Saffron And Memory — The Spice With Serious Evidence
Quick Read Saffron contains compounds called crocin and crocetin that appear to protect memory and slow cognitive decline through multiple
We write a new article on supplements every day , all detailed, with citations.
Once a week we send an email containing a list of the last 7 days articles. Please sign up and also feel free to share the articles with friends and family who you think might benefit.
Simon Parrott – Vitacuity Ltd
Phosphatidylserine (PS) is a fat molecule that forms part of your brain cell membranes and helps them communicate efficiently. The FDA issued a rare official endorsement in 2003 saying PS may reduce dementia risk in older people, based on decades of research. The original studies used PS from cow brain, but switched to soy and sunflower sources in the 1990s due to mad cow disease concerns.
Recent human trials show PS produces modest but real improvements in memory and thinking speed, especially in people already noticing cognitive decline. Doses of 100 to 300 mg daily appear effective. People who benefited most were those whose memory was already below average at the start. PS was safe and well tolerated across all studies, with no serious side effects reported.
Important gaps remain: the strongest evidence used bovine-derived PS, most recent studies combine PS with other ingredients making it hard to know PS’s independent effect, and we lack long-term multi-year studies. PS is not proven as a treatment for established dementia, though early support before significant decline is more promising.
Verdict: PS has consistent scientific support for modest memory improvement in people aged 40 and above who are beginning to notice cognitive decline, though it is not a miracle cure and works best as part of a broader brain-health approach.
What if one of the most rigorously studied brain supplements in history quietly received official government backing, and then almost nobody heard about it?
In 1991, researchers were deep into a body of work that had been quietly accumulating for years: dozens of trials, animal studies, and human investigations all pointing at a curious molecule tucked inside the membranes of your brain cells. By 2003, the US Food and Drug Administration had seen enough to issue a qualified health claim, an extraordinarily rare endorsement for any supplement, acknowledging that phosphatidylserine (PS) *may* reduce the risk of dementia and cognitive dysfunction in the elderly. It was a landmark moment. And then, somehow, the story drifted out of the public conversation.
Here at Vitacuity, we’ve analysed over 1.77 million research papers and selected the most relevant studies on this topic. What we found is a genuinely fascinating scientific journey, one that started with bovine brain extract, navigated a crisis, reinvented itself with soy and sunflower, and has continued producing meaningful human data into 2025. This is not a story of a supplement that failed. It’s a story of science doing exactly what it should: evolving, self-correcting, and getting more precise over time.
Here’s what decades of research has actually taught us.
To understand why researchers got so excited about phosphatidylserine, you first need to picture what’s happening inside every neuron in your brain right now.
Your brain cells are wrapped in a membrane, a thin, flexible barrier that controls everything coming in and going out. Phosphatidylserine is one of the primary structural components of that membrane, sitting predominantly on the inner leaflet (the inside face) of neural cell membranes [1]. It’s classified as a glycerophospholipid, essentially a fat molecule with a serine amino acid attached to it, and your brain contains more of it, proportionally, than almost any other tissue in the body [14].
But PS isn’t just passive scaffolding. It acts as a cofactor for enzymes involved in critical signalling pathways, contributes to neurotransmission, plays a role in managing neuroinflammation, and is involved in synaptic refinement, the process by which your brain fine-tunes its connections [1]. It also serves as a precursor molecule, helping to produce other important phospholipids in your brain cells [1].
There’s a particularly important detail about what happens when things go wrong. The position of PS inside the cell membrane matters enormously. When a neuron begins to die (a process called apoptosis), PS flips to the outer leaflet, essentially raising a molecular flag that signals the immune system to clear the cell [4]. This flip also initiates the blood clotting cascade in other contexts [1]. In healthy brains, PS stays mostly on the inside. In ageing or damaged brains, this balance can be disrupted [4].
Early researchers, mostly working in Italy and the US through the 1980s and 1990s, believed that supplementing PS, initially extracted from bovine (cow) brain cortex, could replenish declining membrane PS levels in ageing neurons, restore signalling efficiency, and support neurotransmitter systems, particularly the cholinergic system (the acetylcholine network so critical to memory) [7]. The animal data was compelling. Then came the human trials.
The first major wave of human research on PS used a form derived from bovine brain cortex. This was the form that accumulated the most clinical evidence through the 1980s and early 1990s, and it was this body of work that eventually caught the FDA’s attention.
