⚠️ Full-text unavailable — permanent verification blocker (Springer paywall; no PMC deposit; no preprint found). gap/no-fulltext-access. Abstract claims confirmed by verifier against PMID 41843349 on 2026-06-30: all eight null-lifespan compound groups, the female harm framing (astaxanthin harm is a JAX-control artifact not confirmed in 2-site reanalysis; mitoglitazone and pioglitazone harm confirmed), and the ITP 3-site design are faithfully sourced from the abstract. Per-compound doses, start ages, p-values, and site-specific effects are NOT in the abstract and require the full PDF before any numeric claim can be verified.
Korstanje et al. 2026 — NIA ITP null cohort (astaxanthin, meclizine, and six additional compounds)
TL;DR
The NIA Interventions Testing Program (ITP) null cohort evaluating eleven interventions (eight named compound groups) in genetically heterogeneous UM-HET3 mice. None significantly extended lifespan in either sex. Critically, this cohort retested astaxanthin, meclizine (meclizine), and mitoglitazone — all of which had shown ITP-positive signals in prior cohorts — at different doses or later starting ages, and all were null. This constitutes the ITP’s own internal negative-replication of those earlier signals, placing astaxanthin and meclizine into contested status. A female harm signal appeared for astaxanthin in the pooled three-site analysis but was attenuated in a two-site reanalysis that excluded an anomalous site, leaving only mitoglitazone (late-start) and pioglitazone with a confirmed negative female effect 1.
ITP design rationale (brief)
The NIA Interventions Testing Program is the highest-rigor preclinical lifespan screen available. Every cohort runs identical protocols simultaneously at three independent sites (The Jackson Laboratory / JAX, UT Health San Antonio, University of Michigan). The genetically heterogeneous UM-HET3 four-way F1 cross (BALB/cByJ Ă— C57BL/6J) Ă— (C3H/HeJ Ă— DBA/2J) captures polygenic variation analogous to outbred humans. Both sexes are tested in all arms. A positive result requires multi-site consistency; a null result at multi-site scale is interpreted as strong evidence against lifespan extension under the tested conditions.
A core feature of the ITP’s design philosophy is deliberate replication of prior positive findings at modified parameters — different doses or starting ages. Harrison 2024 (cohort C2019/C2020) produced positive signals for astaxanthin and meclizine in males; this cohort explicitly re-tests those compounds to probe robustness. Finding null at different parameters is the replication-control system working as designed — it reveals whether the positive signal was narrow in dose-age space or was a false positive.
Compounds tested
The abstract states “eleven compounds” were evaluated, likely reflecting separate arms for different doses or start ages of the same molecule (e.g., late-start mitoglitazone counted separately). The eight compound groups named in the title are:
| Compound | Prior ITP status | This cohort result |
|---|---|---|
| Astaxanthin | Positive — male +12% (Harrison 2024) | NULL — neither sex |
| Meclizine | Positive — male +8% (Harrison 2024) | NULL — neither sex |
| Mitoglitazone | Positive in prior ITP (not Harrison 2024) | NULL — neither sex |
| Pioglitazone | Not previously ITP-positive (novel compound) | NULL — neither sex |
| Alpha-ketoglutarate (AKG) | Not previously ITP-positive | NULL — neither sex |
| Mifepristone | Not previously ITP-positive | NULL — neither sex |
| Methotrexate | Not previously ITP-positive | NULL — neither sex |
| Atorvastatin + telmisartan (combination) | Not previously ITP-positive | NULL — neither sex |
Doses and start ages for each compound in this cohort are not reported in the abstract and require the full paper. The abstract characterises them as “novel doses or starting ages” versus prior ITP cohorts. gap/no-fulltext-access
Key results
Lifespan — all compounds null
None of the eleven tested interventions significantly increased lifespan in male or female UM-HET3 mice pooled across the three sites 1. This includes all three previously ITP-positive compounds (astaxanthin, meclizine, mitoglitazone) when administered under the modified conditions of this cohort.
The abstract does not report compound-level p-values, median lifespan delta percentages, or site-specific effects; those data require the full paper. gap/no-fulltext-access
Non-replication of astaxanthin
astaxanthin did not extend lifespan at the dose and starting age used in this cohort. In harrison-2024-itp-astaxanthin-meclizine, astaxanthin extended median male lifespan by 12% (p=0.003) with consistent cross-site effects (JAX +11%, UT +11%, UM +14%) at ~1840 ppm dietary from 12 months of age. The Korstanje 2026 cohort used a different dose or starting age (specific parameters not stated in abstract); both sexes were null. This is the first direct ITP replication attempt for astaxanthin 1.
