Advanced Topic: Dr. Davis's Private Notes on Preventing/Reversing Cognitive Decline/Dementia


When I was part of a team working in a facility in New Jersey in which residents were to be offered a program to preserve or reverse cognitive decline and dementia, I extensively reviewed the body of existing evidence to get a better handle on what has been shown to work, what has not, reproduced here.

Unfortunately, mainstream conversations on this topic, as cultivated by the work of Bredesen et al, has fallen victim to over-the-top enthusiasm for many strategies that have little or no evidence of efficacy. I therefore regard those conversations as topics of conversation, starting points, but far from conclusive advice. 

Obviously the rapidly changing nature of the evidence means that my notes will need to be updated frequently. Given the length of this information, it is better used as a reference rather than reading material to be read top to bottom. 

The Scientific Evidence Applied to Developing the ReVite Program for Cognitive Preservation/Reversal of Cognitive Decline and Early Dementia

This is an evolving summary of the rapidly-changing evidence we are examining (as well as generating based on our emerging experience) to serve as the foundation for our cognitive preservation/dementia-reversing program.

I’ve highlighted the agents/strategies with at least some genuinely supportive, higher-quality data that either have adequate supportive evidence or appear to be headed in that direction.

Not addressed to date are the combinations of strategies that may yield effects more effective than single strategies in isolation.

Nutritional supplements


Carnitines, of which the acetyl-L-carnitine form is best studied, have been proposed, sometimes alongside alpha-lipoic acid, as a therapy to prevent cognitive decline because of its role in mitochondrial energy generation. Acetyl-L-carnitine is mentioned in passing by Bredesen and only in the context of case studies with no explanation or justification.

How do we reconcile the potential benefits of various carnitine derivatives with the atherogenic potential of carnitine as proposed by the TMAO data from Hazen et al? I believe this is yet another reason to hold the TMAO data suspect (considered separately).

Experimental evidence:

Summary of the experimental evidence suggesting reversal of age-related mitochondrial ultrastructural decay by ALA and acetyl-carnitine:

Acetyl-L-carnitine abolished tau hyperphosphorylation, beta-amyloid accumulation, and cognitive impairment in rats with memory deficits induce in a hyperhomocysteinemia model:

Observational evidence:

In the continuum from normal cognition (n =46), to subjective memory complaints (n = 24), to mild cognitive impairment (n = 18), to AD (MMSE <24; n = 29), serum levels of acetyl-L-carnitine and other acyl-carnitines (e.g. malonyl-, hexenoyl, decanoyl-, myristoyl- and others) all decreased with progressive degrees of cognitive impairment, the dose-response suggesting, but incapable of proving, a cause-effect association. Age was not controlled, a major oversight, as there was also an age-related trend to the cognitive data (controls youngest, subjective memory complaint and MCI group younger, AD oldest), raising the question of an age-related, non-cognitive related association.

People with AD have lower plasma levels of medium-chain acyl-carnitines that correlated with lower prefrontal gray matter volumes and cognitive impairment, as well as lower plasma levels of 2-hydroxybutyric acid (as an index of ketogenesis):

CSF levels of l-carnitine were lower in non-apo E4 carriers in early onset AD that correlated with CSF amyloid-beta levels and MMSE score:

Clinical treatment data:

People with AD (n = 130) were randomized to acetyl-L-carnitine vs. placebo over one year. Both groups showed deterioration in cognitive measures, with less deterioration in the treated group:

A small trial (n = 20) of 1000 mg acetyl-L-carnitine twice per day vs. placebo over 24 weeks showing a non-significant trend towards less deterioration in the treated group:

People with AD were randomized to acetyl-L-carnitine 1000 mg three times per day (n = 112) vs. placebo (n = 117) and followed for 12 months. Of the several cognitive measures used, only MMSE showed less decline in the treated group, specifically reduced decline in attention:

People with AD were treated with 1000 mg acetyl-L-carnitine three times per day vs. placebo (total n = 431) with no difference in rate of decline over one year. Post hoc analysis of those less than 65 years old may have experienced slower decline:

