Fasting-mimicking diet

• Get the benefits of a multi-day fast without a water-only fast.
• Lower weight and body fat, especially visceral fat, while sparing muscle.
• Lower IGF-1, insulin and glucose, signals linked to aging and cancer in animal studies.
• Improve blood sugar control in type 2 diabetes and lower blood pressure.
• Trigger cell renewal and lower biological age, as claimed by its developers.
• Make cancer treatment more effective and less toxic, under study in trials.

The fasting-mimicking diet (FMD) is a fixed meal plan, not a general eating style. In the trial that put it on the map, day one supplied about 4,600 kJ (about 1,100 kcal), with 11% of energy from protein, 46% from fat and 43% from carbohydrate. Days two to five supplied about 3,000 kJ (about 720 kcal) a day, with 9% protein, 44% fat and 47% carbohydrate [1]. Fat is mostly unsaturated, and protein and sugar are kept low on purpose.

People did it for five consecutive days a month, for three months, and ate normally the rest of the time [1]. Later trials repeated it monthly for a year [2].

The commercial version is ProLon, sold by L-Nutra, a company Longo founded. The kit is a box of soups, nut bars, olives, kale crackers, herbal teas and a glycerin drink. A journalist who tried it reported just over 800 kcal a day, a price of $250, and five hungry, irritable days.

Why "mimicking"?

Longo's group noted that patients tolerate water-only fasting poorly, especially during cancer treatment [15]. The FMD was built to keep the signals of fasting, low glucose, low insulin and low IGF-1, while giving enough food to be tolerable [5].

The core idea is that nutrient-sensing pathways, chiefly IGF-1, PKA and mTOR, switch cells from growth into protection and repair when food is scarce, and that refeeding then triggers rebuilding [16].

In mice, four-day FMD cycles shrank several organs, then raised stem and progenitor cell numbers when the mice ate again. Twice-monthly cycles started in middle age extended lifespan, cut cancer incidence and visceral fat, and improved memory [5]. FMD cycles regenerated insulin-producing beta cells in diabetic mice [17], eased a mouse model of multiple sclerosis [13] and reversed colitis damage [14].

In people, the measured effects are smaller and mostly in blood markers:

• IGF-1 fell by 21.7 ng/mL after three cycles, against a rise of 8.7 ng/mL on usual diet [1]. A meta-analysis put the average fall at 19.2 ng/mL [3].
• Glucose and growth factors fell consistently in 101 patients with cancer [11].
• Insulin resistance and liver fat were lower after three cycles in a secondary analysis [6].

Note what has not been measured in people: tissue regeneration and stem cell renewal. Those claims come from the animal work. Autophagy has now been measured once: a 30-person pilot sponsored by L-Nutra found a blood-cell marker of autophagic flux rose after six days, though not at every time point [4].

Healthy adults. In the first randomized trial, 100 adults did three monthly cycles or ate normally. The FMD group lost 2.6 kg against a 0.1 kg gain, with less trunk fat, 4.5 mm Hg lower systolic blood pressure and lower IGF-1. Fasting glucose, LDL cholesterol and CRP did not change significantly for the whole group, though a post hoc analysis found larger effects in people with risk factors [1]. A quarter of the FMD arm dropped out before the end of the randomized phase.

Type 2 diabetes. The strongest trial added a monthly FMD to usual primary care for 12 months in 100 people on metformin or diet alone. The FMD group needed less glucose-lowering medication, and still had HbA1c 3.2 mmol/mol (0.3 percentage points) lower and weight 3.6 kg lower than controls. Glycaemic control improved in 53% of the FMD group against 8% of controls [2]. MRI showed less visceral and subcutaneous fat with no significant loss of abdominal muscle [7]. A smaller German trial saw lower insulin resistance but no overall difference in kidney damage markers compared with a Mediterranean diet [8].

Pooled evidence. A 2025 meta-analysis of 11 randomized trials found lower HbA1c, IGF-1 and blood pressure (4.1/2.3 mm Hg), and no significant effect on lipids or other cardiovascular risk factors [3].

Biological age. A secondary analysis reported a 2.5-year fall in median biological age after three cycles [6]. That was exploratory, used a blood-marker estimate rather than outcomes, and came from the developer's lab.

Behavior. Diet quality between cycles did not change in the diabetes trial, but FMD participants became more active, and many described the cycles as a prompt for other changes [18].

The cancer idea came from mice: short fasts protected normal cells, but not cancer cells, from high-dose chemotherapy, and ten patients who chose to fast around chemotherapy reported fewer side effects [12]. Longo's group then argued that FMDs could make treatments work better while protecting normal tissue [15].

The key human test is the Dutch DIRECT trial. 131 women with HER2-negative breast cancer were randomized to an FMD for three days before and on the day of each chemotherapy cycle, or their regular diet [10].

