NOT MEDICAL ADVICE. I'm not a doctor. Nothing here is a recommendation to use any compound. This is a sourced harm-reduction breakdown. Do your own research, make your own decisions, own the consequences.

All citations traceable to peer-reviewed sources. PubMed IDs (PMID) included where confirmed.

The Problem With the Feed
There's a pattern that's been spreading through PED content lately. Someone posts a video claiming tren wrecked their memory and Dihexa fixed it. Affiliate code in bio. "Ten million times more potent than BDNF." The comments fill up with guys asking where to source it, what dose, whether they can stack it with semax. Then someone else posts a similar video. Then another. The claim gets laundered through enough accounts that it starts to feel like established knowledge.
To be clear: most of the people pushing this aren't malicious. A lot of the bigger voices in this space genuinely try to help the community. But good intentions don't make bad evidence good, and the affiliate model creates real pressure to talk up products whether the underlying science supports them or not. Dihexa is a case where the hype has dramatically outrun the actual literature -- and the gaps in that literature include things that people buying this compound deserve to know about.
The mechanism behind Dihexa is based on real biology. But the specific compound? The evidence is a disaster. The key paper that explains how it works was retracted in 2025 because the data was fabricated. The pharmaceutical company that tried to turn it into a real drug ran three clinical trials and failed all three. Nobody has ever given it to a human being in a controlled study. And there's a legitimate cancer concern that none of the content pushing it is addressing.
This post is three things: what tren and other compounds like it actually do to your brain (it's worse than most people realize), why Dihexa specifically is a bad bet right now, and what the research actually supports as harm reduction. Everything is sourced. If you want to go deeper on anything, the citation list at the bottom will get you there.

Part 1: What These Compounds Actually Do to Your Brain

Most of the Dihexa content online skips past the damage itself and goes straight to "here's the solution." That's backwards. You need to understand what's actually happening first.

Your Brain Is Being Oxidized
The most documented way that nandrolone, trenbolone, and similar compounds damage the brain is through something called oxidative stress. Here's the simple version: your cells constantly produce unstable molecules called free radicals as a byproduct of normal energy production. Your brain has a built-in system to neutralize them -- the main tool is a compound called glutathione (GSH), which acts like an antioxidant sponge inside your neurons. As long as glutathione keeps up, the free radicals get mopped up before they cause serious damage.
What AAS (Anabolic-Androgenic Steroids) do is overwhelm that system. Studies in rats given nandrolone showed measurable damage to cell membranes across the prefrontal cortex (decision-making, judgment), striatum (movement, habit, reward), and hippocampus (memory formation) within 8 weeks -- while simultaneously depleting the brain's own glutathione stores (Turillazzi et al., 2016; El-Shamarka et al., 2020). The compound is consuming the very defense system your brain uses to protect itself.
At the same time, a molecular switch called NF-kB (Nuclear Factor kappa-B) gets flipped on. Think of NF-kB as your brain's fire alarm. When it goes off, it signals the brain's immune cells to flood the area with inflammatory chemicals. That's useful if you have an infection. When it's running constantly because of chronic AAS exposure, it creates a loop: the oxidative damage triggers inflammation, the inflammation generates more oxidative damage, repeat. The brain has no off switch for this while the insult is ongoing (Turillazzi et al., 2016).

The Inflammatory Chemicals Are Not Harmless Background Noise
When that fire alarm (NF-kB) goes off, it causes the release of two specific inflammatory proteins in brain tissue: TNF-alpha (Tumor Necrosis Factor-alpha) and IL-1beta (Interleukin-1 beta). These sound like jargon but the effects are concrete.
TNF-alpha at chronically elevated levels in brain tissue directly kills neurons. That's its job in acute injury -- it signals damaged cells to die. When it's elevated chronically from AAS exposure, it starts doing that to healthy neurons. IL-1beta disrupts synaptic plasticity -- the process by which your brain physically strengthens the connections between neurons when you learn something or form a memory. If you're noticing things not sticking, words not coming as fast, general cognitive sluggishness -- this is part of the mechanism. Both TNF-alpha and IL-1beta were measurably elevated in the hippocampus and prefrontal cortex of rats on nandrolone.

