Brain fog is slow thinking, losing words, rereading the same page, feeling like your brain is running a beat behind. It is not a diagnosis on its own. It can come from bad sleep, low mood, thyroid trouble, medication, long COVID, or low nutrients. If it is new, persistent, or getting worse, see a clinician first.

This piece is for people who have already done the basics. Labs are fine. Sleep is decent. Caffeine helps for a couple of hours, but the fog comes back. Now what?

That is where the uncommon supplements live. They are uncommon for a reason. Most of them were first used for something else, stroke, dementia, muscle, or mood, and people started using them off label for fog because the pharmacology made sense on paper. Paper is not proof. I will walk through each one the way I think about it. What bottleneck is it supposed to fix? What would have to be true for it to work? What did a human trial actually find?

I sort fog into five bottlenecks. Low fuel, like iron or B12 or bad sleep. Low transmitter, like acetylcholine or dopamine running short. Slow membrane repair, where the brain cannot rebuild synapses fast enough. Low energy, where mitochondria cannot keep up. And inflammation or oxidative stress, where the brain is dealing with too much noise. Most common supplements fix fuel. Most uncommon ones try to fix transmitter, membrane, or energy. That is why they sound exciting, and why they often disappoint in healthy people who do not have that bottleneck.

Start with the transmitter bottleneck. Acetylcholine is the transmitter for attention and for laying down new memories. It falls with age, and brain choline uptake falls too. That is the first principles case for choline.

Citicoline, also called CDP-choline, breaks down in the body into choline plus cytidine. Cytidine turns into uridine. So one molecule feeds two pathways at once, acetylcholine synthesis and membrane phospholipid synthesis. That is elegant, and it is why citicoline gets cited so much.

The human trial people quote is Nakazaki and colleagues, 2021. One hundred healthy older adults with age associated memory impairment, 500 mg a day for 12 weeks. Episodic memory improved, and composite memory improved, but both were secondary outcomes, not the primary outcome the trial was powered for. A win on a secondary outcome is real, and weaker than a win on a primary outcome. In July 2024, EFSA looked at the claim for citicoline and memory and said a cause and effect relationship has not been established. One trial at 500 mg for 12 weeks helped, another at 1 g a day for three months did not. My read is, citicoline is plausible for older adults with mild memory complaints, not proven for a healthy 30 year old with fog.

Alpha-GPC is the other choline form, and it is more direct. It is 40 percent choline by weight, it crosses the blood brain barrier fast, and it feeds acetylcholine synthesis directly. A 2024 crossover trial gave 630 mg acutely to healthy men and found Stroop scores improved by 13 points versus 5.2 on placebo. That is a single dose, small study, but it is the kind of acute effect people feel.

Most of alpha-GPC's record is in dementia, not in healthy adults. A 2025 meta analysis pooled three trials, 358 people with dementia, and found alpha-GPC ahead of citicoline on clinician rated scales, with no difference on memory or word fluency tests, PMID 41426989. So in dementia it looks a bit better. In healthy fog, the evidence is thinner. There is also a safety flag I do not want to skip. A Korean cohort study, Lee 2021, found a higher stroke rate in alpha-GPC users, adjusted hazard ratio 1.46. That is observational, not proof of cause, and it needs a clinician conversation, especially if you have vascular risk factors. Do not take alpha-GPC long term without that conversation.

Next bottleneck, membrane repair. Your brain rebuilds synapses constantly, and it needs phospholipids to do it. The Kennedy pathway makes phosphatidylcholine from choline, DHA, and uridine. If any one of those three is short, the pathway slows. That is the first principles case for the stack people call the Mr Happy Stack, uridine plus DHA plus choline.

