Fisetin is a plant flavonol found in foods such as strawberries and apples. It is being studied as a candidate senotherapeutic because senolytic activity has been demonstrated in selected cell and animal models. The evidence base now includes mechanistic studies, multiple preclinical disease models, human pharmacokinetic work, small randomized trials using several dosing strategies, and an expanding clinical-trial pipeline.
Randomized placebo-controlled human trial registry record with no posted outcome results. It documents design and dosing, not efficacy.
Source Trial registry
Purely in vitro study in a single hepatocellular carcinoma cell line infected with Chikungunya virus; no animal or human data.
Sourcedoi:10.1038/s41598-025-09213-6PMID 40645991PMC12254361
Well-designed single-dose double-blind randomised crossover PK study; industry-sponsored — funded by Akay Natural Ingredients (patent holder for the tested formulation), with co-authors from Life Extension, which sells the formulation commercially as Bio-Fisetin.
Sourcedoi:10.1017/jns.2022.72PMID 36304817PMC9574875 Trial registry
Completed Phase 2 placebo-controlled human trial with registry results posted on ClinicalTrials.gov. This record is stronger than a protocol-only entry but remains registry-level evidence without a peer-reviewed results publication identified in the Data Center.
Source Trial registry
Randomized double-blind crossover human trial registry record with no results yet.
Source Trial registry
Active human interventional registry record without posted trial results.
Source Trial registry
Narrative review of in vitro prostate, lung and pancreatic cancer cell line studies plus a single athymic nude mouse xenograft experiment. No human clinical trial data.
In vitro human and mouse breast cancer cell line work plus a single orthotopic mouse mammary tumour model. No human clinical trial data.
Randomized, double-blind, placebo-controlled human trial during chemotherapy. Tier 2 reflects direct controlled clinical evidence, tempered by the small sample, short duration, biomarker endpoints, and the fact that only IL-8 showed a statistically significant between-group difference.
In vitro biochemical, biophysical (surface plasmon resonance) and computational docking study using prostate cancer cell lines. No animal or human data.
Sourcedoi:10.1016/j.canlet.2015.07.030PMID 26235140PMC4570246
In vitro ovarian cancer cell-line work combined with an in vivo SKOV3 xenograft mouse model. No human data.
Purely in vitro cancer cell-line study (human renal carcinoma, hepatoma and normal mouse kidney cells); no animal or human data.
In vivo adenine/potassium-oxonate-induced hyperuricemic nephropathy mouse model combined with in vitro uric-acid-stimulated mouse tubular epithelial (TCMK-1) cell experiments. No human clinical trial data.
In vivo aged mouse model (including a transgenic p16-3MR senescent-cell-clearance model) combined with human endothelial cell culture experiments. No human clinical trial data.
Randomized double-blind supplementation trial in 60 sedentary men with obesity, reported using intention-to-treat ANCOVA. Human intervention evidence warrants Tier 2, but this publication and the companion Maresin-1 paper report different outcomes from the same underlying four-arm trial (IRCT20120129008863N14), so they must not be treated as independent replication.
Sourcedoi:10.3390/nu18030433PMID 41683255PMC12899003 Trial registry
Controlled animal study (high-fat/high-cholesterol diet-fed C57BL/6J mice) combined with in vitro AML12 hepatocyte experiments and novel nanoparticle formulation work. No human data.
Sourcedoi:10.1186/s12951-026-04181-zPMID 41761201PMC13050004
Purely in vitro study in a single primary endothelial cell type (HUVECs); no animal or human data.
A completed, published, randomised controlled human trial with real inflammatory and metabolic biomarker results — genuine Tier 2 human RCT evidence. Not Tier 1 because it is a single, moderate-sized trial (44 participants across four arms) testing a specific population (obese men) and metabolic/inflammatory endpoints rather than the frailty/senescence endpoints most relevant to fisetin's headline longevity claims.
Sourcedoi:10.1080/15502783.2026.2679718PMID 42218768PMC13224698 Trial registry
Multi-modal preclinical mechanistic study (in vivo aged mice, ex vivo plasma perfusion, and in vitro human endothelial cells with rescue) with a strong mechanistic thread from senolytic treatment to circulating SASP factor CXCL12 to endothelial dysfunction. Peer-reviewed in Aging Cell in 2026.