Your Dopamine System Is Getting Specifically Targeted -- But the Story Is More Complicated Than You Think
This is where things get counterintuitive, because many of you reading this have probably experienced the exact opposite on tren: increased drive, intensity, aggression, and sex drive that borders on obsessive. Tren's reputation as "divorce in a bottle" is earned. The libido and drive effects are real and well-documented in the community.
So how do you reconcile "tren turns you into an animal" with "tren damages your dopamine system"? The honest answer is that both are happening, and the relationship between them isn't a contradiction -- it's actually evidence of the problem.
Androgens, especially at supraphysiological levels and especially with tren's extreme receptor affinity, acutely flood dopaminergic circuits -- the reward and motivation systems -- with activity. That's why the drive, the intensity, the hypersexuality are all turned up. But simultaneously, the research shows the underlying neurons are under stress. Cunningham et al. (2009) showed that the same receptor activation that drives these acute effects also triggers a cell death pathway specifically inside dopamine-producing neurons. The dopamine neuron gets activated hard, while a stress cascade builds inside it. The neuron is working overtime and dying for it.
Think of it like a light bulb being run at 200% voltage. It burns brighter -- for a while. Then it burns out. The guys who come off a long tren run and feel flat, unmotivated, anhedonic (nothing sounds good, nothing feels rewarding, just going through the motions) -- that's what depleted or damaged dopamine circuitry actually feels like. The fire during the run and the crash after aren't separate phenomena. They're the same mechanism, different timelines.
On top of that, even the neurons that survive lose sensitivity. Kindlundh et al. (2001) found significant downregulation of dopamine receptors in the brain's reward center after just two weeks of nandrolone. The signal is being sent but the volume knob for receiving it is getting turned down. Tucci et al. (2014) measured actual reduced dopamine, serotonin, and noradrenaline content in the reward center alongside measurably anhedonic behavior in the animals. The flatness, the grey feeling, the low drive that hits some guys weeks or months after a tren run ends -- this is the biochemical basis of it.

Tren Is Worse Than Other AAS for This. Not Slightly -- Categorically.
Three reasons trenbolone stands apart from other anabolic steroids when it comes to brain damage:
No estrogen protection. When your body converts testosterone to estrogen (a process called aromatization), that estrogen turns out to protect neurons. Orlando et al. (2007) demonstrated this by blocking that conversion in rats -- even at low testosterone doses, the neurons became neurotoxic once estrogen was taken out of the equation. Testosterone users get this protection as a side effect of their aromatization. Trenbolone has a specific chemical feature (a 9,11-double bond) that permanently prevents it from ever converting to estrogen. There's no aromatization to manage, which sounds like a benefit -- until you realize you've removed the neuron protection that came with it.
It hits the androgen receptor harder. Trenbolone binds to androgen receptors (ARs -- the receptors that AAS activate in your cells) with roughly 5 times the force of testosterone. Every neurotoxic effect that flows from receptor activation is amplified proportionally.
It does things in the brain that other AAS don't. Havelin et al. (2021) ran a direct head-to-head comparison of testosterone, nandrolone, and trenbolone in live rat brain cells. Trenbolone was the only one that killed neurons, damaged the cell's mitochondria (the cell's power generators), and shut down a gene called Tubb3 that neurons need to build and maintain their structural scaffolding. And Ma and Liu (2015) showed that trenbolone specifically accumulates in the hippocampus -- the region most critical for memory -- and increases production of a sticky protein called Abeta42 (amyloid-beta 42) that clumps together to form plaques in Alzheimer's disease. This happened dose-dependently and was not prevented by adding testosterone alongside the trenbolone. The compound is doing specific structural damage to brain tissue that testosterone and nandrolone don't replicate.

The Human Data: The Effects Are Real, Measurable, and They Don't Go Away Quickly
None of this is purely in rats. In actual humans:
Kanayama et al. (2013) put AAS users through a validated neuropsychological test battery and found measurable, dose-dependent memory deficits -- the more AAS someone had used over their lifetime, the worse their spatial memory performance. Bjornebekk et al. (2019) found significant deficits in memory, processing speed, working memory, and problem solving in 84 AAS users versus 69 non-using controls -- even after statistically removing the effects of age, education, and IQ. These weren't small effects.
On brain scans, Bjornebekk et al. (2017) found AAS users had physically thinner brain tissue across multiple regions, with longer use predicting greater thinning. A 2021 study using machine learning to estimate brain age from scans found that AAS users' brains looked older than their actual age -- the effect was particularly pronounced after 10 or more years of use.
And these effects don't just bounce back when you stop. Animal data found that after stopping nandrolone, the dopamine system took approximately five times the cycle length to normalize. The serotonin system took approximately six times. A 16-week run means potentially over a year before dopamine normalizes -- if it does fully. Kildal et al. (2022) found that former AAS users who had stopped scored just as high as current users on measures of ADHD-like symptoms. A 2025 conference report documented elevated cognitive problems, depression, and anxiety in former users who had been off for an average of almost two years. Some of the neuron death is permanent -- dead neurons don't come back.

Part 2: The Dihexa Problem
What It Is
Dihexa is a small synthetic peptide that came out of research at Washington State University. The researchers found that a certain hormone fragment had unexpected brain-protective effects by activating a growth factor system in the brain -- specifically by making a protein called HGF (Hepatocyte Growth Factor) more potent at activating its receptor, which triggers neurons to grow new connection points. In cell culture experiments, it caused dendritic spine growth (the little branches on neurons where connections form) at extremely tiny concentrations -- which is where the "10 million times more potent than BDNF (Brain-Derived Neurotrophic Factor -- the brain's main repair and growth protein)" claim comes from. It worked at much lower concentrations in a petri dish. That's it. That's the whole basis for the claim. A petri dish experiment. There are no human studies.