A small proprietary study in which 106 participants with AD were randomized to a combination of folate, alpha-tocopherol, B12, s-adenosyl methionine, N-acetyl cysteine, and acetyl-L-carnitine vs. placebo for 6 months, followed by an open-label 6-month extension showed improvement in cognitive measures in the treated arm and during the open-label extension:

A cocktail of folate, vitamin B12, alpha-tocopherol, S-adenosyl methionine, N-acetyl cysteine, and acetyl-L-carnitine slowed cognitive decline over 9 months in 12 participants randomized to treatment or placebo with moderate to advanced dementia:

People with mild-moderate AD (n = 30) were randomized to acetyl-L-carnitine 2500 mg per day for 3 months, then 3000 mg per day for 3 more months vs. placebo. The treated group showed less deterioration in several cognitive measures (Digit Span and verbal fluency):

Cochrane review of the 11 clinical trials of acetyl-l-carnitine showing evidence for benefit on “clinical global impression” but none on objective cognitive measures:


The evidence in total suggest that acetyl-L-carnitine may have modest benefits in slowing cognitive decline at a dose of 3000 mg per day, though many of the positive findings come via relatively tortured post hoc assessments. The data combining acetyl-L-carnitine with alpha-lipoic acid is low in quality and it remains unclear whether the two should be used in combination.

Once again, it is impossible to conclude from the data whether acetyl-L-carnitine simply exerts a nootropic or a neurotrophic effect, as no neuroimaging or long-term follow-up studies have been conducted.

As a personal aside, I have taken acetyl-L-carnitine and did indeed notice a modest increase in energy and focus, but it was short-lived with subsequent doses yielding less and less effect, suggesting a tolerance/tachyphylactic effect. Does this apply to its cognition-enhancing effect, also?

Alpha-linolenic Acid


Summary of experimental data:

ALA, but not linoleic acid, reduced oxidative injury and increased BDNF in a mouse model:

Observational data:

The observational data are fairly consistent in associating low ALA serum levels with increased risk for cognitive decline/dementia.

Increased serum levels of the omega-3 fatty acids, alpha-linolenic acid, stearidonic acid, and eicosatrienoic acid, were associated with greater "fluid intelligence," left frontoparietal volume and total gray matter volume:

Serum ALA levels were inversely associated with progression to dementia:

Higher dietary intakes of ALA were associated with slower cognitive decline but only in apo E4 people:

Total omega-3 and linolenic acid plasma levels were lower in people with mild cognitive decline and dementia compared to controls:

Post-mortem studies of human brains demonstrated dementia neuropathology was inversely associated with seafood consumption (lower density neuritic plaques and neurofibrillary tangles), higher levels of ALA were associated with fewer cerebral microinfarctions, and level of mercury had no association:

In the Rotterdam Study, EPA/DHA and linolenic acid intake by food questionnaire was not correlated with cognitive decline over nearly 10 years:

Human clinical trials:

There is a lack of human clinical treatment trials, with data limited to two studies.

ALA "supplemented" as 1500 mg per day yielded 40% increase in serum BDNF levels in humans:

Supplementation of 384 mg of EPA-DHA, 1.9 g of ALA, or both vs. placebo per day over 40 months in coronary patients yielded no slowing of cognitive decline:


Despite the lack of clinical human treatment data, given the strongly suggestive observational and experimental data, as well as the fact that linolenic acid is an essential fatty acid, purposeful inclusion of ALA-rich foods that can provide 2000 mg or more per day of ALA seems like the best route without need for supplementation. (The Adequate Intake for ALA is 1.1 grams per day for adult females, 1.6 grams per day for adult males.)

Small servings of flaxseed, chia, and walnuts provide several-fold more than 2000 mg, while grass-fed beef provides about 40% more ALA than non-grass fed beef. While pigs do not graze on grass and there is no such designation for pork, pork tends to have 3-fold higher ALA levels than beef or chicken.

