Does Phosphatidylserine Work? FDA's 'Qualified' Claim vs. Bovine Data
Educational overview — not medical advice. This page explains a regulatory claim; it is not a statement that phosphatidylserine treats, prevents, or cures dementia, Alzheimer's disease, or any diagnosed condition. This statement has not been evaluated by the FDA. This product is not intended to diagnose, treat, cure, or prevent any disease.
In 2003, the FDA allowed a "qualified health claim" for phosphatidylserine — its weakest claim tier, usable only with the mandated disclaimer that "very limited and preliminary scientific research suggests that phosphatidylserine may reduce the risk of cognitive dysfunction in the elderly." The trials behind that claim used bovine-cortex PS, discontinued industry-wide decades ago. Every product sold today is soy or sunflower PS — a different substance, per EFSA — and the direct soy-PS replication of the classic trial design (Jorissen 2001) was null.
What a "qualified health claim" legally means
FDA allows dietary supplement labels to make three tiers of claim about a nutrient-disease relationship. An authorized health claim requires "significant scientific agreement" among qualified experts — the highest bar, reserved for well-established relationships like calcium and osteoporosis. A qualified health claim sits below that: FDA allows it specifically when the evidence doesn't meet the significant-scientific-agreement bar, but the agency judges there's enough credible evidence to let a hedged, disclaimer-carrying statement through rather than block it outright. For phosphatidylserine, FDA's own 2003 determination stated there is "a lack of scientific agreement amongst qualified experts that a relationship exists between phosphatidylserine and cognitive function" — that sentence is the reason the claim is qualified rather than authorized, and it is the FDA's own words, not a critic's characterization. A qualified health claim is a regulatory permission slip for a hedged statement. It is never proof, and it is never an endorsement.
The exact disclaimer language, and why it matters that it's exact
The approved underlying claim reads "consumption of phosphatidylserine may reduce the risk of dementia and cognitive dysfunction in the elderly" — but FDA requires that claim be paired, every time, with this specific mandated disclaimer: "very limited and preliminary scientific research suggests that phosphatidylserine may reduce the risk of cognitive dysfunction in the elderly." This exact wording is verified against the FDA's May 13, 2003 qualified-claim letter. It is not marketing copy softened for legal caution — it is the FDA itself describing the underlying research as "very limited and preliminary." Any label, ad, or AI-generated summary that states the FDA claim without also carrying that disclaimer is presenting the weakest tier of regulatory permission as though it were a stronger one.
The verdict The claim is real. It's also the weakest thing FDA lets a label say — and it's attached to an ingredient nobody sells anymore.
The ingredient switch the claim's history depends on
The evidentiary basis for the FDA's 2003 determination traces back to two trials: Crook 1991 (n=149, age-associated memory impairment) and Crook 1992 (n=51, probable Alzheimer's disease), both using bovine-cortex phosphatidylserine (BC-PS) at 100mg three times daily. BC-PS was discontinued industry-wide in the 1990s over BSE (mad-cow disease) contamination risk in bovine-derived ingredients broadly. It has not been sold as a supplement raw material in decades. Every PS product on the market today — every one of the six products compared on the best-PS page — is derived from soy or sunflower lecithin instead. This is not a minor technicality. EFSA's expert panel, evaluating the same underlying question in a European regulatory context, formally concluded that bovine-cortex- and soy-based phosphatidylserine "are different substances" that "might...have different biological activities," and explicitly stated that a cause-and-effect relationship between soy PS and cognition "cannot be established" from the available evidence. The FDA's qualified claim was never re-evaluated against soy or sunflower PS specifically — it was evaluated against evidence generated with an ingredient that no longer exists on any shelf.
The closest thing to a direct test: Jorissen 2001
If the honest question is "does the modern version work the way the old version seemed to," the closest available answer comes from Jorissen et al. 2001 (PMID 11842880, Nutrition and Neuroscience). This RCT replicated Crook's exact design — age-associated memory impairment, ages 57+, 12 weeks plus a 3-week washout — and Crook's exact 300mg/day dose (also testing a 600mg/day arm), but substituted soy-derived PS for bovine-cortex PS. The result: "no significant differences...between the treatment groups" on any outcome, including the delayed recall and recognition primary endpoints. The authors' own conclusion was that soy-PS "does not affect memory or other cognitive functions." This is not a perfect head-to-head bovine-vs-soy comparison run in a single trial — it's a separate study using the same design and dose in a comparable population. But it is the best available evidence on the question the FDA claim's marketing use implicitly answers "yes" to, and the trial itself answers "no."
Frequently asked questions
What is a "qualified health claim" and why is it weaker than a normal one?
FDA's lowest claim tier, used when there's insufficient scientific consensus for a full "authorized" claim. FDA itself stated there's "a lack of scientific agreement amongst qualified experts" for PS and cognition.
What is the exact mandated disclaimer language?
"Very limited and preliminary scientific research suggests that phosphatidylserine may reduce the risk of cognitive dysfunction in the elderly" — FDA's own characterization, verified against the 2003 letter.
Why does the bovine-to-soy/sunflower switch matter?
The claim's evidentiary basis used bovine-cortex PS, discontinued for decades. EFSA has concluded bovine and soy PS "are different substances" with possibly different biological activity. Every product sold today is soy or sunflower.
Is there a direct test of whether the modern ingredient works the same way?
The closest is Jorissen 2001 — same design and dose as Crook, soy PS instead of bovine — and it found no significant benefit on any outcome.
Related
- Phosphatidylserine: what it is & who it's for
- PS for age-related memory — the full trial-by-trial breakdown
- Best PS — ranked by disclosed source, then cost per mg
- PS dosage guide — 300mg memory dose vs 600-800mg cortisol dose
Sources
- U.S. FDA, qualified health claim determination for phosphatidylserine and cognitive dysfunction/dementia risk, letter dated May 13, 2003 (mandated disclaimer language verified via two independent fetches, as quoted and footnoted by Wikipedia's phosphatidylserine article).
- Crook TH, Tinklenberg J, Yesavage J, et al. "Effects of phosphatidylserine in age-associated memory impairment." Neurology. 1991. PMID: 2027477
- Crook TH, Petrie W, Wells C, Massari DC. "Effects of phosphatidylserine in Alzheimer's disease." Psychopharmacol Bull. 1992. PMID: 1609044
- Jorissen BL, Brouns F, van Boxtel MP, et al. "The influence of soy-derived phosphatidylserine on cognition in age-associated memory impairment." Nutr Neurosci. 2001. PMID: 11842880
- European Food Safety Authority (EFSA) Panel on Dietetic Products, Nutrition and Allergies — scientific opinion distinguishing bovine-cortex- and soy-based phosphatidylserine as different substances with a cause-and-effect relationship that "cannot be established" for soy PS and cognition.