What did the UC Irvine team actually find?
A group at UC Irvine made a splash recently with a paper in GeroScience. They combined two compounds, nicotinamide (the amide form of vitamin B3) and EGCG (a green tea catechin), and reported that the pair restored something called GTP inside neurons. Within 24 hours in the dish, the cells' cleanup machinery started moving again and intraneuronal amyloid signal dropped.
The tissue was hippocampal neurons from aged Alzheimer's-model mice. The medial temporal lobe, including the hippocampus, is one of the first regions to show trouble in Alzheimer's, so it is a reasonable place to look. The researchers measured free GTP in those neurons and found it was low, with sluggish vesicle traffic. Then they bathed the neurons in nicotinamide at millimolar levels for about 16 hours, plus EGCG at low micromolar levels. By the end, free GTP rose toward youthful levels, the endocytosis and autophagy markers normalized, intraneuronal amyloid fell, and oxidative damage markers fell too.
The cells looked healthier. In a petri dish, that is exciting. The tissue came from mice, not humans, and these were isolated neurons in a dish rather than a whole brain under physiological conditions. The rest of this article is about what might carry over to a human brain and what probably will not.
Why does GTP matter inside a neuron?
We usually treat ATP as the universal energy molecule. The internal traffic of a neuron runs on a different fuel. The trafficking of vesicles, endosomes, and autophagosomes is run by enzymes called GTPases, with names like Rab7 and Arl8b. These enzymes burn GTP, cycling guanosine triphosphate to diphosphate the same way you cycle ATP to ADP. Burning GTP powers the molecular switches that move cargo along microtubules and fuse cell membranes.
When free GTP inside a neuron falls through aging or chronic stress, those switches stall. Vesicles pile up. Proteostasis, the balance between protein synthesis and protein cleanup, suffers. The cell's cleanup crew slows down, and debris accumulates.
Here is the mechanism the two compounds target. Nicotinamide feeds the NAD salvage pathway, which supports mitochondrial redox and the TCA cycle. The TCA cycle has a specific phosphorylation step, at succinyl-CoA synthetase, that produces GTP directly. More NAD supply, better redox, more mitochondrial GTP, happier GTPases.
EGCG is doing something other than acting as an antioxidant. It donates electrons and pulses the NRF2 pathway, the cell's stress-response system. As the catechin comes in, it adds a small charge to the cell. The stress response kicks into a protective gear. The result is a cell running hotter, with more metabolism, but doing so more safely because the polyphenol primed the protective response. This is a hormetic effect: a mild stress that triggers a resilience program.
How often does a mouse-dish finding reach humans?
This matters more than the headline. Last week I estimated that roughly 10% of mouse findings translate to humans, and I went back to check that number rather than throw a statistic around.
The answer splits by what you are translating. For basic intracellular mechanisms, the kind this study examined, a meaningful fraction replicate in humans once they have been found in animals. That is reasonably good. For treatments, where you dose a whole animal and try to reproduce the effect in a whole human, the rate drops sharply.
This study sits below even that threshold, because it is cells in a dish, not a treated mouse. Brains do not get bathed in millimolar solutions of nicotinamide. To recreate any of this in a person, you have to get the compound through absorption, first-pass metabolism, protein binding, transporters, and the blood-brain barrier. That delivery problem is where the story gets messy, and also where there is some real hope.
What happens when humans actually take nicotinamide?
A Phase 2a trial in early Alzheimer's gave people 1.5 grams of nicotinamide twice a day for 48 weeks (Loy et al., 2024). Safety was good. The primary CSF marker, phosphorylated tau, did not move at the group level.
The pharmacokinetics tell a more interesting story. When the researchers looked across individuals, only a subset of participants had measurable nicotinamide in their cerebrospinal fluid. Many of the others methylated it to MeNAM in peripheral tissue before it reached the brain. In the subset that did show CSF exposure, phospho-tau dropped.
The bottleneck is delivery and metabolism, not the target biology. The compound seems to do something useful when it arrives. The problem is getting it past peripheral methylation and into the brain.
Why is nicotinamide riboside more interesting for the human brain?
Nicotinamide riboside (NR) is a nucleoside rather than an amide, so it can use different transporters to reach the brain. The equilibrative nucleoside transporters, ENT1 through ENT4, sit on the microvascular endothelium of the blood-brain barrier, and ENT2 has been functionally implicated in transport there. Once inside cells, NRK enzymes phosphorylate NR into NMN and onward into NAD.
We have human data on central uptake. A study in Parkinson's patients gave 1 gram of NR per day for 30 days and saw cerebral NAD rise, measured by phosphorus MRS, with a metabolic shift in a brain-relevant direction (Brakedal et al., 2022). In healthy volunteers, a single high oral dose raised brain NAD by downfield proton MRS (Bagga et al., 2020). A third study in older adults used NR for several weeks, raised NAD in neuron-enriched extracellular vesicles, and reduced amyloid-beta 42 and stress-kinase phosphorylation, again among responders (Vreones et al., 2023). There appears to be a responder versus non-responder split in that work, with a large effect in those who respond.
Even high doses were generally well tolerated, with robust blood NAD rises and a modest, transient homocysteine bump in several people (Dollerup et al., 2018). The homocysteine shift is something you can monitor.
