged from prospective monitoring of toxicity in toxicology databases, foremost among such sources is the seminal data collected by Professor Whyte and colleagues, from Newcastle [6-13]. Unfortunately, such data are not well-known or recognised in the psychiatric literature.

We now understand that the only serious interaction between MAOIs and TCAs is caused by excessive elevations of serotonin. Such great elevations of serotonin (5-HT) can only be produced if MAOIs are combined with therapeutic doses of potent serotonin reuptake inhibitors (SRIs). There are a few other drugs that possess SRI properties to the extent that they are occasionally able to reach high enough levels to increase serotonin substantially [14], and therefore become risky (e.g. meperidine, tramadol, chlorphenamine).

Of the TCAs the only ones that possess significant SRI potency are clomipramine and imipramine: however, imipramine is weak and only occasionally causes a severe ST interaction. Indeed, interactions with imipramine are sufficiently uncommon and mild that some practitioners (incorrectly) consider the combination to be safe (see White and Simpson and Gillman [2] for more information).

All the other tricyclics are so much weaker for this SRI effect that there are no reports of serotonin toxicity when these have been combined with MAOIs, even in overdoses (and this agrees with experiments in animals, and much other data).

The references about MAOI & TCA combinations provided below refer to a considerable number of publications describing thousands of patients treated safely with such combinations. In addition to that, many experienced psycho-pharmacologists all over the world use such combinations and would state that they are safe. Certainly, my own first-hand experience of such treatments encompasses several hundred cases without significant adverse effects.

The property of these old TCAs that is beneficial, potentially to an extent which will reduce, or even obviate, the need to follow a tyramine restricted diet, is their ability to block noradrenaline reuptake (NRI).

Ironically, not only is the risk of a tyramine pressor response reduced, but also there is a definite suggestion that some combinations of an MAOI and a TCA can have less ‘day-to-day’ side effects overall than either drug given separately. That is not as strange as it might at first seem, because sometimes one drug can mitigate the side-effects of another — that is what may be occurring with some TCA/MAOI combinations.

The contraindications given in the product information approved by the FDA (often long ago) relate to the earlier period of knowledge referred to above, when the pharmacological properties of these drugs were not precisely known. Why general texts and the PIs have not caught up with modern knowledge is not something I will expound on here. However, it is clear that drug-regulatory agencies, e.g. the FDA and the EMA, deal with drugs on an individual basis and do not necessarily review the overall developments in pharmacology and clinical toxicology. I suspect the explanation is quite simple, it is that dealing with drugs on an individual basis, and not reviewing the ‘big picture’, means that they ‘cannot see the forest for the trees’. That, and the over-reliance on unreliable case reports, may also explain why the warnings they have been giving about ST with the antiemetic drugs (like ondansetron), and the anti-migraine drugs like sumatriptan, have also been criticised. Indeed, I think they are generally ignored.

As a world expert in ST I am well-placed to state that our theoretical knowledge about ST, derived from pharmacology and animal experimentation, is now both thorough and well-established. It allows accurate predictions of which drugs are potentially a risk from this interaction (e.g. see [15]). This knowledge coincides well with an informed interpretation of all the clinical data from trials, and case reports. It is therefore possible to be extremely confident that TCAs (other than clomipramine and imipramine) have no significant effect on elevating serotonin, and no risk of precipitating serotonin toxicity.

There are no other serious or dangerous interactions between (therapeutic doses) of MAOIs and TCAs.

The question of whether these combinations result in an improved antidepressant effect is a different question, which is not addressed in this commentary.

For those not familiar with my numerous papers on serotonin toxicity it should be noted that these contain exhaustive details explaining the pharmacology of this interaction as it pertains to all relevant drugs and gives clear guidance on what is and is not safe.


Notes on ‘Combined MAOI-tricyclic antidepressant treatment: A reevaluation’ [A report to the American College of Neuropsychopharmacology] [16]:

This report (adopted as the ACNP position statement) is now only of historical interest, because its timing was unfortunate, as it was just after 1980 that important explanatory data about serotonin toxicity became known: particularly; the better defined clinical features of ST and its differentiation from other reactions, and the more precisely assayed potency of these drugs for different receptors, particularly the serotonin transporter. Those data provided essential explanatory power to understand this interaction which was not available to White and Simpson at that time.

Nevertheless, certain comments they made are interesting and revealing and some remain true today, as the following quotations reveal:

  1. However, a number of factors have combined to reduce their apparent risk [MAOI+TCA] or at least make it seem more acceptable: appreciation of the rarity of such crises leading to severe complications; better understanding of the mechanism of these crises resulting in prevention and dietary and drug precautions: increased awareness of the toxicity of alternate antidepressant treatments and the risks involved with inadequately treated depression: and evidence that MAOIs may have a unique role in therapy many depressed patients poorly responsive to alternative treatments.
  2. provides little support for the idea that TCP is more problematic when used in combinations with TCAs than our other MAOIs.
  3. informal discussions with many physicians it is obvious that thousands of patients have been thus treated with MAOI TCA combinations.
  4. predicting responders from clinical attributes is unreliable.
  5. the riskiness of the combination is largely superstition.
  6. in 1977 Ayd commented that the prohibition was “one of several myths and unscientific generalisations that plague contemporary psychopharmacology”
  7. [They agree] the hazards of been exaggerated, but that the risks are greater if the TCA is added with within a week or two after starting the MAOI
  8. there is little evidence of an increased rate of mild side-effects with the combination
  9. they mention possible advantages in improving sleep and possibly reducing the tyramine pressor response (citing Ghose [17]).

7: we can now be confident that statement is incorrect: there is no significant evidence that the order in which the drugs are given is relevant; but the mantra ‘start low, go slow’ should always be observed as part of the practice of good clinical pharmacology. Also concerning 9, I concur with this view.

Addendum: for those not familiar with Google searches: Google has, for several years, been placing what their algorithm designates is the ‘best answer’ available to a question in a special box at the top of the search result

As Google states: ‘Introduced in 2015, Google’s Answer Box is the holy grail for internet users. You ask, and Google answers. Appearing above the organic search results on position zero, this box features a snippet of content borrowed from a site Google thinks provides the best answer – not necessarily the #1 ranking result.’

It is relevant to understand that a great proportion of my commentaries appear on page 1 of Google, often at or near the top, and a few of them have been granted ‘holy grail’ status. That may help some people to understand why I think it is largely a waste of time for me to publish still more scientific papers, at least to achieve the objectives that I have.