An experiment with N-acetylcysteine

The case study paper on which my symptom treatment strategy is based uses a combination therapy with daily guanfacine and N-acetylcysteine (NAC). Guanfacine has a clear purpose: it's an ADHD medication, and these have been found in several instances to be useful for reducing long covid's cognitive effects.1 The purpose of NAC is less clearly established; it's known to replenish glutathione levels, it also inhibits the enzyme kynurenine aminotransferase, which is involved in a neurotransmitter-released tryptophan metabolism pathway,2 and it can also modulate calcium entry into mitochondria and reduce inflammatory signalling as a result.

Since oxidative stress, inflammation, neurotransmitters, and tryptophan metabolism have all been implicated in at least certain subtypes of long covid, it seems plausible NAC could help, so I wanted to try it. My doctor was (reasonably) hesitant because he'd never used it before and didn't know much about it. After some more reading, I decided it was relatively low risk for me to try it myself (it's available as a supplement in Canada). I ordered this.3

I tried the NAC at 600 mg/day with food (the dose used in the paper), starting about three weeks after my most recent dosage adjustment of guanfacine. I stayed on the NAC for about three weeks, then went off for two weeks, then on for three weeks again, then off again. I chose this protocol to balance the fact that I seem to be having a rather slow long-term decline in my baseline level of functioning (so I want to do the experiments on as short a timescale as possible), with the fact that I don't really know the timescale on which I should expect NAC to act (so I want to do the experiments on as long a timescale as possible). I also seem to have occasional periods where I feel a little better for a short time, so I wanted to do at least two on/off cycles. Glutathione probably turns over in the body on the scale of hours to days,4 so two weeks is more than enough to modulate that process. I don't know of any good information on the timescale of tryptophan metabolism's involvement in neurotransmission, or accumulation of relevant oxidative damage, so I wanted to try a bit longer than just a couple turnovers of glutathione to be more conservative.

In short, I noticed no difference in fatigue, cognitive performance, or postural tachycardia / dizziness when starting or stopping NAC in either cycle. At least for me, it does not seem to make a large, immediate, persistent difference like guanfacine does. That said, one possible effect it could be having (especially given its involvement in oxidation / reduction and inflammation) is preventing further damage from accumulating, such that I would have seen myself worsening otherwise. This is not a straightforward thing to figure out from single-person studies, so hopefully we'll see a randomized, controlled, double-blinded trial for NAC (perhaps with and without guanfacine) soon.

For now, I'm not going to continue with NAC. While it's generally pretty innocuous, and I've had no adverse effects, I'd generally prefer to be taking fewer things rather than more, and continuing could potentially confuse future experiments.


  1. In addition to the paper with guanfacine, there's methylphenidate and lisdexamphetamine, as well as atomoxetine with positive case studies. I've heard more anecdotal patient reports of positive effects with other ADHD medications too.

  2. The paper explains the pathway and hypotheses with diagrams in a reasonably approachable manner if you're used to reading bio papers, and there's also some good discussion of tryptophan metabolism in long covid in another paper on serotonin reduction in long covid.

  3. My general approach to choosing a supplement vendor if I don't have a trustworthy recommendation is to find one that's sold by a company based in or with a substantial presence in the country where I'm curently living. My rationale is that it's considerably easier for someone to have brought legal action against a company in the same country, which therefore reduces the probability that a company that is still able to sell things in the country is selling completely fake products. I chose this brand as the only option I found that meets the intersection of this criterion and being explicitly gluten-free (since I have celiac disease). That said, I have no tools to do any analysis at home, so the unknown quality of what I ordered is a major potential issue with this experiment.



This work by Colin J. Fuller is licensed under CC BY-NC-ND 4.0