The ITP runs at three independent sites in UM-HET3 mice, genetically heterogeneous, both sexes. A null result here carries more weight than a failed replication in one lab with one inbred strain. Worth keeping that in mind before deciding these results mean nothing. A few things worth reading alongside the primary results: On AKG: the original lifespan claim came from a 2021 study in inbred female mice at one site. ITP is the multi-site heterogeneous replication that study never had. The effect didn’t hold. This is the mechanism-doesn’t-equal-outcome pattern - alpha-ketoglutarate has plausible TCA cycle and mTOR-adjacent biology, but plausibility isn’t outcome. On the glitazone pair: mitoglitazone and pioglitazone failing together is more informative than either alone. Different mechanistic rationale (mitochondrial uncoupling vs. insulin sensitization), same outcome. Worth reading the ITP methods section on how intervention start age interacts with metabolic compounds - that timing variable gets underweighted in most commentary. On astaxanthin: antioxidant-lifespan in mammals has been a consistent series of non-results. C. elegans and Drosophila ROS biology doesn’t port to mice, let alone humans. Not surprising. The ITP results page on the NIA aging site is the primary source. Full paper will probably land in GeroScience or Aging Cell. One thing worth checking in the actual tables: sex-specific survival curves. ITP reports those separately, and a null overall can mask a directional signal in one sex. Not claiming that’s the case here, just worth looking before writing off any compound entirely.