Peer-reviewed ex vivo and animal hypertension research with direct blood-pressure measurements, but no human blood-pressure trial.
Randomized placebo-controlled human trial registry record with no posted results.
Source Trial registry
Peer-reviewed mechanistic study combining diabetic and naturally aged mouse models with endothelial-cell experiments. It directly measures vascular senescence but remains preclinical.
Mechanistically strong preclinical study combining an in vivo mouse model of doxorubicin-induced premature vascular aging with human aortic endothelial-cell experiments. Tier 3 because no human clinical vascular outcome was tested.
Foundational, peer-reviewed preclinical/translational study spanning cell systems, two aging mouse paradigms and ex vivo human adipose tissue. It provides strong mechanistic and animal evidence but no in-vivo human efficacy evidence; the small mouse lifespan result was not reproduced in the later large NIA ITP UM-HET3 study at the regimens tested.
Sourcedoi:10.1016/j.ebiom.2018.09.015PMID 30279143PMC6197652
Peer-reviewed narrative expert review integrating preclinical, pharmacokinetic and early human fisetin evidence. It contains clinically useful synthesis but no preregistered systematic-review methods, formal risk-of-bias grading or pooled treatment estimate, and the available human evidence was sparse and heterogeneous.
Peer-reviewed rationale/protocol publication linked to a Phase 2 randomized placebo-controlled trial. The final ClinicalTrials.gov record reports termination for futility with actual enrollment of 20 and registry results posted in August 2025; the publication itself does not provide efficacy results.
Sourcedoi:10.1111/jgs.17416PMID 34375437PMC8447437 Trial registry
Peer-reviewed narrative review supplemented by in-silico ADME, target and toxicity predictions from multiple web-based machine-learning platforms. It is useful for hypothesis generation but provides no new human pharmacokinetic, safety or efficacy data and should not be treated as clinical evidence.
This study provides early mechanistic in vitro evidence demonstrating the senolytic properties of fisetin and specific BCL-XL inhibitors in targeted human cell lines.
In-vitro mechanistic study in two doxorubicin-induced senescence models (WI-38 fibroblasts and A549 adenocarcinoma cells) with biological triplicates; useful for mechanism and formulation comparison but provides no animal or human efficacy evidence.
Multi-site, genetically heterogeneous (UM-HET3) mouse lifespan study run in parallel across three independently NIA-funded testing centres — the design the field treats as the gold standard for preclinical longevity claims, since a result must replicate across sites to count. Still preclinical (mouse), so capped at Tier 3 alongside other in-vivo lifespan studies; included specifically because a null result from a protocol this rigorous is stronger negative evidence than a single-site positive result would be positive evidence.
Sourcedoi:10.1007/s11357-023-01011-0PMID 38041783PMC10828146
Peer-reviewed human cohort and biomarker-validation study with an exploratory longitudinal fisetin subgroup. Human evidence is directly relevant, but the fisetin analysis was only n=10, self-reported, uncontrolled and non-randomised, so it cannot establish senolytic efficacy.
Tier 3: combines an in vivo bleomycin-induced pulmonary fibrosis mouse model with in vitro senescence assays in A549 and HELF cells. No human data; single research group.
Controlled large-animal (sheep) study — a more translational preclinical model than mouse/rat work, but still an animal model with no human data.
Tier 4: a narrative review focused on fisetin's chemistry, synthesis routes and pharmaceutical preparation methods, not new pharmacological or clinical data.
Recruiting randomized study; registry describes a planned intervention but no results have been posted or peer-reviewed.
Source Trial registry
Registered human pharmacokinetic study awaiting results; the planned fed-versus-fasted comparison is not observed absorption evidence.
Source Trial registry
Recruiting Phase 1/2 trial registry only; neither efficacy nor safety results are available yet.
Source Trial registry
Active randomized clinical trial with no published or posted outcomes; registration cannot demonstrate vascular benefits.
Source Trial registry
Peer-reviewed controlled preclinical mouse study showing topical fisetin effects on UVB-induced skin injury and photoaging-related markers. The model supports topical mechanistic proof of concept only; it does not establish efficacy of oral fisetin supplements or clinical photoprotection in humans.