The findings, reviewed carefully in *Nutrition* (2003) [7], told a nuanced story. For older adults with *moderate cognitive impairment*, people who were struggling but hadn’t yet been diagnosed with full dementia, PS supplementation at 300 mg per day produced “consistently modest increases in recall of word lists” across multiple studies [7]. That’s a meaningful finding. Word list recall is one of the earliest and most reliable indicators of memory decline, and it’s a hard outcome to fake.
For people with probable Alzheimer’s disease, however, the picture was less clear. A single study at that time failed to demonstrate positive effects of bovine-derived PS on memory performance in Alzheimer’s patients [7]. One study, of course, is never conclusive, but it introduced important caution.
There was also a particularly interesting sub-finding that would become a recurring theme across decades of research: the people who benefited most from PS supplementation were consistently those who were *already performing below their peers at baseline* [7]. In other words, PS seemed to do more for people whose cognitive reserves were already being tested. Normal, high-functioning adults in their prime showed less dramatic responses. This pattern isn’t a failure, it’s actually scientifically coherent. If your neurons are well-nourished and your PS levels are healthy, adding more may have a ceiling effect. If they’re depleted, supplementation has more room to work.
Then, in the mid-1990s, the source of PS changed, and the story got more complicated.
Bovine brain-derived PS was effective. But it also came with a practical problem that no supplement company could ignore: the spectre of bovine spongiform encephalopathy, or BSE, mad cow disease. By the mid-1990s, using brain-derived animal extracts in supplements had become commercially and reputationally untenable.
Researchers needed an alternative. The answer came from an unexpected place: soybean lecithin. Using an enzyme called phospholipase D, scientists developed a process called transphosphatidylation, essentially swapping the head group of soy phosphatidylcholine for serine, creating soybean-derived PS (Soy-PS) [10]. It was a genuine feat of biochemical engineering.
The critical question was: does it actually work the same way?
Early rodent studies published in *The Journal of Nutrition* (2001) [6] were encouraging. Soybean-derived PS administered orally to aged rats (24–25 months old) at 60 mg/kg/day for 60 days produced significant improvements in the Morris water maze, a standard test of spatial learning and memory, comparable to the original bovine brain-derived PS. The mechanism appeared to involve increased acetylcholine release and restored sodium-potassium ATPase activity in synaptic terminals [6]. Translated: Soy-PS seemed to be restoring the electrochemical efficiency of ageing neurons, at least in rats.
Earlier rodent work from 1996 in the same research tradition [10] had compared the fatty acid profiles of soy-derived and bovine-derived PS, and found they were quite different. Soy-PS is rich in linoleic and palmitic acids; bovine PS is richer in stearic and oleic acids. Despite this structural difference, both forms restored brain glucose metabolism and reversed scopolamine-induced amnesia (a standard lab model of memory impairment) in rodents [10]. This was reassuring, it suggested that the serine head group, not the fatty acid tail, was the biologically active part for brain function.
Evidence grade for the animal data: Promising to Early Stage, the rodent findings were consistent and mechanistically coherent, but animal studies always require human confirmation.
The human research on soy-derived PS is more limited than the earlier bovine work, but it’s been steadily building, and the pattern is consistent enough to be meaningful.
A double-blind, randomised controlled trial conducted in Japan [9], published in 2010, recruited 78 elderly participants aged 50–69 with mild cognitive impairment (MCI) and assigned them to receive 100 mg Soy-PS per day, 300 mg per day, or placebo for six months. The headline result was modest: neuropsychological scores rose in all groups, including placebo, which suggests a significant practice effect (people get better at cognitive tests simply by taking them repeatedly).
But here’s where it gets interesting, and this mirrors the pattern from the bovine era. In the subgroup of participants who scored *below average at baseline*, both PS doses produced significant memory improvements compared to their own baseline, while the placebo group did not improve [9]. The improvement was predominantly in *delayed verbal recall*, which the researchers noted is precisely the memory function that declines earliest in the trajectory toward dementia [9]. No adverse effects were observed at either dose.
Evidence grade: Promising, one well-designed RCT, 78 participants, six months. Consistent with broader pattern but not yet replicated at scale.
A 2014 pilot study published in *Advances in Therapy* [8] explored a combination of PS plus phosphatidic acid (PA), both derived from soy lecithin. In a double-blind, placebo-controlled trial of elderly participants with memory complaints (n=57 completing the protocol), three capsules daily providing 300 mg PS and 240 mg PA for three months produced significant improvements in memory as measured by the Wechsler Memory Scale, and also appeared to prevent “winter blues”, a mood dip not observed in the placebo group [8]. In a separate arm of the same study, Alzheimer’s disease patients receiving PS+PA maintained their daily functioning scores over two months, while the placebo group declined significantly (P=0.035) [8]. Notably, 49% of PS+PA patients reported improved general condition, versus 26% under placebo [8].
