Nutrition and metabolism · Research Brief
CALERIE tested calorie restriction for two years—not human lifespan
Healthy, non-obese adults achieved about 12% calorie restriction and improved several cardiometabolic risk factors. The trial did not measure lifespan, and its DNA-methylation aging results depended on which clock was used.
Source, correction-status, evidence, safety and prose reviews were completed July 31, 2026.
Evidence box
Promising intermediate outcomes; no lifespan evidence
- Study type
- Two-year randomized controlled trial with secondary analyses
- Studied in
- 218 healthy, non-obese adults aged 21–51
- Participants / sample
- 143 assigned calorie restriction; 75 usual diet; 197 in methylation analysis
- Publication status
- Peer reviewed; methylation paper corrected
- Outcome type
- Metabolism, cardiometabolic risk factors and biomarkers—not mortality
- Development stage
- Human lifestyle intervention research
- Conflicts / funding
- NIH-funded cooperative agreement; paper-specific disclosures checked
- Our assessment
- Promising for selected risk factors; insufficient for longevity claims
The bottom line
CALERIE is one of the strongest randomized human tests of sustained calorie restriction in people without obesity. Participants assigned restriction achieved about 12% fewer calories—not the 25% target—and maintained roughly 10% weight loss over two years. Several cardiometabolic risk factors improved, but the trial was too short and not designed to determine whether anyone lived longer. (primary trial; cardiometabolic analysis)
Why this matters
Calorie restriction extends lifespan in some laboratory organisms, making translation to people a central geroscience question. Human evidence must separate feasibility, weight loss, risk factors and aging biomarkers from disease, disability and mortality outcomes. CALERIE informs the first three; it does not answer the last two.
What researchers did
The multicenter trial randomized 218 adults aged 21–51 with body-mass index from 22.0 to 27.9. The intervention aimed for 25% restriction without prescribing one diet. Eighty-two percent of the restriction group and 95% of controls completed two years. Primary metabolic outcomes were resting-metabolic-rate residual and core temperature; other papers analyzed cardiometabolic factors and DNA methylation.
What they found
The restriction group achieved an average 11.7% reduction in calories and lost 10.4% of body weight. Resting-metabolic-rate residual differed at 12 months but not at 24 months; core temperature changed little. An exploratory analysis reported improvements across measures including blood pressure, lipids, insulin sensitivity and inflammation. A later methylation analysis found a small slowing on DunedinPACE but no significant change in PhenoAge or GrimAge. That paper has an author correction, which is linked below. (methylation analysis; correction)
How strong is the evidence?
Randomization, two years of follow-up and careful metabolic measurement are substantial strengths. But participants were selected healthy volunteers, achieved less restriction than planned, and weight loss is inseparable from many measured changes. The clock analysis was secondary and produces different answers across algorithms. These results support some intermediate benefits in this population, not a validated anti-aging effect.
What this study does not show
CALERIE did not test mortality or lifespan and cannot show that restriction delays every disease, benefits older or frail adults, or is safe for everyone. It does not prove that changing DunedinPACE changes clinical outcomes. Findings from laboratory animals cannot fill those human-outcome gaps.
Safety and conflicts
Sustained restriction can be inappropriate for people who are underweight, pregnant, growing, frail, recovering from illness, or vulnerable to eating disorders. CALERIE used intensive monitoring and nutrition support, so unsupervised restriction is not equivalent. The program was supported by the US National Institute on Aging; readers should consult each linked paper for complete author disclosures.
What happens next
Longer follow-up can test whether risk-factor differences persist and relate to meaningful health outcomes. Replication should predefine the biomarker, population and clinical endpoint, and should report nutritional harms alongside benefits. A clock result should remain a secondary biomarker unless it is validated as a surrogate for outcomes people care about.
Primary sources
Sources, roles and limits
- CALERIE Phase 2 registry
- Identifier
- NCT00427193
- Role
- Protocol and status
- Limitation
- Registration is not government validation or evidence of benefit.
- Ravussin et al., two-year CALERIE trial
- Identifier
- PMID:26187233 · PMCID:PMC4841173 · DOI:10.1093/gerona/glv057
- Role
- Primary randomized results
- Limitation
- Intermediate metabolic outcomes; no lifespan endpoint.
- Kraus et al., cardiometabolic secondary analysis
- Identifier
- PMID:31303390 · PMCID:PMC6707879 · DOI:10.1016/S2213-8587(19)30151-2
- Role
- Exploratory risk-factor outcomes
- Limitation
- Multiple secondary outcomes, not clinical events.
- Waziry et al., DNA-methylation secondary analysis
- Identifier
- PMID:37118425 · DOI:10.1038/s43587-022-00357-y
- Role
- Aging-biomarker analysis
- Limitation
- Small effect on one measure; two other clocks null; corrected publication.
- Author correction to the methylation analysis
- Identifier
- PMID:37161091
- Role
- Correction record
- Checked
- 2026-07-31
Connected topics
Continue through the knowledge system
Disclosures and history
- July 31, 2026 — three research passes and separate evidence/prose reviews completed.
- July 31, 2026 — author correction linked and standing publication authorization applied.