No intervention has generated as much data — or as much disagreement — as cutting calories. The caloric restriction longevity story is almost ninety years old, spans rats, monkeys, and humans, and remains the single most reproducible way to extend lifespan in laboratory animals. And yet in 2026 we still cannot tell a 45-year-old whether 20 percent fewer calories will let them live meaningfully longer. This article walks through the evidence, from the founding McCay experiments to the CALERIE human trial and the popular intermittent-fasting protocols, and tries to give you an honest picture of what the science actually supports.
What Is Caloric Restriction?
Caloric restriction (CR) means chronically eating fewer calories — typically 20 to 40 percent below ad libitum intake — without malnutrition. Every essential nutrient has to be met; only energy is reduced. This is distinct from starvation, fasting mimicking diets, and most forms of intermittent fasting, though the pathways overlap.
Modern usage tends to lump several distinct practices together:
- Chronic CR. Eat every day but at a reduced caloric intake. The classic rodent protocol.
- Time-restricted eating (TRE). Eat within a narrow daily window (commonly 8, 10, or 12 hours). Popular as "16:8."
- Intermittent fasting (IF). Alternate-day fasting (ADF), 5:2 (two low-calorie days per week), or one-meal-a-day (OMAD).
- Prolonged fasting and fasting-mimicking diets (FMD). Multi-day interventions, done occasionally, designed to trigger deep metabolic and autophagic responses. Valter Longo's ProLon is the best-known commercial FMD.
Each has different adherence profiles and different mechanistic fingerprints.
The Science: Shared Mechanisms
Despite their surface differences, these protocols converge on a small number of nutrient-sensing pathways that are deeply tied to aging biology:
- mTOR down. Reduced amino acids and insulin lower mTORC1 signaling, which slows anabolism and lifts the brakes on autophagy. See our autophagy deep dive.
- AMPK up. A low energy state raises the AMP:ATP ratio and activates AMPK, which tilts metabolism toward catabolism and mitochondrial biogenesis.
- IGF-1 down. Insulin-like growth factor signaling drops, echoing the long-lived dwarf mouse models.
- Autophagy up. Both through mTOR inhibition and AMPK activation.
- Sirtuin activation. NAD+ levels rise during fasting, modestly activating SIRT1 and SIRT3.
- Ketogenesis. After roughly 12 to 16 hours without food, the liver starts producing beta-hydroxybutyrate, which acts as both a fuel and a signaling molecule inhibiting HDACs.
This cluster overlaps remarkably with the profiles produced by rapamycin and metformin — which is why people often talk about "CR mimetics."
The Evidence
The rodent foundation
In 1935, Clive McCay and colleagues at Cornell fed rats a diet that was calorically restricted but nutritionally complete and showed that restricted rats lived dramatically longer than controls. This was the first demonstration that lifespan could be deliberately extended, and it launched the field.
Roy Walford and Richard Weindruch at UCLA extended this work through the 1970s and 1980s, showing that CR worked in mice as well, reduced cancer incidence, and delayed virtually every age-related pathology examined. Weindruch's 1988 book "The Retardation of Aging and Disease by Dietary Restriction" is still worth reading.
Rodent CR consistently produces 10 to 40 percent lifespan extension depending on strain, age of onset, and severity. More recent work, particularly from Stephen Spindler and the NIA's Interventions Testing Program, has shown that the magnitude of benefit varies enormously across genetic backgrounds, which foreshadowed the surprises to come in primates.
The rhesus monkey trials
Two long-running CR studies in rhesus macaques — one at the University of Wisconsin (started 1989) and one at the NIA (started 1987) — delivered what looked, at first, like contradictory results.
In 2009, the Wisconsin group (Colman et al., Science) reported that CR monkeys had lower age-related mortality and reduced incidence of diabetes, cancer, and cardiovascular disease. In 2012, the NIA group (Mattison et al., Nature) reported no statistically significant lifespan extension, even though CR monkeys also showed metabolic benefits.
For several years this looked like a major blow to the field. Then in 2017, both groups published a joint analysis in Nature Communications (Mattison, Colman et al.) that reconciled the differences. The discrepancies came down to diet composition (the NIA control diet was healthier, less calorically dense, lower in sugar), age of onset (younger-onset monkeys responded differently), and control feeding practices (the NIA controls were not truly ad libitum). Taken together, the two studies showed that sustained CR improves healthspan and reduces age-related disease in primates, even if the lifespan signal was more modest than the rodent data would have predicted.
The CALERIE human trial
CALERIE (Comprehensive Assessment of Long-term Effects of Reducing Intake of Energy) is the best human CR evidence we have. Non-obese adults were randomized to 25 percent calorie reduction for two years. Most participants achieved closer to 12 percent — adherence is hard.
Ravussin and colleagues (2015, Journals of Gerontology) showed reduced metabolic rate, lower oxidative stress markers, improved cardiometabolic risk factors, and reduced inflammation.
In 2023, Daniel Belsky's group at Columbia published a landmark follow-up in Nature Aging using the DunedinPACE epigenetic clock, showing that CALERIE participants had a measurable slowing of biological aging pace — a small but statistically significant effect. It is the first human RCT evidence that caloric restriction slows the rate of biological aging measured by a molecular clock.