Non-replication of meclizine
meclizine did not extend lifespan at the tested dose/start age. In Harrison 2024, meclizine extended median male lifespan by 8% (p=0.03) at ~544 ppm from 12 months, though with site heterogeneity (UT site 0% effect). The Korstanje 2026 null under different conditions is consistent with that pre-existing concern about robustness 1.
Non-replication of mitoglitazone
Mitoglitazone (MSDC-0602K; a thiazolidinedione structurally related to pioglitazone but designed to minimize PPAR-γ engagement, primarily targeting the mitochondrial pyruvate carrier — mechanism not addressed in the abstract or in the abstract-derived content of this page) had previously shown an ITP-positive lifespan signal; the abstract does not specify which prior cohort. Both early-start and late-start arms of mitoglitazone appear to have been tested (the abstract references “late-start mitoglitazone” separately in the female-harm analysis). Neither arm extended lifespan 1.
Female harm signal — conditional on site
The abstract explicitly addresses a female harm signal and its resolution:
“In females, astaxanthin, late-start mitoglitazone, and pioglitazone were associated with significantly reduced lifespan when pooling the data from all three sites. However, site-specific analysis revealed unusually long lifespans in control females at The Jackson Laboratory, prompting reanalysis using data from the other two sites and only showed a negative effect for mitoglitazone and pioglitazone.”
Summary of the female harm analysis:
| Compound | 3-site pooled harm? | 2-site reanalysis (excl. JAX) harm? |
|---|---|---|
| Astaxanthin | Yes (significant) | No (attenuated; not confirmed) |
| Late-start mitoglitazone | Yes (significant) | Yes (confirmed) |
| Pioglitazone | Yes (significant) | Yes (confirmed) |
The astaxanthin female harm in the pooled analysis appears to be a statistical artefact of JAX’s unusually long control female lifespans in this cohort, not a genuine compound toxicity signal. In the two-site reanalysis excluding JAX, astaxanthin does not show a confirmed female harm effect. Mitoglitazone (late-start) and pioglitazone retain their negative female effects across the two-site reanalysis, making those the compounds with the most concerning female safety signal in this cohort.
The reason for JAX’s anomalously long control female lifespans in this cohort is not stated in the abstract and requires the full paper.
Alpha-ketoglutarate, mifepristone, methotrexate, atorvastatin + telmisartan
These four compound groups had not previously been tested in the ITP (as of the Harrison 2024 cohort). All were null for lifespan in both sexes. Individual mechanistic rationales and the doses used are not stated in the abstract 1. Pioglitazone, a thiazolidinedione with PPAR-Îł agonism and metabolic-syndrome indications, also had not appeared in prior ITP cohorts; it was null for lifespan and associated with confirmed female harm in the reanalysis.
Interpreting the ITP replication logic
The ITP’s replication-by-modification design deliberately creates this type of null result. When a compound produces a positive signal in one cohort, the appropriate scientific response is:
- Determine whether the effect was dose-specific or start-age-specific (positive at one parameter set, null at others)
- Determine whether it was a false positive at the original result’s p-value (ITP uses p<0.05 as a threshold, which will produce ~1 false positive per 20 compound-sex combinations even with perfect design)
A single positive cohort followed by a null replication cohort does not resolve these two possibilities cleanly. It establishes that the positive signal is not robust across parameter variation, which meaningfully raises the prior probability of a false positive or narrow-window effect. For astaxanthin specifically, the prior Harrison 2024 result carried p=0.003 with cross-site consistency — stronger evidence than the meclizine result at p=0.03 with site heterogeneity. The Korstanje 2026 null for both therefore updates them differently: astaxanthin moves from “one robust positive” to “one positive + one null at different conditions”; meclizine moves from “one fragile positive” to “likely false positive or highly dose-narrow.” gap/contradictory-evidence
Extrapolation to humans
| Dimension | Status | Notes |
|---|---|---|
| Pathway conserved in humans? | varies | Compound-specific; for none of these is the pathway to lifespan extension established even in mice |
| Phenotype conserved in humans? | unknown | Lifespan extension in heterogeneous mice has no confirmed human equivalent |
| Replicated in humans? | no | No aging lifespan trials exist for any of these compounds |
Limitations
- Abstract-only extraction: Per-compound p-values, median lifespan changes, site-specific effects, doses, and start ages are not available from the abstract alone. All compound-level quantitative claims in the full paper are currently unverifiable from this page. gap/no-fulltext-access
- “Different doses or starting ages” is the design rationale — the tested conditions in this cohort may not constitute the optimal dose or initiation window; a null result here does not rule out lifespan extension at the parameters used in Harrison 2024.