Preclinical/translational work combining senescent human dermal fibroblasts with aged human skin grafted into nude mice. Human tissue makes the model relevant, but there was no direct fisetin intervention in living human participants.
In vivo mouse (SKH-1 hairless mice) UVB photodamage model. No human data.
Peer-reviewed randomized mouse wound-healing study with direct topical fisetin exposure and skin histology, but no human dermatology trial.
Tier 3: controlled in vivo rat model (STZ-induced diabetes) with a companion in vitro free-radical scavenging assay (DPPH, ABTS). No human data.
Tier 3: controlled in vivo rat model (STZ-induced diabetes), mechanistic hepatic gene/protein expression endpoints. No human data.
Tier 3: controlled in vivo rat model (STZ-induced diabetes), mechanistic hepatic and renal carbohydrate-enzyme endpoints. No human data.
Purely in vitro enzymology, analytical chemistry and computational (molecular docking) study; no cells, animals or humans were used.
Trial registration without posted results; randomization and planned endpoints do not constitute observed clinical efficacy or safety evidence.
Source Trial registry
Tier 4: a narrative book-chapter review, not primary data. Synthesises preclinical (mostly cell-culture and rodent) literature on fisetin across neurological, diabetic, obesity, atherosclerosis, cancer and allergic-inflammatory disease; the tier reflects the secondary, non-empirical nature of the chapter itself rather than the quality of the underlying primary studies it cites (many of which are separately catalogued in this Data Center at their own tier).
Narrative review of in vitro and animal antioxidant, anticancer and neuroprotective studies, plus rodent pharmacokinetic data. No human clinical trial data.
Preclinical mechanistic study combining in vitro human keratinocyte experiments with topical treatment of C57BL/6 mice; hypothesis-generating for hair-growth applications of fisetin.
Peer-reviewed protocol for an ongoing randomized human biological-effect trial. It is important for contemporary dosing and trial design but provides no efficacy results yet.
Sourcedoi:10.1371/journal.pone.0346347PMID 42308222PMC13274875 Trial registry
Peer-reviewed protocol for a planned triple-blind randomized placebo-controlled human trial. It is methodologically important but contains no efficacy results yet, so it remains Tier 4 until outcome data are available.
Sourcedoi:10.1111/bcpt.70290PMID 42633989PMC13500367 Trial registry
Randomized placebo-controlled human registry record with no posted outcome results.
Source Trial registry
A genuine human pilot study using a validated epigenetic biological-age test (TruAge), but uncontrolled (no placebo arm), unblinded, and very small (n=10) — a hypothesis-generating pilot, not a confirmatory trial. Included specifically because it is real, published human dosing data, and because its mixed/null result is directly relevant to honest evidence grading.
SourcePMID 39269340
Peer-reviewed narrative review with a structured registry search; summarizes heterogeneous completed and ongoing studies, many without outcomes. Not a systematic meta-analysis or an independent trial.
Phase 2 randomized placebo-controlled registry record with no posted results.
Source Trial registry
Tier 4: in vitro-only study using the RAW264.7 mouse macrophage cell line. No animal or human in vivo data.
Controlled animal-model study (mouse alcoholic liver disease model). Mechanistic and biochemical endpoints only; no human data.
Controlled long-term dietary intervention in an established mouse model of diet-induced obesity with extensive metabolic, histological and molecular endpoints. Tier 3 because all efficacy evidence is preclinical and cannot establish human weight-loss or metabolic benefit.
Peer-reviewed, randomized preclinical mouse experiments with blinded outcome collection and transcriptomic analysis. The study provides strong proof-of-concept evidence for physical-function and skeletal-muscle transcriptional effects in aged mice, but it does not establish human efficacy or formal equivalence between fisetin and the comparator senolytic approaches.
Randomized clinical trial registry record without posted results; fisetin is tested alone and with losartan in the context of BMAC.
Source Trial registry
Tier 3: primary human OA chondrocyte culture (in vitro) combined with an in vivo mouse surgical OA model (DMM). No human clinical trial data.
