The transcription factor nuclear factor erythroid 2-related factor 2 (Nrf2) is routinely positioned as the master regulator of the cellular antioxidant response. This comprehensive review evaluates both natural phytocompounds (e.g., curcumin, sulforaphane, resveratrol) and synthetic pharmaceuticals (e.g., dimethyl fumarate, bardoxolone, omaveloxolone) acting as Nrf2 activators across various inflammatory conditions, including periodontitis, diabetes mellitus, and neurodegeneration.
While the therapeutic potential for longevity applications is immense—primarily achieved by dampening chronic, low-grade inflammation (“inflammaging”) and preventing iron-dependent cell death (ferroptosis)—the researchers deliver a stark warning regarding pathway overstimulation. The Nrf2 pathway functions as a double-edged sword. Sustained or dysregulated activation can inadvertently shield early-stage malignancies from oxidative damage, promoting tumor cell survival, biosynthesis, and conferring profound resistance to chemotherapeutics such as cisplatin.
Furthermore, the review forcefully confronts the translational bottleneck of Nrf2 bio-modulation. Natural compounds routinely demonstrate efficacy in murine models but fail systematically in clinical trials due to severe pharmacokinetic constraints, including poor aqueous solubility and rapid hepatic conjugation. Conversely, highly potent synthetic activators can trigger severe off-target effects, best illustrated by the cardiovascular toxicity profile observed in clinical trials for bardoxolone methyl. Manipulating Nrf2 for lifespan extension will strictly require pulsed, context-dependent dosing rather than chronic upregulation, necessitating next-generation targeted delivery systems.
Impact Evaluation: The impact score of this journal is 16.9, evaluated against a typical high-end range of 0–60+ for top general science, therefore this is a Medium impact journal.
FWIW, tons of meds and supps people here on this site take, activate the Nrf2 pathway. Speaking just for myself all of these do so: telmisartan, empagliflozin, rapamycin, pitavastatin, ezetimibe, lithium orotate and most carotenoids, such as astaxanthin, lutein, zeaxanthin, mesozeaxanthin, lycopene.
The connection between cancer and excessive dampening of ROS signaling by sustained large doses of exogenous antioxidants has been known for a long time now, where it shows up even with vitamins, such as high dose vitamin E.
It has therefore always been a concern of mine in the back of my mind, that stacking all these “inflammation-dampening” agents may be too much of a good thing and rebound to undesirable effects such as cancer. Studies show benefits of these compounds, but that’s when they are used individually or pulsed or otherwise limited in some way. But stacking all of them on a regular, even daily basis may transpire to be quite deleterious. This is just another example of why one should study drug/supplement interactions - and why folks like Brian Kennedy report that stacking many molecules which have good effects individually are often negative together.
Not all NRF2 activators are equal. I asked Gemini 3 Pro to help identify the compounds, and rank/rate them based on the level of NRF2 activation they provide:
Ranked Hierarchy of Known NRF2 Activators
Note: Compounds are ranked by their Concentration Required to Double (CD) NRF2 activation, based on in vitro pharmacological assays.
Tier 1: Elite Potency (Synthetic Triterpenoids)
Bardoxolone Methyl (CDDO-Me): ~0.001 ÎĽM. The most potent known activator. Highly effective but clinically risky due to target saturation and cardiovascular toxicity [Confidence: High].
Omaveloxolone (RTA 408): ~0.005 μM. Structurally related to CDDO. FDA-approved for Friedreich’s ataxia. Exceptional potency with slightly better safety profiles than Bardoxolone [Confidence: High].
Tier 2: High Potency (Isothiocyanates)
Sulforaphane (SUL): ~0.4 ÎĽM. Derived from glucoraphanin in cruciferous vegetables. The gold standard for natural NRF2 activation. Features a vastly superior in vitro therapeutic index compared to synthetic triterpenoids. Highly actionable for biohackers via broccoli sprout extracts [Confidence: High].
Moringin: Isothiocyanate from Moringa oleifera. Comparable, though slightly weaker, efficacy to sulforaphane [Confidence: Medium].
Dimethyl Fumarate (DMF): ~20.5 ÎĽM. FDA-approved for Multiple Sclerosis. Requires significantly higher concentrations than sulforaphane to achieve the same NRF2 fold-activation [Confidence: High].
Oltipraz: Synthetic dithiolethione. Modest activation, historically investigated for cancer chemoprevention but limited by side effects [Confidence: Medium].
