Fasting-Mimicking Diet

The fasting-mimicking diet (FMD) is a five-day, low-calorie, low-protein plant-based meal plan designed by Valter Longo's lab to trigger fasting-like metabolic responses while still providing some food, unlike a water-only fast. Commercialized as ProLon, it is meant to be repeated periodically, typically monthly.

The key human RCT, led by Wei and colleagues and published in Science Translational Medicine, randomized 100 participants to three monthly 5-day FMD cycles versus an unrestricted diet [1]. The FMD group showed reduced body weight, trunk fat, and total fat, lower blood pressure, decreased IGF-1, and improved lipid and CRP markers, with larger improvements concentrated in participants who started at higher risk on those markers [1]. This pattern, bigger benefit in people who need it more, is consistent with a metabolic-reset mechanism rather than a universal effect.

A 2024 Nature Communications secondary analysis by Brandhorst, Longo, and colleagues pooled two trials (NCT02158897 and NCT04150159) and reported that three FMD cycles were associated with a 2.5-year decrease in median biological age, independent of weight loss [2]. The same analysis modeled a life-expectancy increase from 82.93 to 83.73 years and a reduction in modeled 20-year all-cause mortality risk at age 70 from 81.1% to 57.5% [2]. The University of Southern California's Longevity Institute press release repeats the same 2.5-year figure and quotes Longo directly [3].

What the evidence does not show: those life-expectancy and mortality-risk numbers are modeled projections from biomarker changes, not observed outcomes. No trial has followed FMD participants for 20 years to confirm the modeled mortality reduction actually happened. A 2025 Nutrition Reviews review notes that periodic fasting-mimicking diets extend lifespan by roughly 11% in mouse studies, but cautions that human lifespan extension from FMD remains unproven, with only biomarker proxies studied to date [4]. The gap between mouse lifespan data and human biomarker data is exactly the gap that separates a promising intervention from a proven one.

Who this does not work for, or where the evidence is weakest: FMD trials generally exclude pregnant women, people with type 1 diabetes, anyone underweight, and people with a history of eating disorders, given the low-calorie design. The commercial ProLon product's specific formulation has primarily been tested by Longo's own affiliated research group and collaborators, so independent replication by unaffiliated labs is thinner than the volume of Longo-affiliated publications might suggest. The biological-age reduction finding, while statistically significant in the reported analysis, comes from an epigenetic clock, itself an imperfect proxy for true aging rate, and the modeled mortality figures assume that clock change translates linearly into survival benefit, an assumption not directly tested.

Mechanistically, FMD is proposed to trigger autophagy, reduce IGF-1 and insulin signaling, and promote a stem-cell-based regenerative response during the five-day cycle, based on mouse data from Longo's earlier work, with the metabolic marker improvements in the human RCT offered as indirect human support for the same pathways [1][5].

Plain takeaway: FMD has real RCT-backed short-term metabolic benefits, and a well-conducted secondary analysis links repeated cycles to slower biological aging by one epigenetic clock, but the commonly cited life-expectancy and mortality-risk figures are model projections, not measured outcomes, and human lifespan extension from FMD has not been directly shown.