NR gives you a lever to reach the energy and redox environment of the brain in a way the straight amide does not.
What about EGCG and the liver risk?
EGCG can cross the blood-brain barrier, but levels are typically low after oral dosing, and taking it with food drops absorption further. The bigger concern shows up at the other end. In fasted states or at high bolus doses, EGCG extracts carry liver risk. Regulatory reviews, including EFSA's 2018 opinion (EFSA, 2018) and the UK COT 2024 update, keep around 800 mg/day of EGCG as a practical ceiling for supplemental use, with idiosyncratic hepatotoxicity above that.
The earlier safety literature pointed to extraction methods leaving problem compounds in the extract. The newer reading suggests direct risk from supplementing high doses, especially fasted.
You do not need bath-level concentrations to pulse NRF2. The cell responds to transient electrophilic cues. Green tea as a beverage delivers the signal without the bolus liver exposure. If you supplement the extract, avoid the giant-dose products and consider checking liver enzymes before and a few weeks after, with a physician.
How would you design an N-of-1 around this?
This is education, not medical advice, and any self-experiment with these compounds deserves physician oversight.
Start with the goal, not a product. The goal is neuronal energy and redox, so that GTP-dependent trafficking and autophagy are less likely to bottleneck. You want movement inside the cell.
For most people, NR is the primary lever. We have human evidence of it raising brain NAD, and a workable dose range looks like a gram a day for a month, or split dosing across several weeks. If you push higher or longer, track homocysteine and B-vitamin status, and consider cycling. Pair it with a redox nudge from the polyphenol, starting with green tea as tea before reaching for a supplement.
Set expectations honestly. This is not something you take and feel in an hour. Over weeks to months, the hypothesis is smoother energy metabolism in the brain. The conceptual target with EGCG is a mild NRF2 pulse, not the highest dose you can swallow.
This compound work sits alongside several aging mechanisms I have covered recently, including iron and lithium. There is a pattern emerging where the action is inside the cell rather than in the classic blood-sugar story. I read that as a function of where the field has been looking. After years focused on metabolic phenomena at the whole-body level, the closer look at within-cell metabolism is where a lot of the Alzheimer's biology now lives. If you are thinking about the aging trajectory more broadly, I have written about why the brain's decline starts earlier than most people assume in the critical aging window, and about strategic fasting and brain fog for adjacent metabolic levers.
Questions from the stream
A few audience questions touched on the EEG side of this work, so I will fold them in.
Does coconut oil change the EEG?
It depends on whether you are someone who converts MCT to ketones efficiently. If you are fasted, the EEG slows a little, with peak alpha dropping up to half a hertz and a small dip in power. My hunch is that people who are fat-adapted and used to burning ketones, through diet or exogenous sources, will see a slight speeding up and a power shift with the MCT in coconut oil. Coconut oil also contains monolaurin, which has a special fatty-acid profile, and the only other place humans encounter it is breast milk. Whether there is an acute cognitive effect is an open question. You can test it: map yourself fasted, fed without MCT, and with MCT, and compare. I would bet your alpha speed differs across the three.
Can neurofeedback raise or lower EEG power?
Yes, both. You often decrease overall power in deep fatigue or illness, increase it where it is suppressed, and tamp down focal hot spots where there is excess theta or beta and not enough alpha. Reducing power in a band that is running too hot is a routine part of the training. If you want the longer version of how training reshapes resting bands, QEEG brain mapping and SMR neurofeedback cover the targeting logic, and alpha waves covers what the idle rhythm is doing.
Why is excess delta present, and should you inhibit it?
Delta is the slowest rhythm, the metabolic repair and power-down mode, running about twice a second. You live in it more than you think in it. When delta is excessive and focal, that often points to an injury. When it is broad, it is usually doing something metabolic or inflammatory. I would not normally target delta directly, because it is supporting a healing process. I would look at theta as the likelier culprit, the tissue running in automatic mode, which rides high alongside delta when things are broadly disinhibited. Bring beta up, bring excess theta or alpha down, and the delta picture usually improves.
Can strong plasticity boosters lock in an anxious state?
Plausible, and the mechanism makes sense. Psilocybin and similar compounds create a broad, unchecked plasticity boost. Your brain changes in whatever direction it is experiencing things, and a charged or frightening experience during a high-plasticity window can get seared in. This is not unique to psilocybin. Lion's mane does it, and I have heard anecdotal reports of trauma material locking in under racetams when attention was unusually crisp. A blanket plasticity boost during an intense experience can produce a PTSD-like adaptation toward higher anxiety. If you are interested in shaping plasticity deliberately rather than flooding it, I have written about biohacking plasticity.
The bottom line
Mouse hippocampal neurons in a dish gave us a tractable idea: intraneuronal GTP, redox balance, and autophagy are mechanisms we may be able to nudge in living people. The human-feasible version, supported by existing pharmacokinetic and brain-imaging data, is NR to raise brain NAD paired with green tea as a gentle NRF2 pulse, both under medical supervision. If you want a one-page visual walking through the TCA cycle and the GTP link in more detail, tell me and I will build one for a future stream.