Physiologist Mike Lustgarten says the 9% increase in median mouse lifespan from bone loading is overstated: control mice had a median lifespan of 846 days, compared with about 900 days for healthy mice under the “900-day rule”

Physiologist Mike Lustgarten says the 9% increase in median mouse lifespan from bone loading is overstated: control mice had a median lifespan of 846 days, compared with about 900 days for healthy mice under the “900-day rule”

The discussion began with a post by cardiologist Eric Topol, who described the gain as equivalent to extending human life by nine years. On November 2, Lustgarten compared the control group, mice that received no intervention, with aging researcher Kamil Pabis’s “900-day rule.”

A study from Southern Medical University was published in Cell Metabolism on October 23. Twice a week for four months, a device applied pressure to the ends of the tibiae in 20-month-old C57BL/6 mice. Median lifespan increased by 9,4% (79,5 days). These figures imply a median lifespan of about 846 days in the control group and about 925 days in the bone-loading group. Lustgarten responded:

“The mice that underwent bone loading had a median lifespan of 925 days, only slightly above the benchmark used in the 900-day rule for assessing mouse longevity. The control mice fell short of 900 days (846), so the 9% increase in median lifespan is overstated. Nevertheless, bone loading (that is, weight-bearing exercise) should still be part of everyone’s approach to health and longevity!”

Pabis and colleagues proposed the rule in a 2024 paper. They analyzed lifespan-extension experiments and found that shorter control lifespans were associated with larger percentage gains. In an early version of the paper, posted before peer review, they offered one explanation for this association: an intervention might improve the health of animals whose lives had been shortened by disease, housing conditions, or genetic factors. In that case, the intervention would simply restore the animals’ lifespan to that of healthy mice, while aging continued at the same rate.

The authors argue that most healthy mouse strains, including C57BL/6, should have a median lifespan of about 900 days, plus or minus 50 (850 to 950), which they consider the upper limit of normal lifespan. Under the rule, high confidence in an effect is possible only when control mice live for about 900 days or treated mice live substantially longer than that. Pabis explained the reasoning in a talk:

“There are millions of ways to harm a mouse and cause it to die prematurely, but there is only one way for a mouse to live a truly long life: slow its aging.”

In a 2024 paper, Leonid Peshkin of Harvard Medical School and colleagues compiled data on male C57BL/6J control groups from other studies that reported housing conditions in detail. Median lifespans in these groups ranged from 800 to 970 days. In most established lifespan-extension experiments in males, treated mice lived substantially longer than their own controls, but their lifespans did not exceed the highest of these control medians. The drug rapamycin was the exception.

The same comparison with the rule can be made for an experiment involving a vaccine against senescent cells, published in Advanced Science: 75-week-old C57BL/6 mice received the vaccine, while the control group received saline. Median lifespan was 113 weeks (791 days) in vaccinated mice and 101 weeks (707 days) in mice given saline.

The gains in days were almost equal: 84 days for the vaccine and 79,5 days for bone loading. The percentage gains were 11,9 and 9,4: a shorter control lifespan (707 days in the vaccine experiment and 846 in the bone-loading experiment) produced a larger percentage increase for almost the same number of additional days. The control group in the bone-loading experiment fell 4 days short of the lower bound of the 850 to 950-day range; the vaccine control group fell short by 143 days. The bone-loading group (925 days) was within that range, while the vaccinated group (791) was below it.

On percentage gains, Pabis and colleagues wrote in the early version of their paper: “We advise against reporting percentage increases in lifespan because they fail to convey, and often conceal, essential information about control lifespan.”