Completed randomized placebo-controlled human trial with posted registry results and a published April 2025 OARSI meeting abstract (both describe the same study). Small single-site sample with discordant 75 registry enrollment versus 74 abstract randomization, limited independent replication and no full clinical journal results article independently identified.
Source Trial registry
Completed very small human pilot registry record; no posted outcome results were available at verification.
Source Trial registry
This study provides preclinical in vivo evidence demonstrating the neuroprotective and anti-aging effects of fisetin in a senescence-accelerated animal model.
Controlled human crossover screening trial with placebo periods and within-person comparisons. Tier 2 reflects direct human intervention evidence, but confidence is limited by the small sample, pseudo-randomized design, multiple botanicals, and screening rather than confirmatory efficacy purpose.
Sourcedoi:10.3390/ijerph18052483PMID 33802381PMC7967624 Trial registry
Tier 4: a narrative review by a single author synthesising the author's own and others' preclinical (cell-culture and rodent) work on fisetin and brain ageing. Not primary data in itself, though it references primary rodent behavioural studies (object recognition, Morris water maze) at what would separately be tier-3.
Systematic review synthesising preclinical (cell culture and animal model) fisetin neuroprotection data across multiple neurological diseases. No new primary data, and no human trial data included.
Tier 4: in vitro-only study using rat astrocyte cultures. No animal or human in vivo data.
Tier 3: a controlled in vivo mouse model (5 groups, male Swiss albino mice) with biochemical, immunohistochemical and RT-PCR endpoints. No human data. Mechanistic rather than translational — supports plausibility of fisetin's neuroprotective effect against aluminium-induced neurotoxicity but has not been replicated in other labs or extended to primates or humans.
Randomized, double-blind, placebo-controlled human trial with 192 acute ischemic-stroke patients and condition-specific clinical plus biomarker endpoints. Tier 2 rather than Tier 1 because it is a single-center short-duration adjunct study and does not establish broader neurological or cognitive benefits from routine supplementation.
In vitro study using inducible mouse cortical neuronal cells, rat primary cortical neurons and HEK293 cells; no animal behavioural model or human data.
In vivo APPswe/PS1dE9 Alzheimer's disease double-transgenic mouse model with long-term (9-month) oral fisetin dosing and behavioural, biochemical and eicosanoid analyses. No human clinical trial data.
In vivo mouse model of transient middle cerebral artery occlusion (tMCAO) combined with in vitro murine and human macrophage/microglial LPS-activation assays. No human clinical trial data.
In vitro murine microglial cell line experiments combined with an in vivo mouse model assessing inflammation-related microglial activation and coordination/motor deficits. No human clinical trial data.
Ex vivo rat hippocampal slice electrophysiology/biochemistry combined with an in vivo behavioural memory test (object recognition) in healthy mice. No human clinical data.
Human observational dietary-intake survey, not interventional. Not fisetin-specific research (fisetin is one of eight flavonoids quantified) but is the primary source for the widely cited ~0.4 mg/day dietary fisetin intake figure. Catalogued as a supporting dietary reference, not a pharmacology study.
In vitro cytotoxicity, migration and Matrigel angiogenesis assays combined with an in vivo Lewis lung carcinoma-bearing mouse tumour model. No human clinical trial data.
Tier 3: in vitro mechanistic work across multiple human cancer cell lines combined with an in vivo nude-mouse xenograft tumour-growth endpoint. No human trial data.
In vitro liver cancer cell line work (HepG2, Hep3B, Huh-7, and radio-resistant HepG2R/Hep3BR sublines) combined with an in vivo HepG2 xenograft mouse tumour model. No human clinical trial data.
Peer-reviewed rationale and protocol for a randomized trial, not an outcome report; no published fisetin efficacy results.
Sourcedoi:10.1177/17588359261424668PMID 41835341PMC12979913 Trial registry
Tier 4: narrative review of preclinical/mechanistic animal and cell-culture evidence on fisetin in colorectal cancer, including a dietary-source summary (strawberries 160 μg/g, apples 26.9 μg/g, persimmons 10.5 μg/g fisetin content, drawn from the paper's own cited sources). No original human or animal data generated by this review itself.
