Longevity supplements, checked arrow by arrow

Malondialdehyde fell in pooled placebo-controlled trials; two inflammation markers (CRP, TNF-α) did not. A marker moving is not the same as a health outcome.

• Pooled analysis of randomized trials, 380 people (2022) — primary outcome: improved. Funded by independent sources.

It does — that part of the whiteboard holds in people. Whether higher blood NAD+ makes anyone healthier or live longer is a separate question nobody has answered.

• Double-blind randomized trial, 80 people, 9 weeks (2023) — primary outcome: improved. Funded by the manufacturer.
• Open-label randomized trial, 65 people, 2 weeks (2026) — primary outcome: improved. Funded by the manufacturer.

Pooled trials against control groups, many unblinded. This is berberine; dihydroberberine is a related form with no outcome trial of its own. If you take diabetes medicine, ask your doctor first — the effects add up.

• Pooled analysis of randomized trials (some against usual care, not a placebo), 2,313 people (2019) — primary outcome: improved. Funder not checked.

Striking results in worms, mice and monkeys. In people, lower taurine goes with age-related disease — a correlation, not a test. Its own authors call for human trials.

• Animal study, size not stated (2023) — primary outcome: improved. Funder not checked.

Pooled placebo-controlled trials in adults over 60 found no effect on muscle mass, grip strength or walking speed.

• Pooled analysis of randomized trials, size not stated (2025) — primary outcome: no difference from placebo. Funded by independent sources.

A year-long double-blind trial at 0.9 mg a day found no change in memory. IM8 uses 3 mg; higher doses have not been tested for memory.

• Double-blind randomized trial, 100 people, 52 weeks (2022) — primary outcome: no difference from placebo. Funded by mixed sources.

The famous result depended on an artificial fluorescent tag in the test; with natural substrates, resveratrol did not activate SIRT1. Researchers still disagree about special cases.

• Lab (cell) study, size not stated (2010) — primary outcome: no difference found. Funded by the manufacturer.

Fisetin clears senescent cells in mice. In people, the early signals come mostly from quercetin COMBINED with dasatinib, a prescription cancer drug — not quercetin alone. Fisetin trials are under way.

• Narrative review, size not stated (2026) — primary outcome: result not reported. Funder not checked. Not counted toward the grade.

The trial the page cites compares the powder with a usual diet — no placebo, participants not blinded — and its main outcome is BMI and body measurements, not energy. No results are posted. "95% felt more energy" is a self-report from people who knew they were taking it.

• Open-label randomized trial, 60 people, 12 weeks (2024) — secondary outcome: result not reported. Funded by the manufacturer.

No trial of the product. Its ten compounds range from one that does what it says to a marker (NMN → NAD+) to several with no human test at all — see the diagram above. The doctor recommending it helped formulate it and offers a discount code.

From a sponsored article for IM8 on libertyproject.com (it carries a reader discount code). Its own figure is a hypothetical — “you’re absorbing maybe 20%” — and we found no study measuring supplement absorption in “leaky gut” (PubMed, 2026-10-25). Its point about proprietary blends hiding doses is fair; IM8 does list its amounts.

No trial of any NMN product has measured lifespan or how people age. The NMN rows above are what has been tested: it raises NAD+ in the blood, and has not reliably kept muscles strong. “Third-party tested” checks what is in the bottle, not what it does.

The trial is a preprint, not a published, peer-reviewed paper, paid for by the company that makes Cellitas. Its 32% is a rise in a blood protein (Beclin-1) within the Cellitas group, measured as an exploratory extra; the paper did not test it against placebo, whose group switched to Cellitas after 12 weeks. A blood protein is not autophagy measured in cells.

• Double-blind randomized trial, 42 people, 24 weeks (2026) — secondary outcome: result not reported. Funded by the manufacturer.

We found no placebo-controlled trial of injected, infused or nasal NAD+ for energy or thinking. The one human study of IV NAD+ we found measured tolerability: infusions caused moderate to severe stomach upset, a faster heart rate and chest pressure. A 2026 review calls IV “longevity therapy” experimental. Prescription compounded NAD+ is not an FDA-reviewed product.

• Observational study, size not stated (2026) — secondary outcome: result not reported. Funded by the manufacturer.