Evidence grade: Promising, pilot study with small per-protocol numbers (n=31 PS+PA, n=26 placebo in the memory arm). Encouraging but needs larger replication.
The most recent human data in our database comes from a 2025 randomised, double-blind, placebo-controlled trial published in the *Journal of Affective Disorders* [2]. This is worth examining carefully because it’s both the largest and the most recent study in this collection.
190 participants with mild cognitive impairment (MCI), average age 67.95 years, were recruited in Tianjin, China, and randomly assigned to receive a supplement combination containing 31.5 mg PS, 144 mg alpha-linolenic acid (ALA), 3.6 mg ginkgo total flavonoids, and 0.48 mg Vitamin B, or placebo, for 12 months.
The intervention group showed statistically significant improvements in: – Arithmetic testing (β = 0.688, 95% CI: 0.103–1.274) – Similarity test (a measure of abstract reasoning) (β = 1.070, 95% CI: 0.472–1.667) – Short-term memory (β = 0.600, 95% CI: 0.399–0.800) [2]
The intervention group also showed significant increases in serum ALA, DHA, EPA, acetylcholine, GABA, and serotonin compared to placebo [2]. Mediation analysis suggested that increased serum ALA levels explained approximately 19.7% of the short-term memory improvement, meaning ALA appears to be one mechanism driving the benefit, though not the only one [2].
An important caveat: this was a multi-ingredient formula. We cannot attribute the results to PS alone. The combination of PS, ALA, ginkgo flavonoids, and B vitamins makes it genuinely difficult to isolate PS’s individual contribution.
Evidence grade: Promising, 190 participants, 12 months, well-designed RCT. Strongest human study in this collection, but complicated by the multi-ingredient formula.
One of the more intriguing developments in PS research has been the exploration of marine-derived sources, from krill and squid, as alternatives to soy.
A 2015 animal study [11] examined squid-derived PS in aged rats over seven days and found dose-dependent improvements in Morris water maze performance, alongside biochemical changes in the hippocampal cholinergic system, specifically, restored choline acetyltransferase and acetylcholinesterase activity, and upregulated choline transporter and muscarinic acetylcholine receptor gene expression [11]. The researchers noted that squid PS performed comparably to soy-PS in their models.
A 2012 study [12] on krill-derived PS in rats with chemically induced memory deficits found significant improvement in escape latency in the Morris water maze, alongside preserved acetylcholinergic neurons in the hippocampus and medial septum, and increased CREB (a protein critical for long-term memory formation) in the hippocampal CA1 area [12].
Evidence grade for marine PS: Early Stage, consistent and mechanistically interesting animal data, but human trials are lacking. Marine sources remain a scientifically legitimate avenue that hasn’t yet been tested adequately in humans.
A 2023 randomised, double-blind, placebo-controlled trial published in *Neurology and Therapy* [5] tested *Neuriva*, a commercial supplement combining whole coffee cherry extract and PS, in 138 healthy adults aged 40–65 with self-reported memory problems, over 42 days.
Compared to placebo, Neuriva supplementation produced: – Greater improvements in numeric working memory accuracy (p ≤ 0.024) – Faster reaction times (p ≤ 0.031) – Improved overall accuracy in picture recognition (p = 0.035) [5]
No significant differences were found in BDNF levels, the Everyday Memory Questionnaire, or Go/No-Go tests [5].
Again, this is a multi-ingredient formula, so PS cannot take sole credit. But the findings add to the consistent picture: in adults who are already noticing memory difficulties, PS-containing supplements appear to produce real, measurable improvements in specific cognitive tasks.
Evidence grade: Promising, 138 participants, 42 days, industry-funded but peer-reviewed. Multi-ingredient formula limits attribution.
Our database includes one 2025 RCT examining sunflower-derived PS (100 mg daily for 12 weeks) in 65 healthy, neurotypical children aged 8–12 [3]. The primary outcomes showed no differences between PS and placebo for the total cohort. However, in a pre-defined subgroup of children who were performing consistently below the median across cognitive tasks at baseline, PS supplementation showed benefit on a visuospatial memory task [3].
This mirrors the adult pattern almost exactly: PS appears most useful where there is already a performance gap. The supplement was confirmed safe and well tolerated in this population [3].
