Across 33,780 mice in the Interventions Testing Program, lifespan benefits from 14 candidate geroprotective compounds scaled with body weight in males. Each additional standard deviation of adult weight made treatment 6% more favourable (HR 0.94), and 0.90 for compounds that lowered weight. Median lifespan extension climbed from 6.2% in the lightest quartile to 20.2% in the heaviest, driven largely by control survival falling faster with weight than treated survival. The pattern held for 13 of 14 compounds, tracked adult rather than early-life weight, and was weak in females.
This is animal evidence, and a preprint that has not been peer reviewed, so conclusions may shift. The central implication is uncomfortable for the field: many headline lifespan extensions in ad libitum-fed mice may partly reflect rescuing heavy, metabolically unhealthy controls rather than slowing biological aging itself. That echoes long-standing criticism that overfed lab mice are a poor model of healthy aging, and it parallels the caloric-restriction literature, where benefits shrink against lean controls. For human readers, the takeaway is cautious rather than actionable: drugs such as rapamycin or metformin may show smaller gains in metabolically healthy people, and weight management may already capture some of the benefit. The analysis is retrospective and observational within experiments, weight is not randomized, and the weak female signal remains unexplained, though the authors suggest unusually long-lived controls. Confirmatory in spirit, but a meaningful challenge to how geroprotector results are interpreted.