Valter Longo’s group at USC put old mice (starting at 20 months, roughly a 60–65-year-old human) on five diets and found that a mostly plant-based, low-protein “longevity diet” only worked well when it contained a moderate dose of the amino acid methionine (LDMM). Too little methionine and the mice wasted away; the right amount stripped fat, improved insulin sensitivity, cut frailty and preserved muscle — without cutting calories. Mechanistically the diet raised growth hormone (GH), GLP-1 and FGF21 while lowering IGF-1, an unusual hormonal signature driven by the liver. The diet strongly improved healthspan but did not significantly extend lifespan versus a healthy control diet. A companion re-analysis of 205,000 people suggests high animal-protein (hence high-methionine) diets track with roughly double the diabetes prevalence.
For two decades the anti-aging field has treated protein — and especially the sulphur amino acid methionine — as something to minimise. Cut methionine and rodents live longer. But there is a catch that longevity enthusiasts rarely discuss: severe methionine restriction can also drive muscle wasting and frailty, which is precisely the fate that shadows the long-lived-but-frail elderly of Southern Europe and Okinawa.
The new work from Valter Longo’s lab reframes the problem. Instead of asking “how low can methionine go,” the team asked what specific amount lets a low-protein diet deliver its metabolic benefits without triggering wasting. Their answer is a formulation they call LDMM — a low-amino-acid, high-carbohydrate, olive-oil-rich diet modelled on traditional Mediterranean and Okinawan eating, spiked back up to a moderate 0.3% methionine (roughly half the level in standard chow, but about double what classic methionine-restriction studies use).
The Big Idea is that a single amino acid acts as a hepatic switchboard. In the liver, LDMM raises growth hormone and its downstream signal STAT5, but instead of the usual result — more IGF-1 and more growth — the low-methionine context flips the liver toward a fat-burning, longevity-associated program driven by PPAR-alpha and the hormone FGF21. IGF-1, a growth signal repeatedly linked to cancer and accelerated aging, actually falls. GLP-1, the same incretin targeted by Ozempic-class drugs, rises. The upshot in aged mice: substantial fat loss despite the animals eating more calories, better glucose control, less frailty, better grip strength and rotarod performance, and no loss of lean mass — the exact side effect that plagues GLP-1 drugs.
The honest headline, though, is that these were healthspan wins, not lifespan wins. Started at 20 months, LDMM did not significantly outlive a healthy control diet. Its statistically significant “lifespan extension” was only relative to a junky Western diet and to fasting cycles — comparators that were themselves impaired. The authors are candid that starting so late in life likely capped any longevity payoff.
Finally, the team reinterpreted three Harvard cohorts totalling over 200,000 people. Those eating the most animal protein — and the most methionine — had about double the prevalence of type 2 diabetes, despite otherwise healthier habits, hinting that the mouse mechanism may echo in humans.
Actionable Insights
The take-home is a composition rule, not a magic supplement: a mostly plant-based, low-protein diet appears to deliver its metabolic upside only when methionine sits in a moderate window — neither maximised (animal-heavy) nor crushed to near-zero (classic MetR). In the mice, the moderate-methionine diet produced its benefits without calorie restriction, which is the practically important part for real-world adherence.
Effect sizes worth anchoring on:
- Fat loss: In short-term wild-type experiments, LDMM cut fat mass by roughly 60% from baseline while lean mass was preserved. In aged HET3 mice it produced the largest fat-mass reduction of any group (all comparisons p<0.0001).
- Insulin sensitivity: Significantly better glucose and insulin tolerance than every other group, including the healthy control (GTT p=0.002; ITT p<0.0001).
- Frailty/strength: Lower frailty index and superior body-weight-adjusted grip strength vs control, Western and keto diets (p=0.006 to p<0.0001).
- The human signal: Highest vs lowest animal-protein intake tracked with roughly 2x (+106%) diabetes prevalence and +34% overweight/obesity, alongside +81% higher methionine intake.
Caveat before you act: this is a mouse-mechanism paper with a human association bolted on. There is no human trial of LDMM. The realistic move is dietary-pattern-level — more legumes, vegetables and olive oil, less animal protein — not home methionine dosing.
Context / Source
- Paywalled paper: Methionine-supplemented longevity diet increases growth hormone, GLP-1, and FGF21; reduces frailty; and promotes healthspan.
- Authors / Institution: Fanti, Brandhorst, Navarrete, … Malik, Hu, Longo (corresponding). Longevity Institute, Leonard Davis School of Gerontology, University of Southern California (USA), with collaborators at Harvard T.H. Chan School of Public Health (USA), University of Toronto (Canada), Children’s Hospital Los Angeles, and University of Campinas (Brazil).
- Journal: Cell Metabolism (Cell Press / Elsevier), Vol. 38, August 4, 2026.
- Impact evaluation: Cell Metabolism’s Journal Impact Factor is approximately 37 (JCR 2025; CiteScore ~40.7). The impact score of this journal is ~37, evaluated against a typical high-end range of 0–60+ for top general/biomedical science, therefore this is an Elite impact journal