Protein After 40: Why More Is Not Always Better, and Less Is Not a Longevity Hack

Protein has become a longevity Rorschach test. Muscle-focused advice says to eat more; FGF21 and animal-longevity research is often translated into "eat less." For adults over 40, neither slogan survives contact with the evidence. The useful dose depends on what you eat now, whether you train, and whether your kidneys are healthy.

Established evidence: protein helps most when it corrects a shortfall

The adult protein RDA of 0.8 grams per kilogram of body weight per day is designed to cover the needs of nearly all healthy adults. It is not a demonstrated optimum for every goal, as the National Academies' protein reference report makes clear. One pooled acute tracer analysis estimated that older men may need about 0.4 g/kg in a meal to maximize muscle protein synthesis. That estimate comes from a short-term surrogate, however, not years of strength or disability outcomes (Moore et al., 2015).

Longer trials show why baseline matters. In the 2021 PROMISS randomized trial, 276 older adults who were eating below 1.0 g/kg/day were advised to reach at least 1.2 g/kg/day. After six months, the protein-only advice group completed a 400-meter walk 12.4 seconds faster and produced 32.6 newtons more leg-extension force than controls.

Yet a separate one-year trial in 208 healthy adults over 65 found that raising intake from about 1.1 to 1.5 g/kg/day with supplements did not improve muscle size, strength, or function. A 2026 trial in 295 frail older adults reached the same practical boundary: adding whey and counseling to resistance training produced no overall advantage, although exploratory analyses suggested more benefit among people starting below 1.2 g/kg/day.

The established conclusion is not "more protein." It is "avoid too little protein, then train." Resistance exercise supplies the stimulus; protein supplies material. Once intake is adequate, another scoop may add nothing.

Established evidence: kidney status changes the target

For people with normal kidney function, research has not established that moderately high protein intake causes chronic kidney disease. In the 164-person OmniHeart crossover trial, six weeks at 25 percent of calories from protein increased cystatin C-based estimated filtration by about 4 mL/min/1.73 m² compared with 15 percent protein. That could represent adaptive hyperfiltration; the study could not establish long-term safety or harm.

An 11-year Nurses' Health Study analysis of 1,624 women found no association between higher protein and filtration decline in those starting with normal function. In women with mild impairment, however, higher total and non-dairy animal protein were associated with faster decline. It was observational, but the contrast matters.

Known chronic kidney disease is therefore a different problem. KDIGO's 2024 guideline suggests about 0.8 g/kg/day for adults with stages G3 to G5 and advises those at risk of progression to avoid more than 1.3 g/kg/day. Very-low-protein diets belong under close clinical supervision, especially because undernutrition and sarcopenia are competing risks. Kidney status cannot be inferred from feeling well; eGFR and urine albumin-to-creatinine ratio are the relevant clinical markers.

Emerging evidence: FGF21 is a signal, not a longevity target

FGF21 is a liver-derived hormone involved in adaptation to nutritional stress. In a 2014 human experiment with only eight or nine participants per group, 28 days of protein restriction sharply increased circulating FGF21. A controlled five-week study published in Nature Metabolism in 2025 likewise found that healthy lean men needed more energy to maintain weight on a protein-restricted diet and had higher FGF21.

The human signal is real; the longevity interpretation is not established. A July 2026 randomized feeding study in just 17 men with overweight or obesity compared 0.9 with 1.8 g/kg/day for five weeks. The lower-protein group lost 2.0 kilograms without prescribed calorie restriction and raised FGF21, but lean mass also showed a non-significant downward tendency. The trial measured short-term metabolism, not strength, healthy aging, or lifespan.

A separate 2026 Cell Metabolism paper found better healthspan markers with a low-amino-acid diet in aged mice. Its human component was cross-sectional, not an intervention, and the paper reports relevant commercial interests. Nor can findings about FGF21 in infant growth be imported into a resistance-trained 58-year-old. There is no human trial showing that deliberately raising FGF21 through protein restriction extends life.

Personal experimentation: run an eight-week protein audit

Personal experimentation is not a fourth level of evidence. It is a bounded way to learn whether an established range improves an outcome you can observe.

1. Set the safety boundary. Do not self-test a high- or low-protein diet if you have diagnosed kidney disease, unexplained abnormal kidney markers, or a clinician-prescribed protein target.
2. Measure before changing. Log seven ordinary days. Calculate grams per day and g/kg/day. Record two function markers, such as same-load repetitions on a squat or row and five-chair-stand time.
3. Correct a likely shortfall. If you average below 1.0 g/kg/day and your goal is strength or function, 1.2 g/kg/day is a defensible eight-week test because it has been used in older-adult trials. It is not a universal prescription. If you already consume 1.2 to 1.6 g/kg/day, evidence does not justify automatically going higher.
4. Spread and substitute. Divide the target across three meals. Roughly 0.3 to 0.4 g/kg per meal is a practical heuristic, not a magic threshold. Replace lower-value calories rather than stacking shakes on top. Use a mix of legumes, soy foods, dairy, eggs, fish, or poultry. Prospective cohort evidence favors replacing some processed meat protein with plant protein, though it cannot prove causality (Song et al., 2016).
5. Apply a training stimulus. Use progressive resistance training two or three times weekly. Keep the program stable enough to interpret the nutrition change.
6. Review observable results. Track load or repetitions, chair-stand time, weekly body weight and waist, hunger, digestion, and adherence. Clinician-ordered eGFR and urine albumin are safety data, not performance scores. Continue only if the intended marker improves without an unwanted trend.

Mini-case: Elena's 84-gram test

Elena is a hypothetical 58-year-old who weighs 70 kilograms, strength-trains twice weekly, and averages 58 grams of protein, or 0.83 g/kg/day. Her test target is 84 grams. She shifts breakfast from 10 to 25 grams, lunch from 18 to 29, and keeps dinner near 30, while replacing rather than adding calories.

For eight weeks she tracks goblet-squat repetitions at a fixed load, five-chair-stand time, waist, body weight, and digestive comfort. The decision at week eight is based on those results, not on whether she reached an impressive number.

Errors to avoid

• The RDA-is-optimal error: treating 0.8 g/kg as the best target for every active adult.
• The FGF21 scoreboard error: assuming a higher hormone level means more longevity.
• The mouse-to-human leap: copying a low-amino-acid rodent diet without human outcome data.
• The powder-without-stimulus error: adding supplements while resistance training stays absent or inconsistent.
• The silent-kidney assumption: escalating protein without accounting for known CKD or abnormal eGFR or albuminuria.
• The dinner-dump error: leaving nearly all protein for one meal when a workable distribution is available.

Your next action

Start with a seven-day average, not a supplement order. Established evidence supports correcting low intake and pairing protein with resistance training. Emerging FGF21 research is worth watching, but not acting on as a longevity protocol. A personal experiment earns its keep only when strength, function, body composition, tolerance, and safety markers are observable.

Sources