Phosphatidylserine is a phospholipid itself, about 15 percent of the brain's phospholipid pool, sitting in the inner leaflet of neuronal membranes where it keeps them fluid and anchors enzymes. The early trial used bovine brain extract, Crook and colleagues, 1991, 149 people aged 50 to 75 with age associated memory impairment, 300 mg a day for 12 weeks, with better learning of names and faces, paragraph recall, and phone number memory. Cenacchi and colleagues, 1993, had 497 older adults with moderate impairment, 300 mg a day for six months, with better everyday memory and mental agility. Soy derived phosphatidylserine, Kato-Kataoka and colleagues, 2010, 78 older Japanese adults, 100 to 300 mg a day for six months, helped delayed verbal recall in the lowest baseline group. It is one of the only supplements with an FDA qualified health claim for age associated cognitive decline. That does not mean it works for a healthy young adult. It means the older adult data is the best of the uncommon group.

Uridine monophosphate is the weak link in that stack, and I want to be clear about it. In rats, uridine raises brain phosphatides and can increase striatal acetylcholine release, and after six weeks it increased dopamine release in aged rats. In people, there is no trial that isolates uridine alone for cognition. The human trials used it inside a multinutrient drink with choline, DHA, EPA, phospholipids, vitamins and selenium, and the primary outcomes were mixed. So the mechanism is beautiful, the human proof for uridine alone is missing. If you take it, you are betting on the pathway, not on a trial. Studied human dose in those mixed drinks was 250 to 500 mg a day, but do not read that as a recommendation.

Next, the energy bottleneck. Mitochondria make ATP, and the brain uses a lot of it. When energy drops, thinking slows first. That is the case for creatine, and for CoQ10, PQQ, and NAD precursors.

CoQ10 sits in the electron transport chain. It helps existing mitochondria work better. The strongest human data is cardiac, not cognitive. Q-SYMBIO showed fewer major cardiac events in heart failure, and CoQ10 helps statin related muscle symptoms in some trials. For brain fog, the case is first principles plus statins. Statins lower CoQ10, and some people on statins feel foggy and achy. That does not prove CoQ10 fixes fog, but it is a reason to talk to your clinician if you are on a statin and feel foggy, rather than adding a supplement on your own.

PQQ is different. It does not just help existing mitochondria, in cell studies it activates CREB and PGC-1 alpha and increases mitochondrial biogenesis, the creation of new mitochondria. In people, the trials are small and Japanese. Itoh and colleagues, 2016, 41 healthy older adults, 20 mg a day for 12 weeks, limited the decline in Stroop attention performance. Shiojima and colleagues, 2022, 64 healthy adults aged 40 to 79, 12 weeks of PQQ, improved memory, attention, reaction time and cognitive flexibility. Those are real trials, and they are small. No human trial has shown PQQ actually increases mitochondrial density in the brain. It measures test scores, not mitochondria. So call it promising, not proven. Do not count on it as a standalone fix.

NAD precursors, NMN and NR, are the same story, louder. NAD falls with age, and NMN and NR raise NAD in blood. Animal data was dramatic. Human data is not. A recent open label study in 42 healthy adults aged 45 to 65, NMN alone and NMN in multi ingredient formulas for 42 days, found no statistically significant improvements in cardiopulmonary or cognitive function, the primary objectives. The multi ingredient group did a bit better on memory, but not significantly. That does not kill the idea, it just says the spectacular mouse results have not turned into human cognition results yet. If you take NMN, you are paying for mechanism and safety data, not for a proven fog benefit.

Two more in this energy and repair group, because people ask. Spermidine is a polyamine that triggers autophagy, the cell's cleanup process. A 2021 trial in older adults with subjective cognitive decline found spermidine improved memory consolidation versus placebo. That is one trial, and it is interesting because the mechanism, clearing damaged mitochondria, fits fog after viral illness or poor sleep. Urolithin A triggers mitophagy, a more specific cleanup of damaged mitochondria. A 2022 trial found it improved muscle endurance in middle aged adults by 10 to 12 percent. Muscle is not brain, but the pathway is shared, and that is exactly the kind of first principles leap you should label as a leap, not as proof.