There are also benefits to ALA outside of cognitive preservation, such as a modest reduction in cardiovascular risk. (The negative reports associating ALA with increased prostate cancer risk have been debunked/refuted.)

Alpha-lipoic Acid

The anti-oxidative mitochondrial nutrient, alpha-lipoic acid, or ALA, has been studied for its cognitive-preserving effects, though often in combination with acetyl-carnitine and other presumptive mitochondrial-influencing nutrients. ALA is naturally occurring cofactor for the mitochondrial enzymes pyruvate dehydrogenase and alpha-ketoglutarate dehydrogenase of the Kreb’s cycle.

Note that there are two enantiomers: R- and L-ALA, with the dextrorotatory form exerting biological effects, not the levorotatory form. In all data, the racemic mixture of ALA was used.

Summary of the experimental evidence suggesting reversal of age-related mitochondrial ultrastructural decay by ALA and acetyl-carnitine:

Experimental evidence:

ALA administered in a rat model of vascular dementia (bilateral carotid occlusion) resulted in less cognitive impairment, decreased production of reactive oxidative species, and increased hippocampal levels of reduced glutathione:

ALA administered in a mouse model of AD exerted an insulin-mimetic effect that increased brain glucose uptake, activation of the insulin receptor substrate and improved synaptic plasticity:

Clinical treatment data:

A pilot study 39 people with AD were randomized to omega-3 fatty acids (675 mg DHA + 975 mg EPA per day), omega-3 (675 mg DHA + 975 mg EPA per day)+ ALA (600 mg/day), or placebo and tracked with cognitive measures (ADAS-Cog, MMSE, ADL/IADL) over 12 months. Combination treatment slowed cognitive and functional decline (MMSE and IADL) over placebo:

126 people with AD (MMSE 12-26) with (n = 61) or without (n = 65) type 2 diabetes were given ALA (uncontrolled) 600 mg/day in combination with conventional dementia treatment over 16 months. Participants with DM given ALA showed significant slowing of cognitive measures, with 43% of participants with DM showing improved MMSE compared to non-DM given ALA:

53 people with Down’s syndrome were randomized to a combination of alpha-tocopherol 900 IUs, ascorbic acid 200 mg, and ALA 600 mg per day vs. placebo over two years with no improvement in cognitive measures nor a slowing of decline:


While limited, the existing data suggest a potential cognition-preserving effect, though we cannot say whether it is a nootropic or neurotrophic effect based on the human clinical data. The experimental data make it tempting to believe that there is indeed a neurotrophic effect but, until we have confirmatory evidence such as volumetric MRI or improved psychometric measures after stopping treatment, we can only presume that the effect is limited to a modest nootropic effect.


Ashwaghanda is listed as a basic supplement in ReCODE as an agent to stop/reverse neurodegenerative disease. Unfortunately, there is no rationale or science listed in the program nor the book.

The evidence for ashwaghanda:

A summary of the experimental model evidence in mice and in vitro, which is varied and compelling (on the surface):

If we believe the experimental data, ashwaghanda can:

Reverse numerous forms of cancer

Protect the brain from stroke

Reverse neurodegenerative damage

Protect from aspergillus infection

Increase sexual interest/function, i.e., aphrodisiac properties

Subdues mania in bipolar illness

Improves insulin resistance

Normalizes cortisol responses

Protects neurons against changes associated with Parkinson’s disease

Protects against tooth decay by degrading the biofilm produced by Streptococcus mutans

Reverses macular degeneration

Reduces pulmonary hypertension

Immunomodulatory activity including suppression of anaphylactic reactions

Exert antifungal effects against a variety of pathogenic fungi

Substantially improves cardiovascular endurance and generates substantially greater muscle growth with strength training

Yields reduced pain and inflammatory markers in rheumatoid arthritis

Has a nootropic effect

Reduces anxiety

Yield beta-blocking and ACE-inhibitor effects

Reduces alcohol craving in experimental alcoholism

Reduces blood sugar and insulin

Exerts anti-hyperlipidemic effects

Enhances efficacy of radiation therapy for cancer

Human data:

Reverses fatigue and improves survival in chemotherapy-treated breast cancer:

Reduces symptoms of obsessive-compulsive disorder:

Enhanced sexual function in females:

Improved cardiorespiratory performance:

Reduced joint pain, swelling, and inflammatory markers in rheumatoid arthritis:

Improved cognitive measures in bipolar illness:

No data on dementia or cognitive decline.