The Evidence Is a Wreck
The total published scientific literature on Dihexa is about six papers. Here's the honest breakdown:
The paper that established exactly how Dihexa works -- the mechanism that every vendor and content creator is citing when they explain it -- was formally retracted in April 2025. Washington State University investigated and confirmed the data was fabricated. Not disputed, not questioned -- fabricated. The co-authors responsible had their contributions formally disavowed.
The original behavioral study -- which showed Dihexa reversed memory impairment in rats that had been given a drug to cause amnesia -- received a formal "expression of concern" from the journal in 2021 due to data integrity issues. It hasn't been retracted, but it's been flagged (McCoy et al., 2013, PMID: 23592440).
One legitimate independent study exists -- a Chinese research group treated genetically engineered Alzheimer's model mice with oral Dihexa and found improvements in memory and reductions in brain inflammation. This one is in good standing and is the strongest remaining evidence. It is a mouse study (Sun et al., 2021, Brain Sciences).
One independent study produced a completely negative result -- researchers tested Dihexa in a rat model of Huntington's disease neurotoxicity. Zero benefit. No protection of any kind across any measure tested (Wells et al., 2024, Journal of Huntington's Disease).
Also worth knowing: in every study where Dihexa did show any effect at all, it only worked in animals that already had cognitive impairment from a drug or disease. It had no effect on animals with normal brain function. The entire "nootropic for healthy people" angle has zero support in the published data.

No Human Has Ever Taken It in a Controlled Study

Zero clinical trials on ClinicalTrials.gov. Zero human pharmacokinetic data -- meaning nobody knows what dose actually reaches the brain, how long it stays, or what it does to human tissue. Zero published case reports. It is sold as a "research chemical" with a label that says "not for human use."
A pharmaceutical company called Athira Pharma tried to turn Dihexa into a real drug. They took a modified version called fosgonimeton and ran actual clinical trials in humans with Alzheimer's disease and Parkinson's disease dementia. Three trials. All three failed -- the treatment groups were statistically indistinguishable from placebo. Athira then laid off 70% of their staff, paid $10 million to settle a securities fraud lawsuit, and paid another $4 million to the Department of Justice (DOJ) for concealing research misconduct from the government agency that had funded the research. The researcher who originated the work had her PhD revoked by Washington State University.
That's the clinical history of this compound's lineage.

There Is a Real Cancer Concern and Nobody Is Talking About It
This is the part that should give anyone serious pause.
The receptor system that Dihexa activates -- the HGF/c-Met system -- was originally discovered as a cancer-causing gene. When this system gets abnormally activated, it tells cells to divide, spread, grow new blood vessels to feed themselves, and resist dying. These are the defining features of cancer progression. The same pathway is involved in kidney cancer, lung cancer, liver cancer, stomach cancer, brain cancer, and several others. It's implicated so reliably in cancer that the pharmaceutical industry has developed multiple FDA-approved drugs -- capmatinib, tepotinib, cabozantinib, and others -- whose sole purpose is to block this pathway in cancer patients.
Dihexa chronically activates the pathway those cancer drugs exist to turn off.
No long-term safety studies have ever been run for Dihexa. The clinical trials that tested its pharmaceutical derivative excluded anyone with any history of cancer -- meaning even the limited human safety data that exists tells us nothing about cancer risk. The Alzheimer's Drug Discovery Foundation (ADDF) specifically warned in their review that Dihexa could theoretically promote tumor growth and cancer progression, and flagged the complete absence of any long-term safety evaluation.
Now think about the typical advanced stack context. If you're already using testosterone, GH (Growth Hormone), and IGF-1 LR3 (Insulin-Like Growth Factor 1 Long R3 -- a modified growth factor that specifically drives cell growth and division), you are already running multiple systems that promote cell growth simultaneously. Adding a compound that activates a known cancer-promoting pathway on top of that, with zero human safety data, is a risk that has never been characterized. Nobody has studied that combination. Not even the researchers who developed Dihexa.
One more practical detail: Dihexa has an estimated half-life of about 12 days in rats. If something goes wrong, you cannot clear it quickly.

The Bottom Line on Dihexa
The excitement in this community is based on a single flashy potency comparison from a petri dish, a mechanistic paper that turned out to contain fabricated data, aggressive marketing by peptide vendors presenting cell culture results as clinical evidence, and a self-reinforcing cycle of anecdotes that traces back mostly to a single 2015 Reddit post about a one-week experiment. Self-reported user results are all over the map -- including people who report cognitive slowing, attention problems, and many who noticed nothing at all.
Nobody pushing this compound with an affiliate code is doing a full literature review before they post. This is that review.

