Spermidine has been marketed as a brain-protecting, longevity-adjacent supplement. The SmartAge randomized clinical trial, published in JAMA Network Open, tested that idea in people with subjective cognitive decline. Twelve months of about 0.9 mg per day of spermidine did not improve memory performance more than placebo.
Why spermidine got so much hype
Spermidine is a polyamine, a small molecule found in all living cells and in foods such as wheat germ. The excitement comes mostly from laboratory and animal research. In those models, spermidine has been linked to autophagy, the cellular housekeeping process that clears damaged components, and to beneficial effects on brain and cognitive health. The SmartAge authors describe that evidence in their own framing of the study.
That is a tidy story: boost a recycling pathway, protect the aging brain. The authors also cite preliminary evidence of memory improvement in people with subjective cognitive decline (SCD), meaning people who feel their memory is slipping but don't meet criteria for dementia. SmartAge was designed to test whether that early signal held up over a longer period and in a larger group.
The trial: who, what, and how much
SmartAge was a 12-month, randomized, double-masked, placebo-controlled phase 2b trial run at a single academic clinical research center in Germany. It ran from January 2017 to May 2020.
- Participants: 100 adults aged 60 to 90 with SCD (mean age 69; 49% female), recruited through health care facilities and advertisements.
- Groups: 51 received spermidine and 49 received placebo, randomly assigned 1:1.
- Intervention: A wheat-germ-derived supplement supplying about 0.9 mg of spermidine per day. The 750 mg extract also contained roughly 0.5 mg of spermine and 0.2 mg of putrescine. The placebo was microcrystalline cellulose.
- Completion: 89 of 100 participants (89%) finished the intervention.
- Registration: ClinicalTrials.gov NCT03094546.
The primary outcome was the change in memory performance from baseline to 12 months. It was measured as mnemonic discrimination on the Mnemonic Similarity Task, which tests whether a person can tell a previously seen image from a very similar new one. The analysis followed the intention-to-treat principle, so everyone randomized was counted.
The primary result: null
The primary outcome showed no benefit. The between-group difference in mnemonic discrimination was -0.03 (95% CI, -0.11 to 0.05; P = .47). The confidence interval spans zero, and the point estimate leans slightly negative rather than positive. The authors report that secondary outcomes, which covered additional neuropsychological, behavioral, and physiological measures, also showed no significant changes.
Their conclusion is that longer-term spermidine supplementation in people with SCD "did not modify memory and biomarkers compared with placebo." Adverse events were balanced between the groups, so the supplement looked tolerable at this dose, but that is not the same as it working.
What about the exploratory signals?
The abstract notes that exploratory analyses "indicated possible beneficial effects" on inflammation and verbal memory. Treat that cautiously. These were exploratory, not the prespecified primary test, and the trial's main finding and its secondary outcomes were null. The authors say such signals would need validation in future studies, and they frame that validation as testing at higher dosage. Exploratory findings can generate hypotheses, but they are not evidence that a supplement works, and they shouldn't appear on a product label.
What this means for supplement buyers
Don't expect a memory boost from this dose. If you are an older adult worried about cognitive decline and considering a spermidine supplement at roughly this level, this trial gives no evidence of a memory benefit over a year. Marketing that leans on autophagy and animal data is describing biology, not a proven outcome in people.
Animal promise did not translate here. The trial is a clear example of the gap between preclinical results and human trials. A mechanism that looks compelling in animal models can still fail to move a clinical endpoint at a realistic supplement dose over a realistic time frame. This study tested one dose for one year. It does not show that spermidine can never matter, but it does not support the confident claims common in the supplement market.
Safety is not efficacy. Balanced adverse events are reassuring, but they don't justify spending money on a benefit the trial couldn't find.
Limitations worth keeping in mind
- Single dose level. Only about 0.9 mg per day was tested, so the study can't say whether higher doses would behave differently. The authors themselves point to higher dosage as a next step.
- Population. Participants had subjective cognitive decline, not dementia or mild cognitive impairment. The results don't necessarily apply to other groups, such as younger adults, people with diagnosed neurodegenerative disease, or people with very different diets.
- Single center, modest size. With 100 participants at one German site, the trial was sized as a phase 2b study, not a definitive large-scale confirmation.
- Duration. Twelve months may be too short, or the Mnemonic Similarity Task may not capture what spermidine could influence, though the broader secondary outcomes were also null.
- Exploratory analyses. The possible inflammation and verbal memory signals came from exploratory analyses and have not been confirmed.
The bottom line
SmartAge is a well-designed, preregistered, placebo-controlled test of a popular idea, and the idea did not pass on its main question. For now, the honest summary is that at about 0.9 mg per day for 12 months, spermidine did not improve memory performance in older adults with subjective cognitive decline compared with placebo. Whether higher doses or different populations would change that is an open question for future research, not a claim supplement buyers can rely on today.
