Autophagy — the cell's recycling process, suppressed by active mTORC1 and switched on when mTORC1 is inhibited — is the textbook explanation for how rapamycin extends lifespan. But "autophagy goes up when you give rapamycin, and rapamycin extends lifespan" is a correlation, not proof that autophagy is the mechanism. This is one of the Atlas's ten flagged open questions: an evidence gap, not a settled fact.
18 studies in the Atlas touch autophagy directly (9 animal, 9 mechanistic/in vitro) — a substantial mechanistic case that autophagy changes when mTORC1 is inhibited.
What's missing is the causal experiment: block autophagy genetically (e.g. knock out an ATG gene) in an animal also given rapamycin, and see whether the lifespan benefit disappears. If it does, autophagy is required. If lifespan extends anyway, something else in the mTORC1 pathway is doing the work and autophagy is a correlated side effect, not the mechanism. This exact experiment is the Atlas's flagged evidence gap — read the full breakdown, including the proposed test, on the open question page: Is autophagy actually required for the mammalian lifespan benefit?
If autophagy isn't strictly required, then drugs or lifestyle interventions that boost autophagy through a completely different pathway (independent of mTOR) might not deliver the same longevity benefit rapamycin does — an assumption a lot of "autophagy-boosting" supplement marketing quietly skips over.