Intermittent fasting human trials
Intermittent fasting has the advantage of much better adherence, but the human outcome data is mixed. Mark Mattson at NIH pioneered much of the mechanistic IF work and wrote influential 2019 reviews in the New England Journal of Medicine. Satchin Panda at the Salk Institute championed time-restricted eating based on circadian biology and published the first human TRE trials showing metabolic benefits independent of calorie reduction. Valter Longo at USC developed the fasting-mimicking diet concept and has small RCTs showing improvements in biomarkers after monthly 5-day cycles.
The honest summary of IF RCTs: modest weight loss, modest metabolic improvements, and — in most head-to-head trials — no clear advantage over conventional calorie restriction when total calories are matched.
Current Interventions: What People Are Actually Doing
- 16:8 time-restricted eating. By far the most popular. Skip breakfast or dinner; eat in an eight-hour window. Easy adherence, real but modest metabolic effects.
- 5:2. Two non-consecutive days per week at about 500 to 600 calories. Better adherence than chronic CR for most people.
- Alternate-day fasting. Harder to sustain; effective for weight loss but with higher dropout rates in trials.
- Fasting-mimicking diet (ProLon). A commercial 5-day kit done once a month or quarterly. Designed by Longo; small trials show biomarker effects.
- Chronic 10 to 15 percent CR. The protocol closest to CALERIE. Very hard without structure; risk of micronutrient deficiency and muscle loss if protein is not prioritized.
The biohacker meta-pattern is a daily 10 to 12-hour eating window, frequent exercise, high protein intake, and occasional multi-day fasts — essentially borrowing the best of each approach while trying to avoid the muscle-mass downsides.
Connection to Gene Editing and Peptides
The search for CR mimetics is really a search for drugs and therapies that trigger the same nutrient-sensing signature without requiring hunger. Rapamycin is the closest thing we have. Metformin partially overlaps. Acarbose, canagliflozin, and NAD+ precursors each hit fragments of the signature. See the NMN/NR NAD precursors evidence review for the adjacent NAD story.
Gene editing plays a more speculative role. Long-lived dwarf mice (Ames, Snell, growth-hormone-receptor knockouts) show the extreme of reduced IGF-1 signaling, and the Laron syndrome human population — naturally IGF-1-deficient due to GHR mutations — has remarkably low rates of cancer and diabetes. In principle, modest base-editing of nutrient-sensing pathways could mimic CR without the diet, but this is laboratory-stage thinking, not clinical practice.
On the peptide side, GLP-1 receptor agonists (semaglutide, tirzepatide) are now the dominant practical tool for achieving sustained caloric deficit in free-living humans. They are, in effect, pharmacologically enforced CR. Whether their longevity effects recapitulate real CR will take years of follow-up to know.
Limitations and Open Questions
Adherence is the biggest problem. CALERIE enrolled highly motivated adults and only achieved about half the intended restriction. Real-world adherence is worse.
Muscle mass. CR without adequate protein and resistance training causes lean mass loss, which is the opposite of what you want for older adults. This is the main reason most longevity clinicians now recommend higher protein (1.2 to 1.6 g/kg) and strength training alongside any caloric deficit.
Genetic background matters. The ITP data and the monkey reconciliation both suggest that CR's magnitude of effect depends on genotype. Population-average results may understate or overstate the effect in any given individual.
We still do not have a human CR lifespan trial. CALERIE measured biomarkers and biological age, not death. The 20-plus year follow-up studies that could answer the question are only now starting to accumulate.
The "just eat less, live longer" shortcut is not supported. In humans, the curve linking BMI to mortality is J-shaped, and underweight older adults have higher mortality than moderately overweight ones. The monkey and human data suggest benefits from modest restriction in metabolically healthy adults, not a linear "less is always more."
FAQ
Does intermittent fasting extend lifespan in humans?
No human trial has measured lifespan endpoints. IF improves metabolic markers and modestly slows biological age proxies, which is suggestive but not proof.
Is 16:8 as good as 25 percent CR?
Probably not, if the only thing 16:8 does is shift meal timing without changing total calories. Several RCTs have found that TRE without calorie reduction produces minimal metabolic change. Its main benefit seems to be as a simple tool for calorie control.
What was the Wisconsin vs. NIA monkey conflict?
It looked like a contradiction — one study showed CR extended lifespan, one did not — but the 2017 joint analysis showed it came down to diet composition, age of onset, and control feeding. Both studies now support healthspan benefits.
Is CALERIE the best human CR evidence we have?
Yes. It is the only long-term randomized trial of sustained calorie restriction in healthy non-obese adults. The Belsky 2023 Nature Aging paper on biological aging pace is the strongest molecular-level outcome data.
Can I just take rapamycin instead?
Rapamycin hits one of the main CR pathways (mTOR) more cleanly than diet, but it does not replicate the full CR signature (AMPK, autophagy, ketogenesis, inflammation). It is complementary, not equivalent.
Should older adults do CR?
Most longevity clinicians advise against aggressive caloric restriction in adults over 65 because of muscle and bone loss risks. Protein-preserving strategies, exercise, and mild time-restricted eating are more defensible in this group.
