Plant or Animal Protein After 70? The Mortality Split

After 70, the usual advice to swap animal protein for plant protein runs into an awkward fact: the two dedicated mortality cohorts in older adults point in opposite directions. One linked a higher share of animal protein to lower mortality; the other linked each additional animal-protein serving to higher mortality. Neither study can prove that protein source caused either outcome. Together, they show that nutrition advice built largely in younger and middle-aged adults does not settle the question for people in their seventies and eighties.

The practical answer is less dramatic but more useful. Older adults need enough total protein, enough essential amino acids, and resistance training when medically appropriate. Protein source can affect how easily a meal delivers those amino acids, especially leucine, but a planned plant-based diet can meet the anabolic challenge in active older adults. The mortality evidence does not justify declaring animal protein either a longevity treatment or a dietary villain.

Two cohorts, opposite mortality signals

In Tuscany, the InCHIANTI Study followed 1,139 community-dwelling adults with a mean age of 75; 56% were women. Participants completed a validated food-frequency questionnaire, and researchers followed them for as long as 20 years. Each one-percentage-point increase in animal protein as a share of total energy intake was associated with a 4% lower risk of death from any cause (hazard ratio [HR] 0.96; 95% confidence interval [CI], 0.93–0.99) and a 7% lower risk of cardiovascular death (HR 0.93; 95% CI, 0.87–0.98). Both estimates were statistically significant. Plant protein showed no statistically significant association with mortality in this cohort.

That result does not establish that eating more animal protein prolonged life. Diets high in animal protein can differ in many other ways: food preparation, total diet quality, physical function, income, illness, and eating habits. Still, the long follow-up gives the study more protection than a short study against reverse causation—the problem in which illness changes diet before death rather than diet changing mortality risk.

Then comes the Concord Health and Ageing in Men Project, or CHAMP, in Sydney. This prospective cohort enrolled 794 older men, with a mean age of about 81, used a validated diet-history interview, and followed participants for a median of 3.7 years. Its findings ran in the other direction. Each additional daily serving of animal protein was associated with 12% higher all-cause mortality (HR 1.12; 95% CI, 1.00–1.26) and 23% higher cancer mortality (HR 1.23; 95% CI, 1.02–1.49). Each additional serving of plant protein was associated with 25% lower all-cause mortality (HR 0.75; 95% CI, 0.61–0.92).

The lower boundary of CHAMP's all-cause estimate sits exactly at 1.00, so that result is statistically borderline. The cancer and plant-protein estimates exclude the null. Even so, these are relative associations, not absolute risk differences, and they cannot tell us how much any individual's chance of dying changed.

The honest conclusion is a contradiction, not a tidy average: InCHIANTI linked animal protein with lower mortality, while CHAMP linked it with higher mortality in older Australian men.

Why the studies cannot be blended into one answer

The cohorts differ on nearly every feature that could shape the result. InCHIANTI measured protein as a percentage of total energy, included women and men, and took place in a Mediterranean Italian setting. CHAMP counted daily servings, included men only, and studied an Australian population roughly six years older on average. Their hazard ratios therefore do not measure the same dietary change.

CHAMP's shorter follow-up also leaves more room for reverse causation. At age 81, pre-existing disease, frailty, appetite changes, and reduced ability to shop or cook can alter food choices over a few years. It is plausible, though unproven, that declining health pushed some men toward particular animal-based foods. Conversely, InCHIANTI's 20-year window cannot eliminate confounding, but it provides a much larger temporal buffer.

Food context may matter as much as the protein label. A systematic review on ultra-processed foods, frailty, and sarcopenia found that higher ultra-processed-food consumption was associated with frailty and sarcopenia risk. Neither mortality cohort reportedly adjusted for ultra-processed foods as a separate variable. Processed meats and processed dairy products could carry different dietary co-exposures than fish, eggs, yogurt, beans, or tofu. That is a plausible explanation for heterogeneous findings, not proof that processing caused either cohort's result.

Aging muscle raises the protein-quality question

Older muscle responds less strongly to amino acids and exercise than younger muscle, a phenomenon called anabolic resistance. Research on habitual protein intake and muscle physiology across age describes this reduced muscle-protein-synthesis response and the importance of adequate essential amino acids for preserving muscle.

Leucine is one of the amino acids involved in triggering muscle protein synthesis. In a cross-sectional study, Lixandrão and colleagues found that higher daily leucine intake was associated with greater lower-limb muscle mass and strength in older adults. Association is not proof that leucine caused stronger muscles; people with better diets and more muscle may differ in many other ways. A systematic review of post-exercise muscle-protein-synthesis studies found that ingested leucine dose was associated with the size of the response in older adults. Mechanistic estimates often place the older-adult per-meal leucine threshold around 2.5 to 3 grams, versus 1.5 to 2 grams in younger adults, but no trial here directly establishes a universal threshold for every person over 65.

Animal proteins such as whey, eggs, fish, poultry, and meat generally deliver more leucine per gram than most plant proteins. That makes meal planning easier, not inevitable. The key distinction is between an unplanned low-protein plant diet and a diet designed to provide adequate protein and essential amino acids.

A randomized trial challenges the "animal is anabolic" shortcut

A 2025 randomized crossover trial by Domic et al. compared a well-balanced vegan diet with an omnivorous diet in active older adults. It found equivalent daily mixed muscle-protein-synthesis rates. This is experimental evidence on a surrogate outcome, not mortality, but it directly shows that plant-based eating does not inherently impair daily muscle protein synthesis when the diet is well designed in active older adults.

The qualifier matters. The trial does not establish the same result for sedentary, frail, or institutionalized adults, who may face larger barriers to eating enough protein. It also does not show that vegan and omnivorous diets produce equal long-term survival. It answers a narrower question: whether a balanced vegan pattern can support daily mixed muscle protein synthesis under the trial's conditions.

Exercise may carry more weight than fine-tuning leucine. In a 2025 randomized trial in older women consuming optimized protein, resistance training increased basal muscle protein synthesis and reversed frailty measures, while leucine supplementation alone did not. A 2025 meta-analysis of seven randomized trials involving 349 older people with osteosarcopenia likewise found that strength training improved body composition and physical performance across protein-intake levels.

What broader substitution studies can—and cannot—tell us

A 2022 systematic review of 12 observational studies found that isocalorically replacing animal-protein calories with plant-protein calories was generally associated with better aging-related health outcomes. That broader adult evidence helps explain why plant-forward guidance is common. But the review covered adults broadly older than 18 and did not report age-stratified analyses above 65. It cannot resolve the conflict between cohorts focused on people in their seventies and eighties. It also excluded clinical trials by design.

What to do with the uncertainty

• Prioritize adequate total protein intake. The evidence here supports adequacy as more important than choosing a side in the animal-versus-plant argument.
• Include resistance training when health status and clinical guidance allow it. The randomized evidence gives training a stronger footing than leucine supplements alone.
• Plan plant-predominant meals deliberately. Larger or complementary servings may help meet protein and leucine needs that animal foods supply more densely.
• Prefer minimally processed protein foods where possible. The evidence links ultra-processed-food intake with frailty and sarcopenia, although it does not isolate protein source as the cause.

No age threshold at 65, 70, or 75 currently identifies animal protein as the preferred source for mortality or muscle preservation. Nor does the evidence show that plant protein becomes ineffective after 70. The mortality cohorts disagree, and the strongest direct dietary trial addresses muscle protein synthesis rather than death. For now, the defensible target is a diet that delivers adequate protein and essential amino acids, paired with resistance exercise and centered on foods a person can eat consistently. That is less marketable than a protein-source verdict, but it is where the evidence lands.