Combines computational network pharmacology, single-cell/spatial transcriptomics, a rat ligature-induced periodontitis model, and in vitro human periodontal ligament fibroblast experiments. No human clinical trial data.
Sourcedoi:10.1016/j.identj.2024.12.012PMID 39755534PMC12142780
Tier 3: in vivo mouse pharmacokinetic and Lewis lung carcinoma tumour-growth study, following formulation development/characterisation work. No human data.
Tier 3: in vitro anticancer/apoptosis assay in MCF-7 breast cancer cells plus in vivo mouse (C57BL/6) oral bioavailability comparison, following nanoparticle formulation and characterisation. No human data.
Sourcedoi:10.1080/10717544.2016.1245366PMID 28156161PMC8241160
Tier 3: in vitro cytotoxicity in HCT116 colon cancer cells plus in vivo antitumour efficacy in a 4T1 breast cancer xenograft mouse model, following PLA nanoparticle formulation and optimisation. No human data.
Tier 3: in vivo mouse pharmacokinetics (plasma and tumour tissue) combined with in vitro cytotoxicity/anti-angiogenic assays. No human PK data for fisetin exists at this evidence tier.
Tier 3: controlled in vivo rat pharmacokinetic study (intravenous and oral dosing) with in vitro antihemolysis assays of serum metabolites. No human PK data.
Methodological review, not primary fisetin research. Cited across the cluster only to source the standard FDA body-surface-area allometric scaling factor used to convert mouse doses to human-equivalent doses — not as fisetin-specific evidence.
Tier 3: controlled in vivo mouse model of LPS-induced septic acute kidney injury with mechanistic biochemical and histopathological endpoints. No human data.
Two independent mouse models of chronic kidney disease (adenine diet and unilateral ureteral obstruction) plus in vitro tubular epithelial cell work and correlative human kidney biopsy staining. No human interventional data.
Sourcedoi:10.1038/s41401-023-01156-wPMID 37696989PMC10770410
Peer-reviewed structured narrative review with PubMed, Google Scholar and ClinicalTrials.gov searches through December 2025, focused on reproductive and gynecologic fisetin evidence. It provides broad mechanistic mapping but no formal risk-of-bias assessment, certainty grading or pooled effect estimate, and direct human reproductive-health efficacy evidence is absent.
New randomized double-blind placebo-controlled Phase 2 registry record with no results.
Source Trial registry
In vitro human bronchial epithelial cell (BEAS-2B) work combined with an in vivo ovalbumin-sensitized mouse model of allergic asthma. No human clinical data.
Open-label two-center feasibility pilot. Walking and chair-stand measures improved from baseline, while lung function and reported health did not. Adverse events included one serious event. Direct human intervention or pharmacokinetic evidence. This is an individual study, not replicated highest-tier clinical proof; sample size, controls and the measured endpoint limit its interpretation.
Sourcedoi:10.1016/j.ebiom.2018.12.052PMID 30616998PMC6412088
Open-label phase 1 senolytic pilot. Selected adipose, skin, and circulating senescence-related markers decreased. Direct human intervention or pharmacokinetic evidence. This is an individual study, not replicated highest-tier clinical proof; sample size, controls and the measured endpoint limit its interpretation.
Sourcedoi:10.1016/j.ebiom.2019.08.069PMID 31542391PMC6796530
Animal intervention experiments with ex vivo human adipose work. The combination reduced selected senescence-related tissue findings and improved physical function and remaining survival in old mice. Animal or translational experimental evidence. Findings concern the tested model and do not establish clinical efficacy, a human dose or a longevity benefit.
Peer-reviewed Phase 2 human trial protocol with a registered, actively recruiting randomized study, but no efficacy results yet.
Sourcedoi:10.1186/s13063-024-08474-2PMID 39434114PMC11492760 Trial registry
Completed randomized Phase 2 human intervention trial with results posted to ClinicalTrials.gov. Tier 2 reflects direct human randomized evidence, but the registry results have not undergone journal peer review and the study tested several senolytic interventions in a relatively small female-only cohort.
Source Trial registry
Veterinary/agricultural animal trial in broiler chickens under heat stress. Not a human model and not directly translatable to human dosing or physiology; included in the Data Center for mechanistic antioxidant interest only.