Evidence grade: Early Stage for children, one small RCT, no significant primary outcomes, interesting subgroup finding. Requires replication before any conclusions can be drawn.
Honesty is built into everything we do at Vitacuity, so let’s be direct about the gaps.
The bovine-to-soy transition muddied the evidence base. The strongest human evidence, the work that convinced the FDA to issue its qualified health claim, was done with bovine brain-derived PS. Soy-derived PS has a different fatty acid profile [10], and while the animal data suggests it works similarly, we don’t have the large, long-duration human trials with soy-PS that existed for bovine-PS. This is the single most important gap in the literature.
Most positive human studies involve multi-ingredient formulas. The 2025 RCT [2] combined PS with ALA, ginkgo flavonoids, and B vitamins. The 2023 Neuriva trial [5] combined PS with coffee cherry extract. The 2014 PS+PA study [8] combined two phospholipids. Isolating PS’s independent contribution is genuinely difficult.
The “below median baseline” pattern is consistently observed but not fully explained. Why does PS seem to work specifically in people already struggling? Is it a floor effect? Is it that PS is most bioavailable when neuronal membranes are already depleted? The mechanism here needs dedicated investigation [7, 9, 3].
We don’t have long-duration, large-scale RCTs. The longest human trial in our database is 12 months [2], and it used a multi-ingredient formula. For a supplement being considered in the context of dementia prevention, a decades-long process, we need multi-year data.
The conflict in Alzheimer’s research needs context. Early studies showed no effect in confirmed Alzheimer’s patients [7], but later work showed stabilisation of daily functioning [8]. The difference may come down to disease stage, PS may be more useful before significant neurodegeneration, not after it. This is a crucial distinction that has never been definitively tested.
Marine-derived PS (krill, squid) has compelling animal data and no human trials [11, 12]. This is a genuinely underexplored avenue.
Here’s how a sensible, informed person should think about phosphatidylserine in 2025.
The research journey started in the 1980s, hit a real inflection point with the FDA’s qualified health claim in the early 2000s, and has continued producing meaningful, if not definitive, human data right up to this year. The pattern across decades is remarkably consistent: PS supplementation tends to produce modest but measurable improvements in memory and cognitive function, particularly in people who are already noticing decline. It doesn’t appear to perform miracles in high-functioning adults with no complaints. But for the 40–65 age group who are beginning to notice that names, words and details don’t stick the way they used to, that’s precisely where the evidence points.
The doses used in human trials cluster around 100–300 mg per day, with most positive findings at the higher end [8, 9, 2]. The supplement is consistently reported as safe and well tolerated across all studies in our database, including in elderly, MCI, and Alzheimer’s populations [1, 2, 3, 8, 9].
PS is naturally found in the brain, declines with age, and the body synthesises it from dietary precursors, but synthesis capacity also decreases as we get older [1, 14]. This is not a manufactured deficiency. It’s a real biological trend.
Practical steps:
– If you’re 40+ and starting to notice memory lapses, the evidence genuinely supports trying PS at 100–300 mg daily. It’s not a miracle, but the science across multiple decades and multiple trial designs consistently points in the same direction. – Look for sunflower or soy-derived PS, these are the commercially available, BSE-risk-free forms. The animal data confirms they work through similar mechanisms to the original bovine-derived PS [10, 6]. – Consider PS as part of a broader brain health stack, the best-performing trials tended to combine PS with omega-3 fatty acids or phosphatidic acid [2, 8]. This likely reflects synergistic effects on neuronal membrane health. If you’re already taking an omega-3 supplement, adding PS makes even more sense. – Be patient. The meaningful human trial durations ranged from three to twelve months. Don’t expect overnight results and don’t give up after four weeks. – PS is not a treatment for dementia. The evidence for prevention and early cognitive support is more encouraging than the evidence for established disease. Earlier intervention appears to matter.
The FDA qualified health claim was issued because the evidence crossed a threshold that government regulators rarely acknowledge for supplements. That the story faded from public consciousness says more about the supplement industry’s marketing priorities than about the science itself.
The science didn’t disappear. It kept accumulating. And in 2025, it still points in the same direction it did in 2003, just with more nuance, more precision, and a more honest understanding of who benefits most.
That, ultimately, is what good science looks like.
