Cut calories without malnutrition and animals live longer — that result has held up across yeast, worms, flies, and rodents for the better part of a century. The obvious question is whether it transfers to people. You cannot run a decades-long survival trial in humans, so the field did the next best thing: CALERIE, the Comprehensive Assessment of Long-term Effects of Reducing Intake of Energy. It is the most rigorous randomized test of sustained caloric restriction ever run in healthy, non-obese adults, and reading it honestly means separating what it actually showed from what the headlines wanted it to show.
Why it mattered
Caloric restriction is the reference intervention in aging research: reliably extend lifespan in multiple species and you have something worth reverse-engineering. But animal restriction protocols are severe — often 30–40% below normal intake — and come with trade-offs in growth, reproduction, and immune defense that no one would prescribe to a healthy person.[3] CALERIE was designed to ask a narrower, answerable question: if healthy adults sustain meaningful restriction for two years, does their aging biology actually move in the direction the animal data predict?[1]It is the human bridge for a hypothesis that also drives interest in rapamycin and metformin as drugs that might mimic restriction without the hunger.
The design
CALERIE-2 randomized roughly 220 healthy, non-obese adults across three US sites, about two to one, to either a 25% caloric-restriction target or an ad libitum (eat-as-usual) control diet, for two full years.[1] This was not a weight-loss study in people with obesity — participants started at normal or modestly overweight body weight, which is precisely what makes it a clean test of restriction itself rather than of reversing metabolic disease. The trial tracked a broad panel of outcomes: body composition, cardiometabolic markers, immune and inflammatory measures, and later, from stored samples, molecular measures of biological aging.
What the trial found
Cardiometabolic risk. Even though participants were healthy to begin with, sustained restriction improved the standard cardiometabolic risk factors — LDL and total cholesterol, blood pressure, C-reactive protein, and insulin-sensitivity measures all moved favorably relative to control.[2] The signal was consistent across markers rather than resting on any single number.
Immune and anti-inflammatory biology. A mechanistic analysis found that about 14% restriction over two years improved thymopoiesis — the thymus's production of new T cells, which normally declines steeply with age — alongside mobilization of the fat that accumulates inside the aging thymus.[3] Restriction also reprogrammed fat tissue toward an anti-inflammatory state and suppressed a gene, PLA2G7, encoding a pro-inflammatory protein. Deleting that gene in mice reproduced several restriction-like benefits, including lower age-related inflammation and inflammasome activity.[3] That points at a druggable target — the same logic behind anti-inflammatory longevity candidates like spermidine.
Pace of aging. The most-cited result came from a post-hoc DNA-methylation analysis. Restriction produced a small but statistically detectable slowing of the pace of aging, measured by the DunedinPACE algorithm — but it did not significantly change the epigenetic clocks that estimate biological age and predict disease and death, such as PhenoAge and GrimAge, and the authors described the effect sizes as small.[4] A slower rate of aging is a genuinely interesting readout; it is also a biomarker, not a lifespan.
The honest limitations
Start with the number in the hero: the target was 25%, but participants sustained only about 12% restriction on average over the two years — roughly half of what was prescribed, and it drifted upward as the trial went on.[2] That is not a footnote. It means CALERIE tested a milder intervention than the one animal lifespan studies use, so the human effects it found are arguably the floor of what full-dose restriction might do — and also a blunt reminder that sustained restriction is genuinely hard to maintain. Beyond adherence: the population was healthy and non-obese, so the results may not generalize to everyone; two years is a sliver of a human lifespan; and CALERIE measured biomarkers, not hard outcomes. There is no mortality data, no disease-endpoint data, and no claim — from the investigators or from us — that anyone in the trial will live longer.
What it means for CR-mimetic drugs
Here is the throughline. CALERIE established that restriction moves human aging biology in a favorable direction — cardiometabolic risk, inflammation, thymic function, pace of aging — while also showing that almost no one can sustain the restriction that animal studies demand. That combination is the entire rationale for CR-mimetics: molecules that trip the same nutrient-sensing pathways restriction acts on, without requiring lifelong hunger. Rapamycin (mTOR) and metformin (AMPK/ metabolic signaling) are the headline candidates, and the specific finding that suppressing PLA2G7 reproduces restriction-like benefits is exactly the kind of target this line of work is hunting for.[3] But a mimetic can only ever be as good as the thing it mimics, and CALERIE mimics a two-year, half-dose, biomarker-level benefit — not a proven extension of human life. This is a research explainer, not medical or dietary advice; sustained caloric restriction has real risks and is not something to attempt without clinical supervision.