- N per group not confirmed. ITP-typical per-arm n is ~100–200/sex across three sites; the abstract does not report these values. gap/needs-n-confirmation
- Cause-specific mortality not measured in standard ITP lifespan protocols.
- Female JAX control anomaly is not explained in the abstract; whether it reflects a cohort-level environmental factor, batch effect, or statistical fluctuation is unknown and is relevant to interpreting all female results in this cohort.
- Mechanism unknown for the compounds that were null: whether null at these conditions reflects absent efficacy, wrong dose, wrong timing, or a genuine class-level failure is not determinable from a lifespan screen.
- The pioglitazone female harm could reflect PPAR-γ agonism–driven physiological effects (fluid retention, hormonal) specific to female UM-HET3 mice, or may be a statistical artefact at marginal p-value; full paper required to assess.
Gaps
- gap/no-fulltext-access — Per-compound doses, start ages, and lifespan statistics require the full PDF (CC-BY-NC-ND 4.0 but not available via web extraction at this time)
- gap/needs-n-confirmation — Per-arm n values not reported in abstract
- gap/contradictory-evidence — Astaxanthin: Harrison 2024 male +12% (p=0.003, cross-site consistent) vs Korstanje 2026 null at different dose/age; see astaxanthin for full discordance framing
- gap/contradictory-evidence — Meclizine: Harrison 2024 male +8% (p=0.03, site-heterogeneous) vs Korstanje 2026 null; see meclizine
- gap/no-mechanism — Female harm of late-start mitoglitazone and pioglitazone: confirmed in 2-site reanalysis but mechanism not addressed in abstract
- gap/dose-response-unclear — For all previously-positive compounds: the effective dose-age window (if any) is not defined by a single positive + single null data point
Cross-references
- harrison-2024-itp-astaxanthin-meclizine — the cohort this study fails to replicate; astaxanthin male +12% (p=0.003) + meclizine male +8% (p=0.03); verified: true
- astaxanthin — compound page; full ITP discordance framing; see “Korstanje 2026 — non-replication with harm signal” section
- meclizine — compound page; see “Korstanje 2026 — NULL at different dose or starting age”
- cellular-senescence — linked via astaxanthin mechanism hypothesis (Nrf2/oxidative-stress → senescence protection); not tested as a direct hallmark endpoint in this cohort
- mitochondrial-dysfunction — linked via astaxanthin/mitoglitazone/pioglitazone mechanism hypotheses
- deregulated-nutrient-sensing — alpha-ketoglutarate (AKG) is an mTOR-pathway metabolite; PPAR-γ agonism engages metabolic sensing
Cited by (anticipated downstream uses)
- astaxanthin — non-replication anchor; female harm signal nuance
- meclizine — non-replication anchor
- Any future pioglitazone, mitoglitazone, alpha-ketoglutarate, or mifepristone compound pages
Footnotes
-
korstanje-2026-itp-null-cohort · n=~100–200/sex/arm (ITP-typical; per-arm n unconfirmed, abstract-only) · in-vivo · model: UM-HET3 genetically heterogeneous mice, 3 sites (JAX/UT/UM) · eleven interventions including astaxanthin, meclizine, mitoglitazone, pioglitazone, AKG, mifepristone, methotrexate, atorvastatin+telmisartan · none extended lifespan in either sex · female harm in 3-site pooled analysis for astaxanthin + late-start mitoglitazone + pioglitazone; astaxanthin harm not confirmed in 2-site reanalysis (excl. JAX); mitoglitazone + pioglitazone harm confirmed in 2-site reanalysis · doi:10.1007/s11357-026-02201-2 · PMID 41843349 · GeroScience 2026 (online 2026-03-17) · abstract-only verification 2026-06-30 · gap/no-fulltext-access ↩ ↩2 ↩3 ↩4 ↩5 ↩6