Part 3: What Actually Has Evidence Behind It

First: IV and IM Glutathione Also Isn't Doing What You Think
Before covering what works, the other popular "neuroprotective" protocol in this community deserves the same scrutiny. Injectable glutathione -- whether IV (intravenous) drip or IM (intramuscular) shot -- is everywhere. The pharmacology doesn't really support it for brain protection.
Here's the problem. Your brain makes its own glutathione internally -- it doesn't import it from your blood. The reason is structural: the BBB (Blood-Brain Barrier), the highly selective filter between your bloodstream and your brain tissue, doesn't let intact glutathione through in meaningful amounts. An enzyme sitting right on the surface of the blood-brain barrier immediately chops up any glutathione arriving from the blood. What crosses instead are the broken-down building blocks, which brain cells then reassemble into new glutathione on the other side.
IV glutathione has a half-life of about 2-3 minutes in the blood before this enzyme destroys it. IM glutathione takes longer to reach peak blood levels but faces the same barrier once it gets there. The one properly controlled clinical trial that tested injectable glutathione in Parkinson's disease patients -- a disease where brain glutathione depletion is a major feature -- found no benefit over placebo (Hauser et al., 2009, PMID: 19230029). Oral glutathione is even less useful -- your gut breaks it apart before it can be absorbed (Witschi et al., 1992, PMID: 1592668). The Michael J. Fox Foundation, after funding research into this, publicly stated that researchers were unable to show that administered glutathione crosses the blood-brain barrier.
The one exception is intranasal glutathione -- sprayed directly up the nose -- which can bypass the blood-brain barrier by traveling along nerve pathways. One small study in Parkinson's patients confirmed it raised glutathione levels in the brain when delivered this way (Mischley et al., 2017, PMID: 28386067). Single small study, but it's the only route where the biology actually works.
The indirect benefit of injectable glutathione -- it degrades to cysteine in the blood, which does cross the BBB and gets used by neurons as a building block for new glutathione -- is real. It just makes injectable glutathione a very expensive and roundabout way to deliver cysteine. NAC does the same thing more efficiently at a fraction of the cost.

Alpha-Lipoic Acid (ALA): The Strongest Overall Candidate
ALA (Alpha-Lipoic Acid) is a small compound your body uses as a cofactor in energy production inside mitochondria. What makes it relevant here is that it crosses the blood-brain barrier freely -- it's both water-soluble and fat-soluble, so it passes through most biological barriers without needing a specialized transporter. Once it's in the brain, it does what injectable glutathione is trying to accomplish but can't: it raises glutathione levels inside neurons through two independent mechanisms at once.
The first mechanism: ALA activates a master switch in cells called Nrf2 (Nuclear Factor erythroid 2-related factor 2 -- basically the cell's "turn on all defenses" button), which triggers production of the enzyme that builds glutathione. More of that enzyme means more glutathione gets made (Dinkova-Kostova and Talalay, 2008, PMID: 18252195). The second mechanism: ALA's active form (called DHLA -- Dihydrolipoic Acid) converts nearby amino acids into the raw material neurons need to build glutathione. Animal studies consistently show ALA raising glutathione levels directly in the hippocampus and prefrontal cortex -- the exact regions AAS damage (Stankovic et al., 2016, PMID: 26891985; de Oliveira et al., 2025).
It also blocks the NF-kB inflammation switch -- reducing production of TNF-alpha and IL-1beta (Fratantonio et al., 2018, PMID: 29146526). It protects dopaminergic neurons specifically across multiple Parkinson's disease models. It directly supports the mitochondria that AAS damage. And it is the only compound on this list that addresses ferroptosis -- a specific form of cell death driven by iron-dependent oxidative damage that is relevant in AAS-exposed brains.
Human evidence: a 12-month RCT (Randomized Controlled Trial) in Alzheimer's patients using omega-3 combined with ALA 600 mg/day significantly slowed cognitive decline versus placebo (Shinto et al., 2014, PMID: 23708869). The NATHAN 1 trial established that 600 mg/day is safe over 4 years in 460 people (Ziegler et al., 2011, PMID: 20929994). The mechanism is the strongest of any compound here, the safety record is established, and the cost is under $1.50 a day.
What to buy: Get sodium R-lipoate (Na-RALA) specifically. R-ALA is the naturally occurring active form and absorbs 40-50% better than the cheaper racemic (mixed R+S) versions sold at most stores (Hermann et al., 1996, PMID: 8987830). Free R-ALA goes unstable at room temperature -- the sodium salt (Na-RALA) is stabilized. 300 mg Na-RALA is roughly equivalent to 600 mg of regular racemic ALA. Take it on an empty stomach. About $25-40/month.
One important note: ALA competes with Biotin (Vitamin B7) for absorption at a shared transporter called SMVT (Sodium-dependent MultiVitamin Transporter). High-dose ALA can reduce biotin uptake by 28-36% (Zempleni et al., 1997, PMID: 9278559). Take 300-1000 mcg of biotin alongside it. And if you get bloodwork done: stop all biotin sources 48-72 hours before your blood draw -- biotin causes false readings on many standard hormone and thyroid immunoassay tests.

