There is no single longevity diet. The category has been claimed by carnivores, vegans, Mediterraneans, Okinawans, low-carbers, high-carbers, fasters, and grazers, each pointing at a subset of the evidence that supports them. This is not because the science is silent. It is because the honest answer, once you strip away marketing, is that several patterns work reasonably well and no pattern has been shown to be dramatically superior in humans.
Recognizing what the different diets share is more useful than picking a winner.
What the strongest human evidence supports
The dietary pattern with the deepest human evidence base for cardiovascular mortality is the Mediterranean pattern, largely because of trials like PREDIMED, which randomized thousands of adults at cardiovascular risk to Mediterranean diets supplemented with olive oil or nuts and found reductions in major cardiovascular events compared to a lower-fat control.
Other patterns with defensible longitudinal data include:
- The DASH diet, developed for hypertension, with strong effects on blood pressure and reasonable outcome data
- Traditional plant-forward diets across cultures, associated with lower all-cause mortality in cohort studies
- The MIND diet, a hybrid of Mediterranean and DASH designed for cognitive endpoints
What these patterns share is more revealing than what separates them: heavy on vegetables, legumes, and whole grains; regular fish; olive oil or other unsaturated fats as the main added fat; modest red and processed meat; limited ultra-processed food; nuts and seeds daily; alcohol either absent or moderate.
The dietary features that repeatedly correlate with lower mortality are also the boring ones. There is no exotic ingredient or timing trick that outperforms the basics.
Where the debates actually live
Most internet food arguments are about second-order questions. Are seed oils harmful? Is saturated fat rehabilitated? Is fruit sugar as bad as added sugar? Should you eat only during an eight-hour window?
The honest reading of the literature on most of these is that the effect sizes, if they exist, are small compared to the effect of overall dietary pattern. A person eating whole foods, mostly plants, in reasonable quantities is unlikely to see their outcomes materially altered by whether they cook in olive oil or avocado oil. A person eating primarily ultra-processed food will not be rescued by switching cooking oils.
The growing body of work on ultra-processed foods is one area where the effect size does appear meaningfully larger. Cohort studies and a small NIH inpatient trial by Kevin Hall have shown that ultra-processed foods drive greater caloric intake and worse metabolic profiles even when nutrients are matched. This is currently the most defensible single dietary target.
Protein and older adults
One place where the standard advice is likely wrong is protein for older adults. Standard RDAs, set to prevent deficiency, appear inadequate for preserving muscle mass in aging. Most geriatric nutrition experts now recommend intakes closer to 1.2 to 1.6 grams per kilogram per day for older adults, higher for those doing resistance training or recovering from illness. Concerns about kidney load in people with normal kidney function have not been substantiated in trials.
What the fringe patterns get right and wrong
Low-carb and ketogenic diets improve short-term glycemic control and can drive weight loss, both of which likely improve longevity outcomes. Long-term adherence is difficult and long-term mortality data are limited. Carnivore diets have essentially no long-term human evidence and depart sharply from any observed high-longevity population. Fasting protocols, in trials, appear to produce metabolic benefits that are largely mediated by weight loss, not by any unique fasting mechanism.
Specific supplements marketed for longevity, from resveratrol to NAD precursors, have not been shown to change hard outcomes in humans, though several remain under study.
The bottom line
The evidence points to a small number of dietary patterns that mostly agree with each other: plant-forward, minimally processed, adequate protein especially with age, moderate or no alcohol. The disagreements at the margins get most of the attention and matter less than the shared core. If you already eat this way, further optimization has diminishing returns.