[1] Phosphatidylserine: A comprehensive overview of synthesis, metabolism, and nutrition. (2024). *Chemistry and Physics of Lipids*. DOI: 10.1016/j.chemphyslip.2024.105422 | https://pubmed.ncbi.nlm.nih.gov/39097133/
[2] Effects of a food supplement containing phosphatidylserine on cognitive function in Chinese older adults with mild cognitive impairment: A randomized double-blind, placebo-controlled trial. (2025). *Journal of Affective Disorders*. DOI: 10.1016/j.jad.2024.09.131 | https://pubmed.ncbi.nlm.nih.gov/39317299/
[3] The cognitive effects of supplementation with sunflower phosphatidyl serine in healthy children aged 8 to 12 years: a randomized controlled trial. (2025). *Nutrition Journal*. DOI: 10.1186/s12937-025-01264-9 | https://pubmed.ncbi.nlm.nih.gov/41318468/ | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12772107/
[4] Role of Enzymes Capable of Transporting Phosphatidylserine in Brain Development and Brain Diseases. (2024). *ACS Omega*. DOI: 10.1021/acsomega.4c05036 | https://pubmed.ncbi.nlm.nih.gov/39157110/ | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11325426/
[5] A Randomized, Double-Blind, Placebo-Controlled, Parallel Study Investigating the Efficacy of a Whole Coffee Cherry Extract and Phosphatidylserine Formulation on Cognitive Performance of Healthy Adults with Self-Perceived Memory Problems. (2023). *Neurology and Therapy*. DOI: 10.1007/s40120-023-00454-z | https://pubmed.ncbi.nlm.nih.gov/36929344/ | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10195946/
[6] Oral administration of soybean lecithin transphosphatidylated phosphatidylserine improves memory impairment in aged rats. (2001). *Journal of Nutrition*. DOI: 10.1093/jn/131.11.2951 | https://pubmed.ncbi.nlm.nih.gov/11694624/
[7] “Brain-specific” nutrients: a memory cure? (2003). *Nutrition*. DOI: 10.1016/s0899-9007(03)00024-8 | https://pubmed.ncbi.nlm.nih.gov/14624946/
[8] Positive effects of soy lecithin-derived phosphatidylserine plus phosphatidic acid on memory, cognition, daily functioning, and mood in elderly patients with Alzheimer’s disease and dementia. (2014). *Advances in Therapy*. DOI: 10.1007/s12325-014-0165-1 | https://pubmed.ncbi.nlm.nih.gov/25414047/ | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4271139/
[9] Soybean-derived phosphatidylserine improves memory function of the elderly Japanese subjects with memory complaints. (2010). https://pubmed.ncbi.nlm.nih.gov/21103034/
[10] Pharmacological effects of phosphatidylserine enzymatically synthesized from soybean lecithin on brain functions in rodents. (1996). https://pubmed.ncbi.nlm.nih.gov/8708821/
[11] Oral administration of squid lecithin-transphosphatidylated phosphatidylserine improves memory impairment in aged rats. (2015). *Progress in Neuro-Psychopharmacology and Biological Psychiatry*. DOI: 10.1016/j.pnpbp.2014.07.004 | https://pubmed.ncbi.nlm.nih.gov/25058912/
[12] Krill-Derived Phosphatidylserine Improves TMT-Induced Memory Impairment in the Rat. (2012). https://pubmed.ncbi.nlm.nih.gov/24116297/
[13] Effect of intracerebroventricular administration of soybean lecithin transphosphatidylated phosphatidylserine on scopolamine-induced amnesic mice. (2000). *Japanese Journal of Pharmacology*. DOI: 10.1254/jjp.84.86 | https://pubmed.ncbi.nlm.nih.gov/11043460/
[14] Phosphatidylserine and the human brain. (2015). https://pubmed.ncbi.nlm.nih.gov/25933483/
[15] Phosphatidylserine. Monograph. (2008). https://pubmed.ncbi.nlm.nih.gov/18950250/
This article is for informational purposes only and does not constitute medical advice. Food supplements should not be used as a substitute for a varied and balanced diet and healthy lifestyle. If you are pregnant, breastfeeding, taking medication or have a medical condition, consult your doctor before taking any supplement. These statements have not been evaluated by the Food and Drug Administration (FDA) or the Medicines and Healthcare products Regulatory Agency (MHRA). This product is not intended to diagnose, treat, cure, or prevent any disease.
Quick Read Saffron contains compounds called crocin and crocetin that appear to protect memory and slow cognitive decline through multiple
Quick Read Dark chocolate contains compounds called flavanols that may help your brain work better by improving blood flow and
Free guide: The research behind brain health supplements
Plus our weekly research digest — straight to your inbox.
Just leave your email address.