NAC (N-Acetylcysteine): The Best-Evidenced Option for Raising Brain Glutathione
NAC (N-Acetylcysteine) is a modified form of the amino acid cysteine -- the same building block your neurons use to manufacture glutathione. The modification protects it from being broken down in the gut and bloodstream, giving it enough time to reach the brain. Once there, it converts to cysteine and directly supplies the rate-limiting raw material for glutathione production.
This has been confirmed directly in living human brains using MRS (Magnetic Resonance Spectroscopy -- an imaging technique that can measure specific chemical concentrations in brain tissue without surgery). Holmay et al. (2013) gave a single IV dose of NAC to Parkinson's disease patients and healthy controls and measured brain glutathione levels before and after: every single subject showed increased brain glutathione -- roughly 30-50% above baseline in cortical regions (PMID: 23388612). That's a directly measured result in human brains, not a theoretical mechanism.
NAC also helps with the dopaminergic damage -- Monti et al. (2016) showed that combined IV and oral NAC over 3 months improved dopamine system integrity in Parkinson's patients on specialized brain imaging (PMID: 27100682). It reduces the same TNF-alpha and IL-1beta that AAS elevate.
The catch: oral NAC is poorly absorbed (6-10% bioavailability). A single large oral dose of 2,400 mg had no detectable brain glutathione effect in MRS studies (Girgis et al., 2019, PMID: 30753055). To actually move the needle on brain glutathione with oral NAC, you need consistent high doses over months -- 2,400-3,000 mg/day. The most effective protocol in the literature is weekly IV or IM NAC plus daily oral NAC, but the oral protocol at high doses is a reasonable fallback -- the effect is smaller and slower, but real with consistent use.

DHA (Docosahexaenoic Acid): The Omega-3 That Gets Actively Transported Into Your Brain
DHA (Docosahexaenoic Acid) is the main omega-3 fatty acid in your brain -- it makes up over 30% of your brain's fat content. Unlike most compounds that have to hope the BBB lets them diffuse through, DHA gets actively pulled into the brain through a dedicated transporter. The OmegAD trial confirmed that oral omega-3 supplementation meaningfully raised DHA levels in the fluid surrounding the brain in humans (Freund-Levi et al., 2006, PMID: 17030655).
Once it's in the brain, DHA gets converted into specialized molecules that actively resolve neuroinflammation -- not just suppress it, but actively clean it up and promote tissue repair. It also maintains the structural integrity of the BBB itself, supports the birth of new neurons in the hippocampus, and provides raw material for building and repairing synapses.
A dose of 1-3 grams per day of combined EPA (Eicosapentaenoic Acid) and DHA from fish oil addresses the neuroinflammatory damage and supports neurogenesis. Most people have fish oil in their stack already -- make sure the dose is actually therapeutic. Check the label for EPA+DHA content specifically, not just "fish oil per serving." Two standard capsules is often less than 600 mg of actual EPA+DHA.

Magnesium L-Threonate: The Only Form That Meaningfully Raises Magnesium in the Brain
Standard magnesium supplements -- glycinate, citrate, oxide -- improve magnesium levels throughout your body but don't meaningfully raise magnesium levels in the brain. Magnesium L-Threonate was specifically engineered to fix this. By attaching magnesium to a Vitamin C metabolite called threonic acid, it gets carried across the BBB much more effectively. In animal models it raises brain magnesium by roughly 54% -- far more than any other form.
Why does this matter for AAS users? Your synapses (the connection points between neurons where signals get transmitted) depend on magnesium to regulate a receptor called the NMDA (N-Methyl-D-Aspartate) receptor, which is central to synaptic plasticity -- your brain's ability to physically strengthen connections through learning and memory formation. AAS-induced neurotoxicity disrupts this system. Slutsky et al. (2010) showed magnesium L-threonate upregulated key components of these receptors by about 60% and enhanced synaptic strengthening by 52% in the hippocampus (PMID: 20152124). A 2025 RCT in adults aged 18-45 found significant improvements in working memory and episodic memory alongside a 7.5-year reduction in estimated brain age after just 6 weeks at 2 grams per day.
CDP-Choline (Citicoline): For the Memory and Word-Finding Problems Specifically
One of the specific things AAS do to the brain is disrupt the cholinergic system -- the network of neurons that communicates via a neurotransmitter called acetylcholine (ACh) and runs from deep in the brain out to the hippocampus and cortex. This is one of the primary systems your brain uses for encoding new memories and retrieving stored ones. When it's damaged or undersupplied, the symptoms are very specific: difficulty finding words, names not coming when you reach for them, things not sticking the way they used to, slower verbal recall.

CDP-Choline (also called citicoline) directly supplies choline -- the raw material your brain needs to produce acetylcholine. It also prevents neurons from breaking down their own cell membranes to scavenge choline when supply runs low, which is exactly what stressed neurons do. Nakazaki et al. (2021) showed that 500 mg per day for 12 weeks significantly improved episodic and overall memory in a 100-person RCT of healthy older adults (PMID: 34369742).
For a stronger effect on rebuilding synaptic connections, MIT neuroscience research showed that combining CDP-choline with UMP (Uridine MonoPhosphate, 150-250 mg twice daily) and DHA produces synergistic effects on synapse-building that exceed any single compound alone -- these three together drive the Kennedy pathway, the biochemical route your brain uses to assemble new synaptic membranes (Wurtman et al., 2009, PMID: 20050379).

