The mouse lifespan data behind rapamycin's reputation as a longevity drug didn't use one fixed protocol. Two separate findings matter: rapamycin still works when started late in life (not just from birth), and it still works when given only briefly rather than continuously. Neither finding has been replicated with a matched dosing protocol in a human lifespan trial — that trial doesn't exist and may never be practical to run.
Late-onset, continuous dosing
HAR2009C — rapamycin fed starting at 600 days of age — late-middle-age for a mouse — extended median lifespan by 9–14% in both sexes. This was the finding that established rapamycin doesn't need to start early to work.
Brief, late-life dosing
BIT2016C — just 3 months of rapamycin, given late in life, increased subsequent life expectancy by up to 60% — the lifespan benefit doesn't require lifelong daily dosing, at least in mice.
Why intermittent/pulsed dosing is a live research question, not settled practice
If a few months of treatment can extend lifespan as much as continuous dosing in mice, the practical implication for humans is large: intermittent dosing could in principle deliver most of the benefit while limiting cumulative exposure to side effects like the mTORC2-linked insulin resistance discussed in the Atlas's side-effects answer page. But the specific schedule — how often, how much, for how long — hasn't been established in humans at longevity-relevant doses. This is tracked directly by two of the Atlas's open questions: muscle-sparing pulsed mTORC1 inhibition and mTORC1-selective, mTORC2-sparing dosing.
What human trials have actually tested
KRA2018B — a safety-first pilot RCT in older adults (n=25, ages 70–95) testing daily low-dose rapamycin over 8+ weeks for basic tolerability — not a lifespan endpoint.
MOE2025B — the first completed long-term RCT of rapamycin for healthy human aging (48 weeks, n=114) — still not a lifespan endpoint, and its primary metabolic endpoint showed no significant change.