Problem: Virtually ALL experimental and human data were generated in India, where most of the world’s ashwaghanda is grown.

Hmmm. Do we believe the (too-good-to-be-true experimental and clinical) data, nearly all of it generated by potentially biased sources?


Bacopa monnieri

Bacopa is an Ayurvedic preparation with a history dating back over 1000 years with long-purported benefits of improved cognition, in addition to a list of other purported effects that include anti-inflammatory, anti-convulsant, analgesic, anti-microbial (a red flag in my view, as agents such as berberine that also has antimicrobial effects, there is potential for altering bowel flora in uncertain ways), neuroprotective, anti-oxidative and others.

Experimental evidence:

The experimental evidence is plentiful and robust, in total suggesting numerous plausible means of exerting a brain-protective effect.

The active component is uncertain but likely resides in the dozen “bacosides” that occur along with nicotine, several flavonoids, and others. This is a summary of the experimental evidence including inhibition of beta-amyloid deposition, acetylcholinesterase inhibition, choline acetyltransferase activation, monoamine enhancement, and increased cerebral blood flow, including that involving hippocampal tissue:

Effects appear most prominent in enhancing dopamine and serotonin (nootropic effects), but no effect on norepinephrine in cerebral tissue. Increased dendritic branching and interconnection has been observed in brain preparations (suggesting a neurotrophic effect).

Clinical treatment data:

Bacopa as part of a multi-component treatment (Bacopa monnieri, L-theanine, Crocus sativus, copper, folate, B-complex, and vitamin D) vs. placebo on cognitive measures in patients with dementia with MMSE of 20-27, n = 30, suggesting improved cognition with treatment:

Extracts of Bacopa monnieri (whole plant), Hippophae rhamnoides (leaves and fruits), and Dioscorea bulbifera (bulbils), 500 mg per day vs. donepezil 10 mg per day vs. placebo in a randomized, controlled trial over 12 months in both cognitively normal (n = 97) and people with AD (n= 103) showed improved cognitive measures and measures of oxidation and inflammation in the plant extract group compared to both donepezil and placebo groups:

(Though no commercial affiliations/funding was disclosed, this study felt like something that was very commercial.)

A combination of bacopa (100 mg extract containing 20 mg bacosides), phosphatidylserine (soy-derived; 30 mg); vitamin E (form unspecified), astaxanthin 2 mg, and micro algae dry extract (74 mg), uncontrolled, improved cognitive measures in people (n = 102) with mild cognitive impairment, MMSE 22-28, over 3 months:

Bacopa extract 125 mg twice per day (55% bacoside content) vs. placebo in people (n = 35) over age 55 with memory complaints but MMSE >24 suggested improved cognition with treatment over 12 weeks:

Randomized trial bacopa extract 150 mg vs. placebo in 60 cognitively normal medical students over 2 weeks suggested improved cognitive measures (nootropic effect), especially attention and memory, that persisted for 15 days after the intervention (suggested a durable neurotrophic effect):

Proprietary randomized, controlled study of bacopa extract 320 mg or 640 mg vs. placebo on cognitive measures in healthy volunteers (n = 17) demonstrating acutely improved cognitive measures, mood, and reduced salivary cortisol within 2 hours of administration:

Bacopa extract 300 mg or 600 mg vs placebo in healthy elderly (n = 60) over 12 weeks improved memory, attention, and cognition:

Bacopa extract 300 mg per day vs. placebo in cognitively normal people (n = 98) over 12 weeks improved measures of memory:

Bacopa extract 300 mg per day vs. placebo in cognitively normal people (n = 54) over 12 weeks improved measures of memory:

Meta-analysis of 12 prospective, randomized trials (n = 437) of bacopa on cognitive testing of at least 12 weeks duration suggests that cognitive improvements are experienced:

Curiously, studies cited and labeled “unbiased” or “good quality” I found to be potentially biased (e.g., branded proprietary products used) or low quality, especially with small numbers of participants, uncontrolled, or including a smorgasbord of other treatment components. This suggests that this meta-analysis (among several others I found) was likely biased.


It’s the same familiar theme: Plentiful and compelling experimental data suggesting plausible biological neuroprotective mechanisms, but small, sloppy clinical studies, purported by some to be sufficient to prove cause-effect, including what appear to be biased meta-analyses. All studies are short-term, do not involve any imaging measure, and do not look at long-term outcomes.

I think one conclusion can be made with moderate confidence: Bacopa is likely an effective nootropic agent, i.e., enhances cognition during the period of administration. But there are virtually no data beyond that in experimental preparations to make us believe that there is any neurotrophic effect, i.e., no durable effect upon removal of treatment, no volumetric MRI or other imaging measure of disease, no change in long-term outcomes.  

We are back to the same question we have with piracetam, vinpocetine, acetylcholinesterase inhibitors, dimethylaminoethanol, and other nootropics: Is demonstration of a nootropic effect enough for us to consider inclusion in a program that purports to stall/reverse neurotrophic effects and cognitive decline? I don’t think it is. Note that there has been NO head-to-head comparison of the various nootropics to help us make sense of superiority of one agent or another, either.


Citicoline, 250 mg twice per day, is listed as one of the supplements in ReCODE without justification except to say that it “supports synaptic growth and maintenance.” Citicoline (cytidine-5′-diphosphate choline) is a precursor for the neurotransmitter, acetylcholine, and the phospholipid, phosphatidylcholine, and is the most prescribed agent for dementia in parts of Europe.


There are a number of other acetylcholine precursors, such as dimethylaminoethanol and choline, but there is experimental evidence to suggest that citicoline is also involved in phospholipid metabolism to synthesize phosphatidylcholine essential for brain function. Choline toxicity (acetylcholine effects such as tachycardia and gastrointestinal hyperactivity) are much less prominent compared to other acetylcholine precursors but for unclear reasons.


Experimental data:

Experimental data suggest that citicoline supports neuroplasticity, anti-apoptotic, SIRT1-increasing, cardiolipin-sparing, cerebral ischemia-protective effects, summarized here:


Clinical treatment studies:

There are a number of studies that examine citicoline’s nootropic effects, such as:

Citocoline 2000 mg per day improved verbal memory in people with “inefficient memories” (n = 32) vs. placebo over 60 days:


The Citicholinage Study was a retrospective case-control study in which combined therapy with an (unspecified) acetylcholinesterase inhibitor and citicoline 1000 mg per day or placebo (n = 448) over 9 months in people with Alzheimer’s dementia was assessed. Improved cognitive measures were seen at 3 months, no change at 9 months. (Oddly, study authors claimed that the combination slowed disease progression, not recognizing the difference between a nootropic and a neurotrophic effect.)


The CITIRIVAD Study used the same retrospective case-control design of Citicholinage (above) but specifically compared the acetylcholinesterase inhibitor, rivastigmine, with or without citicoline 1000 mg per day in people over age 65 with Alzheimer’s or mixed dementia (n = 174). Combined treatment was superior in maintaining cognitive measures over 9 months compared to rivastigmine alone. Once again, the study authors appeared not to understand the distinction between nootropic and neurotrophic effects, making claims about the slowing of disease progression.