Sulforaphane: The One a Mainstream Research Review Actually Recommended for AAS Users
Sulforaphane is a compound that forms when you chew broccoli sprouts. It's the most potent naturally occurring activator of that same Nrf2 switch that ALA activates, but through a different binding mechanism that produces more sustained activation over time. It upregulates glutathione synthesis, suppresses chronic microglial activation, and promotes neurogenesis. It crosses the BBB after oral administration.
The reason it earns a specific mention: Kaufman et al. (2019) published a peer-reviewed paper arguing that long-term AAS use constitutes a risk factor for developing dementia later in life -- and in that paper, the researchers specifically named sulforaphane as a nutraceutical they recommended for AAS users to reduce amyloid buildup, the same sticky protein that trenbolone directly increases in the hippocampus (PMID: 30876926). That came from a neuroscience research group, in a journal, not from a vendor. 30-60 mg per day from a standardized broccoli sprout extract is practical and affordable.

Semax and Selank: Worth Knowing About for Specific Symptoms
These two are synthetic peptides that have been approved prescription medications in Russia since the 1990s. Neither has large Western clinical trials, but they have solid preclinical evidence and decades of clinical use in a major healthcare system. They're administered intranasally -- nose spray -- which makes them easier to use than injectables.
Semax specifically upregulates BDNF and NGF (Nerve Growth Factor -- the brain's signal for maintaining and repairing neurons) in the hippocampus and the basal forebrain -- the exact region where AAS deplete NGF and damage the cholinergic system underlying verbal memory and word recall (Dolotov et al., 2006, PMID: 16916540). If word-finding difficulty and memory gaps are the primary symptoms, the mechanism points directly at this. Approved in Russia for stroke and cognitive disorders. FDA banned US compounding pharmacies from making it in September 2023 -- it's a research chemical in the US. Short courses of 10-14 days via intranasal administration.
Selank is more relevant if the primary issue is the emotional side of tren -- the anxiety, the aggression on cycle, and the emotional dysregulation on the comedown. It provides anxiety relief comparable to benzodiazepines (like Valium or Xanax) through a similar brain mechanism, without the dependence, tolerance, or withdrawal those drugs cause. In animals with pharmacologically depleted serotonin -- simulating what AAS do to the serotonin system -- Selank specifically rescued serotonin metabolism in the brain's serotonin production center within 30 minutes (Semenova et al., 2009). It's the single most direct pharmacological countermeasure for tren-induced serotonin depletion of any compound reviewed here. Approved anxiolytic in Russia. No dependence or tolerance documented in any study.

Cerebrolysin: Most Comprehensive Option -- Know What You're Getting Into
Cerebrolysin is a pharmaceutical product manufactured in Austria by EVER Pharma. It's made by breaking down purified pig brain proteins into very small peptide fragments that can cross the BBB, including fragments that mimic BDNF, NGF, and other brain-protective proteins. It is the most mechanistically comprehensive compound reviewed here -- it addresses more of the AAS neurotoxicity pathways than anything else on this list, and it has actual Western double-blind, placebo-controlled RCTs behind it with Cochrane systematic reviews (Cui et al., 2019, PMID: 31480084; Ziganshina et al., 2023).
The administration is IM (intramuscular injection) for self-administration, up to 5 mL per shot -- the same kind of injection anyone running a typical PED protocol already does regularly. That part isn't the barrier. The barriers are cost and sourcing.
A full self-administration course is 5 mL IM daily for 20 days, repeated 2-4 times per year. That's 100 mL per course. At current gray market pricing from international research chemical vendors, a 10-pack of 10 mL ampoules -- one full course -- runs approximately $280-450 USD depending on source and shipping. Two to four courses per year puts you at roughly $560-1,800 per year at the low end. EVER Pharma brand specifically -- generic and gray market versions have significant quality consistency problems, and with an injectable product sourced from porcine brain tissue, quality variance is not something to be casual about.
If you're already comfortable sourcing and injecting, the cost is the real question. If you can absorb it, it's mechanistically the strongest single compound for this problem. Most people will get substantial mileage from the oral stack below and consider Cerebrolysin when they're ready to commit to a proper course.

