Exploring MAOIs for Depression

Ben Mann
3 min readMar 5, 2024

I’m considering trying an MAOI (monoamine oxidase inhibitor) instead of my current SNRI (serotonin-norepinephrine reuptake inhibitor) for treating depression. While there are some potential risks, like interactions with tyramine in foods leading to acute hypertension, it seems these may be overblown according to this paper (though it doesn’t look very reputable and needs further investigation).

Notably, Scott Alexander of Slate Star Codex loves MAOIs. He argues they can be very effective even for treatment-resistant depression.

The paper discusses how tyramine response can be measured relatively safely:

A standard procedure was developed in humans for giving increasing doses of Tyr until the systolic blood pressure (SBP) was elevated by >30 mm Hg: this is referred to as the Tyr30. The oral Tyr30 in unmedicated subjects, fasted and given Tyr capsules (Tyr ingested in a meal has a lesser effect, so Tyr30 is greater), is approximately 500 mg (200–800 mg). It is a time-consuming procedure, so the number of tests done is small and data must be interpreted with that in mind. Inter-subject variation (n=55) was estimated to be threefold (Bieck and Antonin 1989).

Essentially, you can test tyramine response by taking increasing doses and monitoring blood pressure with a home BP cuff until systolic BP rises 30 mmHg. This Tyr30 dose varies by 3x between people and is higher when tyramine is consumed with food vs fasting.

The key takeaway seems to be:

for a majority of patients on MAOIs, the estimated Tyr dose (in a meal) required to provoke a problematic pressor response is in excess of 50 mg… For other drugs such as moclobemide, selegiline, or rasagiline that have much lower Tyr potentiation, the amount of Tyr needed to induce a problematic pressor response would be commensurately (at least ×5–10) higher, i.e. at least 200 mg (Goren et al. 2010; Knoll and Magyar 1972).

Even if blood pressure does spike from tyramine while on an MAOI, the risks of acute events may be lower than feared:

Short-lived (1–2 h) elevations of BP in ‘hypertensive urgency’ range from 180 to 220 mm/Hg and do not engender a significant acute risk of serious consequences…

Large increases of BP rarely precipitate subarachnoid haemorrhage, most occur ‘spontaneously’ or with short and modest BP increases associated with, for instance, defaecation

So by starting low, testing response, and being cautious with tyramine intake, the risks of hypertensive crisis on MAOIs are likely manageable. Tyramine content also varies a lot by food preparation:

Most pizza styles use mozzarella cheese, which has no Tyr, although other cheese styles may be used (cheddar, Edam, etc.); these are ‘commercial’ styles that are stored frozen and are highly unlikely to be high in Tyr… Such considerations illustrate why simple tables and lists of ‘prohibited’ foods are of limited use and why the advice and knowledge of an informed physician are required.

Interestingly, one study found no dietary restrictions were needed with selegiline transdermal patches, due to low, stable blood concentration from the patch delivery. However, selegiline patches are very expensive at $2000/month compared to only $40/month for clomipramine and phenelzine pills.

Another analysis found the MAOIs clomipramine and phenelzine were among the most effective antidepressants, in line with Scott Alexander’s conclusions, though the study sizes are quite small, with a mean of only n=58 per treatment arm across trials.

Money plots for acceptability (did patient continue taking the drug?) and efficacy (how well did it work while taking it?)

Overall, MAOIs seem to be underused options for treatment-resistant depression that may be worth trying, with proper medical guidance, before giving up on medication altogether. The dietary and drug interaction risks, while real, are likely manageable in most cases. Cost may be the bigger barrier for some of the newer MAOI options. Based on all this, I plan to phenelzine since I’m not as concerned about acceptability and am mostly interested in trying the best thing. If that doesn’t work well, I’ll try clomipramine.



Ben Mann

Software engineer, tinkerer, aspiring mad scientist