Now the inflammation and oxidative stress bottleneck. NAC, N-acetyl cysteine, was first a mucolytic and a paracetamol overdose antidote. Off label, people use it for brain fog because cysteine is the rate limiting building block for glutathione, the brain's main antioxidant, and because NAC modulates glutamate through the cystine glutamate antiporter. In plain terms, it helps the brain clear oxidative noise and calm excess glutamate signalling. Human trials are moderate, usually 600 to 1,800 mg a day in divided doses for 2 to 4 weeks, with the best signal when oxidative stress or toxin exposure is involved. It can upset the stomach, it smells like sulfur, and it interacts with nitroglycerin and some blood pressure drugs. It is not a daily focus pill.

Curcumin fits here too. It is anti inflammatory and antioxidant in lab studies, but it barely gets absorbed. The trial people quote is Small and colleagues, 2018, a bioavailable form in non demented adults for 18 months, with better memory and lower amyloid and tau signals on brain scans. That is a long trial, and it used a form designed to get into blood. Regular turmeric powder will not do the same thing. If your fog follows a viral illness or tracks with high hs-CRP, curcumin is more plausible. If your inflammation markers are normal, expect less.

Lion's mane fits the repair bottleneck, but through a different route. It does not supply a building block. It signals the brain to make more of its own nerve growth factor, through hericenones in the fruiting body and erinacines in the mycelium. Mori and colleagues, 2009, PMID 18844328, 30 adults aged 50 to 80 with mild cognitive impairment, 3 g a day for 16 weeks, improved cognitive scores from week 8 onward, and the gain faded four weeks after they stopped. That fade is actually reassuring, it looks like a real ongoing signal, not a fluke. But it was 30 older adults with impairment, not healthy young adults with fog. A 2023 pilot in young adults found faster attention an hour after a single 1.8 g dose, and no change on overall score in another small study. So lion's mane is moderate evidence for mild impairment, thin for everyday fog.

What I would not put in a stack, even though people ask, is methylene blue and lithium orotate. Methylene blue can interact dangerously with SSRIs and cause serotonin syndrome, and the human cognition trials are few and small. Lithium orotate is sold as a low dose mood supplement, but lithium has a narrow safety window, interacts with kidney function, thyroid, and diuretics, and needs blood monitoring even at low doses. First principles does not make a risky compound safe. I am leaving both out on purpose.

How would I use all of this, if labs and sleep were already fine? I would pick one bottleneck, not five supplements at once. If fog is attention and word finding, try a choline form first, citicoline 500 mg a day for 12 weeks, and track one test, like paired associate memory or just how often you lose a word. If fog is memory and names, try phosphatidylserine 300 mg a day for 12 weeks, same tracking. If fog is energy and you are older or sleep deprived, try creatine plus CoQ10, not PQQ plus NMN plus spermidine all at once. If fog follows stress or a virus, try NAC or curcumin, and check inflammation markers first. Give each one 8 to 12 weeks, except the acute ones like caffeine plus L-theanine which you judge in hours. Stop what does not help. Keep what does, and check interactions before you add the next one.

A protocol is not a longer list. It is a shorter list, in the right order, with the interactions checked first.

This is education, not medical advice. It is not a diagnosis, and it does not replace professional care. Off label use, especially for alpha-GPC, NAC, saffron, and any mitochondrial stack, needs a clinician conversation first, especially if you take prescription medicine, are pregnant, or have kidney, liver, thyroid, or heart issues.

Sources: Nakazaki and colleagues, Journal of Nutrition, 2021. Sagaro and Amenta, Frontiers in Neurology, 2025, PMID 41426989. Crook and colleagues, Neurology, 1991. Cenacchi and colleagues, Aging, 1993. Kato-Kataoka and colleagues, Journal of Clinical Biochemistry and Nutrition, 2010. Mori and colleagues, Phytotherapy Research, 2009, PMID 18844328. Small and colleagues, American Journal of Geriatric Psychiatry, 2018. Itoh and colleagues, 2016. Shiojima and colleagues, 2022. Xu and colleagues, Frontiers in Nutrition, 2024.