163 people were enrolled 6 weeks following ischemic stroke in an open-label, randomized, parallel study of citicoline vs. usual treatment observed over 2 years with citicoline-treated participants demonstrating less cognitive decline:


A summary of the negative outcomes of the two prospective clinical trials in ischemic stroke, the Citicoline Brain Injury Treatment Trial (COBRIT; n = 1213) and the International Citicoline Trial on Acute Stroke (ICTUS; n = 2298), are summarized here:



The unexpectedly unsophisticated literature on citicoline fails to distinguish nootropic from neurotrophic effects, much like the literature supporting use of prescription acetylcholinesterase inhibitors. Note that there are no data demonstrating prevention or reduction in risk for cognitive impairment, only improvement in cognitive testing in established dementia in studies of low quality. At best, citicoline potentiates the nootropic/cognitive benefits of acetylcholinesterase inhibitors. The science to include citicoline is weak.



Coenzyme Q10/Ubiquinol

Coenzyme Q10/ubiquinol is included in the mix of supplements in the ReCODE program. While positive effects have been demonstrated in animal and limited human trials at high doses for Parkinsonism, ALS, and Huntinton’s disease, the only data on cognitive decline/dementia are mouse data and surrogate measures in humans:

Less hippocampal atrophy with mega-dose CoQ10 in mice:

Reduced amyloid plaque deposition in mice:

Reduced amyloid deposition in mice with the presenilin 1-L235P SNP:

Serum CoQ10 levels are inversely associated with dementia in humans (thereby suggesting association, not necessarily causation):

Lower serum CoQ10 levels in humans with Lewy body dementia:

A mix of antioxidants (vit C, E, alpha lipoic acid, and CoQ10 1200 mg) was associated with accelerated cognitive decline in humans:

So there are no good data suggesting that CoQ10 (with virtually no data on ubiquinol) has efficacy in cognitive decline. There are data, on the other hand, of CoQ10’s effectiveness in reducing blood pressure and improving left ventricular function in heart failure/dilated cardiomyopathy, as well as reversing muscle pain/weakness on statin drugs. And it is benign, though expensive.


Despite the popularity of this supplement, probably the most widely-ingested nutritional supplement for preservation of cognitive health, the sum total of data over the years have demonstrated little if any cognitive benefit in normal people and in those with cognitive decline. Even though it is not included in the DeCODE program, people may be taking ginkgo, believing it may be of benefit. An awareness of the data may therefore be useful.

A summary of the experimental evidence suggesting increased cerebral perfusion, anti-oxidative, amyloid beta-reducing, and nootropic neurotransmitter effects:

Clinical treatment studies:

A number of smaller studies demonstrated varied effects, some with improvement of cognitive decline in cognitively normal people and those with mild cognitive decline, while others showed no effect.

An example of one of the smaller studies suggesting short-term benefit:

Gingko biloba as EGb 761 180 mg per day vs. placebo over 6 weeks in people (n = 262) 60 years or older with normal cognition was associated with modest improvement in several measures of cognitive function:

This is one of the analyses from the GEM Study, funded in part by the National Center for Complementary and Integrative Health, that essentially settled the question—adequately-powered, a cohort of people at risk, long-term follow-up, using the preferred preparation EGb 761 from the German Schwabe Pharmaceuticals. Ginkgo biloba 120 mg twice per day was associated with no difference in cognitive measures over 6.1 years (n = 3069) vs. placebo in cognitively normal people age 72-96 years:

Another report from the same GEM Study group (above) Gingko 120 mg twice per day vs. placebo yielded no difference in cognitive measures in participants aged 75 years or older with normal cognition (n = 2587) or MCI (n = 482) over 6.1 years of treatment:

A meta-analysis of short-term trials of gingko as 240 mg per day of EGb 761 for 22-26 weeks showed moderate effects on slowing progression of cognitive decline and dementia (Alzheimer’s and vascular), but no effect on preventing the development of cognitive decline in normal people (?):


The rigor, number of participants, and length, I believe, of the GEM Study trump the smaller, shorter analyses and suggest that gingko is not associated with any appreciable preservation of cognitive health, despite the promising findings from experimental models.