What's the best longevity diet?

This week the newsletter is answering three of the questions we keep getting in reply. Wednesday's is really two questions in one: should I be vegan after 50, and what actually is the best diet for longevity?

Short honest answer. Plant-forward Mediterranean, with fish 2 to 3 times a week and enough protein to hold your muscle. Not strict vegan. Not keto or carnivore. That is the center longevity researchers have kept returning to for two decades, and the evidence backing it up has only gotten stronger while the loud dietary waves on either side of it have receded.

The most-cited cohort supporting a plant-based case for longevity is Adventist Health Study 2, Michael Orlich and colleagues at Loma Linda, JAMA Internal Medicine 2013. 73,308 Seventh-day Adventists tracked for around 6 years. The result the vegan wave built on is that vegetarians and vegans had lower all-cause mortality than non-vegetarians. What the marketing rarely mentions is which vegetarian pattern actually had the strongest signal. It was the pesco-vegetarian arm. Vegans did well. Pesco-vegetarians did better.

The larger and longer-running British replication is EPIC-Oxford, 65,000 adults tracked since the 1990s at the University of Oxford. All-cause mortality did not differ between vegetarians and meat-eaters, hazard ratio 1.00. Vegans were slightly elevated at 1.14, though not to statistical significance. Vegetarians did carry a higher risk of stroke and bone fractures. Not the story anyone wanted, but the story the data told.

The Blue Zones have been reflexively cited as evidence of plant-based longevity for a decade. The actual dietary intake surveys from those regions tell a more specific story. Sardinian centenarians eat roughly 47% grains, 26% dairy, 12% vegetables, 5% meat and fish. Okinawan centenarians historically ate more animal protein than the contemporary Japanese average. Nicoyan diets are built on beans and corn but include pork, eggs, and dairy weekly. "95 to 100% plant-based" is a book positioning. It is not what the source surveys actually measured. The actual pattern across Blue Zones is plant-forward, not plant-only.

And the gold-standard randomised trial for cardiovascular longevity is PREDIMED, Ramón Estruch and colleagues, NEJM 2018 (republished after a randomisation correction). 7,447 high-risk Spanish adults. Mediterranean diet with extra-virgin olive oil or nuts cut heart attack, stroke, and cardiovascular death by 31% versus a low-fat control. This is the strongest RCT evidence for any dietary pattern in longevity research. Mediterranean is plant-forward. It includes fish, eggs, dairy and some meat.

Meanwhile the pendulum swung the other way in the early 2020s. Keto went mainstream, carnivore extremism arrived alongside it. The American Heart Association's expert review panel ranked ketogenic diets the worst of all evaluated patterns for cardiovascular disease prevention. Carnivore trials that have run so far show LDL cholesterol moving from a median of 157 to 256 mg/dL in one 2025 German cohort. There is no long-term cardiovascular outcome data for carnivore. Short-term "I feel great" reports are the elimination-diet effect plus weight loss, not the meat doing the work.

Valter Longo at USC, who has run more human fasting-mimicking-diet trials than anyone else, recommends a primarily plant-based diet with fish 2 to 3 times a week, low-mercury species only, and after age 65 explicitly more protein from fish, eggs, and white meat to preserve muscle. His framework is pescatarian, not vegan. Walter Willett at Harvard, whose Mediterranean-adherence work underpins the Harvard Healthy Eating Plate, recommends the same essential architecture. Peter Attia, from the protein-forward end of the longevity conversation, argues for 1.6 g/kg of protein daily in most adults from a plant-forward base plus regular fish and lean meat.

Three researchers approaching the question from three different angles converge on the same plate. Legumes, whole grains, nuts, seeds, and vegetables carry the majority of it. Fish 2 to 3 times a week. Eggs and modest dairy. Meat as an occasional food. Enough protein to hold muscle, which after 50 means more than the 0.8 g/kg RDA, closer to 1.2 to 1.6.

Plant-forward Mediterranean is the defensible longevity framework at 50 and beyond. It has the strongest RCT evidence, the strongest cohort convergence, and the endorsement of the researchers with the most skin in the game.

Strict vegan can still work after 50, but the execution bar is higher. You need explicit protein planning at 1.2 to 1.6 g/kg from a leucine-adequate blend of legumes, tofu, tempeh, seitan, and plant protein powder. Reliable B12 supplementation. Algae-sourced omega-3. Attention to iron and iodine. If you are drawn to veganism for ethical or environmental reasons, those are legitimate reasons, and the diet is compatible with the longevity goal if the planning is done. It is not a shortcut to it.

Keto and carnivore have no long-term longevity evidence and mounting cardiovascular concerns. Short-term metabolic improvements are real for some people. They come from the elimination effect and the weight loss, not the meat.

Friday takes the third reader question of the week. Bone health after 50. What actually protects density, what does not, and where the DEXA line sits for adults who want to still be moving well at 80.

Until Friday.

Longevity Daily / The Building Decades