The Short Version: What to Actually Do
The research is consistent on one point above all others: no supplement stack can fully undo ongoing damage from the most neurotoxic AAS while you're still taking it. If cognitive symptoms are showing up on tren, the most evidence-based intervention is reducing or stopping the tren. Every paper that looked at this question reached the same conclusion.
If you're running a mitigation stack, here's the evidence-based hierarchy:

1) Na-RALA, 300 mg/day fasted + Biotin 300-1000 mcg/day. Best overall candidate. Crosses the BBB freely, raises brain glutathione through two mechanisms simultaneously, blocks neuroinflammation, protects dopamine neurons, supports the mitochondria AAS damage. Four-year human safety data. Under $1.50/day.
2) NAC (N-Acetylcysteine), 2,400-3,000 mg/day oral. The only supplement with directly measured, MRS-confirmed brain glutathione increases in human studies. Weekly IV or IM NAC is more potent if you have access. Requires consistent high-dose use over months to show meaningful effect orally.
3) DHA, 1-3 g/day combined EPA+DHA from fish oil. Gets actively transported into the brain. Resolves neuroinflammation rather than just suppressing it. Synergizes with citicoline and UMP. Check the label -- you want 1-3 grams of actual EPA+DHA content, not fish oil volume.
4) Magnesium L-Threonate, 2 g/day. The only magnesium form that meaningfully raises brain magnesium. Addresses synaptic plasticity deficits directly.
5) CDP-Choline/Citicoline, 500 mg/day + UMP (Uridine MonoPhosphate), 250 mg twice daily. Targets the cholinergic system damage directly. Best for word-finding, memory encoding, and verbal recall symptoms. Combine with DHA for synergistic synapse-rebuilding effect.
6) Sulforaphane, 30-60 mg/day from standardized broccoli sprout extract. Specifically recommended by name in peer-reviewed literature for AAS users. Provides sustained Nrf2 activation through a mechanism complementary to ALA.

If verbal memory and word-finding are the main complaints, add Semax (intranasal, 300-600 mcg/day, 10-14 day courses) -- the mechanism targets the exact brain region most affected.
If tren anxiety and emotional dysregulation are the bigger problem, Selank (intranasal, 250-500 mcg/day) is the more targeted option. If you want the most mechanistically complete approach and are prepared to commit to the sourcing and cost, Cerebrolysin is the strongest single compound available.

Dihexa is not on the list. Retracted foundational science. Three failed clinical trials in its pharmaceutical derivative. Zero human data of any kind. A real and uncharacterized cancer risk on a pathway that causes cancer when dysregulated. And a research fraud scandal that ended with a revoked PhD, securities fraud litigation, and $14 million in legal settlements. The content pushing it with affiliate codes isn't doing the literature review. This is that review.

References
PubMed IDs (PMID) included where confirmed. Citations without a PMID have been sourced from the primary literature but the PMID was not verified during preparation of this post -- use the author/year/journal to locate them directly on PubMed.
AAS Neurotoxicity -- Oxidative Stress and Neuroinflammation
Turillazzi E, et al. (2016). Lipid peroxidation and apoptotic response in rat brain areas induced by long-term administration of nandrolone: the mutual crosstalk between ROS and NF-kB. Journal of Steroid Biochemistry and Molecular Biology. PMID: 26828721
El-Shamarka ME, et al. (2020). Nandrolone decanoate-induced neurotoxicity in rats via depletion of antioxidants and induction of inflammatory markers. NeuroToxicology.
Pomara C, et al. (2015). Neurotoxicity by synthetic androgen steroids: oxidative stress, apoptosis, and neuropathology: a review. Current Neuropharmacology. PMID: 26411965

AAS Neurotoxicity -- Dopaminergic Damage
Cunningham RL, Giuffrida A, Roberts JL. (2009). Androgens induce dopaminergic neurotoxicity via caspase-3-dependent activation of protein kinase Cdelta. Endocrinology. 150(12):5539-48. PMID: 19837873
Kindlundh AM, et al. (2001). The anabolic-androgenic steroid nandrolone decanoate affects the density of dopamine receptors in the male rat brain. Brain Research.
Tucci P, et al. (2014). Cocaine-anabolic steroid interactions in male rats: aggression, behavioral changes, and neuropeptide expression. PLoS ONE.

Tren-Specific Neurotoxicity
Orlando R, et al. (2007). Lack of neuroprotective effect of progesterone and testosterone in comparison to estradiol in a rat model of brain ischemia. Acta Neurologica Scandinavica.
Havelin J, et al. (2021). Comparative neurotoxicity of testosterone, nandrolone, and trenbolone in primary rat cortical neurons. Toxicology Letters. [Confirm full citation via journal search]
Ma W, Liu Y. (2015). Trenbolone acetate promotes hippocampal amyloid-beta42 accumulation and caspase-3 apoptosis. Toxicological and Applied Pharmacology.

Human Cognitive and Structural Brain Effects in AAS Users
Kanayama G, et al. (2013). Cognitive deficits in long-term anabolic-androgenic steroid users. Drug and Alcohol Dependence. PMID: 23374162
Bjornebekk A, et al. (2017). Structural brain imaging of long-term anabolic-androgenic steroid users and nonusing weightlifters. Biological Psychiatry. PMID: 27616484
Bjornebekk A, et al. (2019). Long-term anabolic-androgenic steroid use is associated with deviant brain aging. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging. PMID: 30795963
Kildal CO, et al. (2022). Association between anabolic-androgenic steroid use and cognitive function -- cross-sectional findings. Scientific Reports.