Curcumin, Quercetin, Epigallocatechin Gallate (EGCG), Fisetin, Resveratrol: >20.0 to 50.0+ ÎĽM. These compounds require massive concentrations for direct KEAP1 modification. In vivo NRF2 activation is often secondary to the xenobiotic stress they induce rather than specific KEAP1 binding. Poor systemic bioavailability renders oral monotherapy highly inefficient for NRF2 targets [Confidence: High].
Astaxanthin: Carotenoid antioxidant. Tested by the ITP for lifespan (failed to extend median lifespan in either sex) [Confidence: High].
Tier 5: Botanical Blends
Protandim: A patented mixture of Ashwagandha, Bacopa, Green Tea (EGCG), Milk Thistle (Silymarin), and Turmeric (Curcumin). Designed for synergistic activation. Tested in the ITP for lifespan [Confidence: High].
For those unaware: bardoxolone methyl and its analogs are the most effective known activators of the Nrf2 bypass pathway, but they also increase the incidence of cardiovascular death, heart failure events, non-fatal heart attacks, and non-fatal strokes. This is exactly what I’ve been saying — don’t focus on surrogate markers, focus on hard endpoints, because surrogate markers carry too little information. I see many friends who are obsessed with covering as many anti-aging pathways as possible or improving surrogate markers, and as a result they take a bunch of drugs with little or no evidence for anti-aging effects. In fact, that’s why I made my first post on this forum. I’ve shared a lot of correct anti-aging concepts before — those interested can go check them out. @DrFraser@KarlT
Of course, this is exactly the mistake I’ve made before. I think I’ve mentioned that when I first started out as a biohacker, the combination of anti-aging compounds I was taking was far more extensive than Bryan Johnson’s — simply because I wanted to achieve theoretical perfection. I believe this is one of the most common mistakes beginners make. For example, I considered taking retrotransposon inhibitors to target genomic instability, one of the 12 hallmarks of aging (now updated to 14), simply because retrotransposons are a major cause of aging-related genomic instability. But after reading tens of thousands of papers, I’ve since matured. Looking back on my anti-aging journey, I’ve really stepped into a lot of pitfalls.
I think whether you choose to use these things should also depend on two main things:
What is your baseline?
How strong is the evidence for taking them?
For example, astaxanthin has a positive lifespan extension in the ITP, and it’s been used in several human clinical trials. So together that’s pretty strong evidence. And the IPT was in mice which predominantly die of cancer, so I feel reasonably confident to say that if it was strongly promoting tumour formation, we would have seen it.
To my knowledge, evidence for sulphurophane and others is much weaker, and I think curcumin is fairly “debunked” in general nowadays.
Still I probably wouldn’t be taking every single molecule if I didn’t have a “problem” to address, such as the diabetes, neurodegeneration etc mentioned in the original post. And if you actually have cancer, you’re taking potentially bigger risks, since all of the data on your treatments (radiation, chemo etc) is based on patients presumably not taking extra molecules. So you might assist, or interfere with, your prescribed treatment.
Gemini is factually incorrect here. Astaxanthin did extend lifespan in the male mice.
I think if he asked a follow up question and mentioned the previous higher dose result about astaxanthin, Gemini would correct its answer. This seems to happen a lot with every AI app.
My concern as well. Yet I continue to err in the direction of ensuring that I consume a somewhat strategically balanced array of anti-oxidants. Based on current knowledge, I see no free lunch in navigating this choice point. The one concession I make, with little knowledge of its substance, is supplementing at levels close to what a perfect diet might contain. I do not megadose on any single anti-oxidant. I also rotate them a bit, perhaps as might occur in daily diet variations.
I agree, we have to be very cautious when looking at mechanism - as, even if that mechanism is proven to be great for longevity, the substance used to activate that mechanism can have many other untoward activities. Even then, we really need excellent safety data before even considering offering any therapy based upon logic without data of efficacy.
This is where staying up on the literature on anything I use of this type and carefully assessing and reassessing risk:benefit must be part of the role of any prescribing or advising physician.
I’ve changed my view on a few things over time based on this constant reassessment. Having really smart patients who keep on researching has made this easier.
Are you using the paid-for versions of these AI, or the free versions? I’m curious because we’ve seen some significant differences in the results from queries on free vs. paid Gemini in “Pro” mode. It seems the free uses back-rev software (earlier versions) and probably a lot less computational effort. I usually use the most intensive option available in Gemini, the “Pro” mode.
I pay for Grok and ChatGPT but not for Claude and I don’t have Gemini
I still get ChatGPT and Grok missing certain studies when it answers. When I follow up with a link to the study it missed, it revises its whole answer to account for it. A little frustrating that it doesn’t aggregate it from the start.