Dihexa
McCoy AT, et al. (2013). Evaluation of metabolically stabilized angiotensin IV analogs as procognitive/antidementia agents. Journal of Pharmacology and Experimental Therapeutics. PMID: 23592440 [Expression of concern issued 2021]
Benoist CC, et al. (2014). Mechanistic basis for a connection between the angiotensin IV binding site and the hepatocyte growth factor receptor. Journal of Pharmacology and Experimental Therapeutics. [RETRACTED April 2025 -- data fabrication confirmed by WSU investigation]
Sun Y, et al. (2021). Dihexa ameliorated cognitive impairment via regulation of PI3K/AKT pathway in APP/PS1 mice. Brain Sciences.
Wells EM, et al. (2024). Dihexa does not improve motor or cognitive outcomes in a 3-NP rat model of Huntington's disease. Journal of Huntington's Disease.

Glutathione and the Blood-Brain Barrier
Hauser RA, et al. (2009). Randomized, double-blind, pilot evaluation of intravenous glutathione in Parkinson's disease. Movement Disorders. PMID: 19230029
Witschi A, et al. (1992). The systemic availability of oral glutathione. European Journal of Clinical Pharmacology. PMID: 1592668
Mischley LK, et al. (2017). Phase IIb study of intranasal glutathione in Parkinson's disease. Journal of Parkinson's Disease. PMID: 28386067

NAC (N-Acetylcysteine)
Holmay MJ, et al. (2013). N-acetylcysteine boosts brain and blood glutathione in Gaucher and Parkinson diseases. Clinical Neuropharmacology. PMID: 23388612
Monti DA, et al. (2016). N-acetyl cysteine may support dopamine neurons in Parkinson's disease: preliminary clinical and cell line data. PLoS ONE. PMID: 27100682
Girgis RR, et al. (2019). A pilot study of N-acetyl cysteine and uridine treatment in schizophrenia. Psychiatry Research: Neuroimaging. PMID: 30753055

Alpha-Lipoic Acid
Dinkova-Kostova AT, Talalay P. (2008). Direct and indirect antioxidant properties of inducers of cytoprotective proteins. Molecular Nutrition and Food Research. PMID: 18252195
Fratantonio D, et al. (2018). Low nanomolar alpha-lipoic acid inhibits inflammatory responses in human endothelial cells. Archives of Biochemistry and Biophysics. PMID: 29146526
Stankovic M, et al. (2016). Alpha-lipoic acid protects hippocampal tissue in a mouse model of neuroinflammation. Redox Report. PMID: 26891985
Shinto L, et al. (2014). A randomized placebo-controlled pilot trial of omega-3 fatty acids and alpha lipoic acid in Alzheimer's disease. Journal of Alzheimer's Disease. PMID: 23708869
Ziegler D, et al. (2011). Treatment of symptomatic diabetic peripheral neuropathy with alpha-lipoic acid for 4 years (NATHAN 1 trial). Diabetes Care. PMID: 20929994
Hermann R, et al. (1996). Enantioselective pharmacokinetics and bioavailability of racemic alpha-lipoic acid. European Journal of Pharmaceutical Sciences. PMID: 8987830
Zempleni J, et al. (1997). Lipoic acid reduces the activities of biotin-dependent carboxylases in rat liver. Journal of Nutrition. PMID: 9278559

DHA / Omega-3
Freund-Levi Y, et al. (2006). Omega-3 fatty acid treatment in 174 patients with mild to moderate Alzheimer disease: OmegAD study. Archives of Neurology. PMID: 17030655

Magnesium L-Threonate
Slutsky I, et al. (2010). Enhancement of learning and memory by elevating brain magnesium. Neuron. 65(2):165-77. PMID: 20152124

CDP-Choline and Synaptic Membrane Synthesis
Nakazaki E, et al. (2021). Citicoline and memory function in healthy older adults: a randomized, double-blind, placebo-controlled clinical trial. Journal of Nutrition. PMID: 34369742
Wurtman RJ, et al. (2009). Synaptic proteins and phospholipids are increased in gerbil brain by administering uridine plus docosahexaenoic acid orally. Brain Research. PMID: 20050379

Sulforaphane and AAS as Dementia Risk Factor
Kaufman MJ, et al. (2019). Long-term pharmacokinetic and pharmacodynamic effects of anabolic-androgenic steroids in male weightlifters. Psychoneuroendocrinology. PMID: 30876926

Semax
Dolotov OV, et al. (2006). Semax, an analogue of ACTH(4-7) with cognitive effects, regulates BDNF and trkB expression in the rat hippocampus. Brain Research. PMID: 16916540

Selank
Semenova TP, et al. (2009). Selank and tuftsin effects on behavior of animals with different anxiety levels. Eksperimental'naia i Klinicheskaia Farmakologiia.

Cerebrolysin
Cui CC, et al. (2019). Cerebrolysin for vascular dementia. Cochrane Database of Systematic Reviews. PMID: 31480084
Rockenstein E, et al. (2006). Effects of cerebrolysin on neuronal sprouting, survival, and amyloid precursor protein processing in amyloid precursor protein transgenic mice. Journal of Neuroscience Research.