• Toxicity, the primary endpoint, did not differ between groups.
• Tumor response on imaging was more frequent with the FMD (odds ratio 3.2), and in a per-protocol analysis a near-complete pathological response was more likely (odds ratio 4.1).
• Adherence was the problem: only 20% of FMD patients complied in every cycle, and a planned phase 3 extension was dropped.

A single-arm trial in 101 patients found the FMD safe alongside standard treatments, with lower glucose and immune changes linked to stronger antitumor responses [11]. In mice, FMD cycles made hormone therapy for breast cancer work better, and patients on hormone therapy showed the same drop in insulin, leptin and IGF-1 [9].

None of this means people with cancer should start an FMD on their own. Weight and muscle loss can be dangerous during treatment, and the trials ran with oncology teams.

Marketing and social media often describe the FMD as a way to trigger autophagy, clear damaged cells, renew stem cells and turn back the clock. Here is how those claims map to evidence.

Outside scientists quoted in 2017 press coverage made the same point: the human base was thin and mostly animal, and there were no side-by-side comparisons with other fasting or diet programs. Chris Kresser, who uses a homemade version with patients, has said the benefits have been overstated relative to the human data.

Readers should know who stands to gain, because it shapes how much weight to give each study.

• The company. L-Nutra was founded by Longo in 2009 and sells ProLon kits and related foods. Press coverage in 2017 reported that Longo kept majority ownership.
• The university. USC licensed intellectual property to L-Nutra and may receive royalties. L-Nutra supplied the diet used in the trials [1] [6].
• The pledge. Longo and a coauthor hold equity in L-Nutra and did not take part in collecting or analyzing data in the 2024 analysis; the paper states that 100% of Longo's equity will go to charitable organizations, including the Create Cures Foundation he founded [6].
• Trial funding. The 12-month diabetes trial was co-funded by L-Nutra along with public Dutch funders [2]. In the breast cancer trial, Longo had equity in L-Nutra and another author had shares in a company that invested in it [10]. Authors of key reviews hold FMD patents under negotiation with L-Nutra [15].

None of this makes the results wrong. The Dutch diabetes trial was assessor-blinded and run in primary care [2]. But the field would be stronger with more trials run independently, against active comparators such as a Mediterranean diet or simple calorie restriction.

Against a water fast, the FMD keeps some protein and calories coming in. That matters because about two thirds of the weight lost in 5 to 20-day water fasts is lean mass [19]. In mice with colitis, FMD cycles reversed gut damage that water-only fasting did not [14], though that comparison has not been repeated in people.

Against intermittent fasting and the ketogenic diet, the main difference is rhythm. The FMD concentrates restriction into five days a month and leaves 25 or so days of normal eating. Whether that beats daily approaches is unknown, because the trials mostly compared the FMD with usual eating rather than with another diet [1] [2]. For context, daily calorie restriction produced a 7.5 kg loss over two years in the CALERIE trial [20], and intermittent and continuous restriction give similar weight loss [21]. The one trial with an active comparator, a Mediterranean diet in diabetes with kidney damage, found no difference in its main outcome [8].

Searches for "fasting-mimicking diet and keto" often ask whether the FMD puts you in ketosis. It is designed to push metabolism toward a fasting state, but published human trials report glucose, insulin and IGF-1 more consistently than ketones [1] [11].

The trials used L-Nutra's packaged plan. If you build your own, the published targets are the reference point: about 1,100 kcal on day one, then about 720 kcal a day for four days, with protein near 9 to 11% of energy, fat 44 to 46% (mostly unsaturated) and carbohydrate 43 to 47%, mostly from vegetables [1].

Typical DIY menus use:

• vegetable soups and broths,
• small portions of nuts and olives,
• vegetables such as kale, mushrooms and squash,
• herbal tea and water,
• no meat, fish, eggs, dairy, added sugar or alcohol.

Whether a homemade plan matches the packaged one depends on hitting the protein, sugar and calorie targets, which is hard to do by eye. Weighing food and logging it in an app is the most reliable check.

How often?

The healthy-adult trial used three monthly cycles [1]. The diabetes trial used monthly cycles for a year under medical supervision [2]. Longer-term schedules for healthy people, such as a few cycles a year, have not been compared in trials.

Exercise during the five days

Light activity such as walking is the common advice. There is little published data on hard training during an FMD cycle, and low intake makes heavy sessions harder and riskier.

• Adding meat, eggs, protein shakes or sugary snacks, which breaks the low-protein, low-sugar design.
• Going far below the target calories, which turns the plan into a water fast.
• Taking glucose-lowering drugs at usual doses without medical advice, which risks low blood sugar.
• Expecting regeneration or anti-aging effects that have only been shown in mice.
• Bingeing on day 6, which wipes out much of the calorie deficit.
• Doing it once and expecting lasting changes; trials used repeated monthly cycles.

• Pregnancy or breastfeeding.
• A current or past eating disorder.
• Underweight (BMI under 18.5) or frailty.
• Diabetes treated with insulin or sulfonylureas, unless a clinician supervises and adjusts medication.
• Active cancer treatment, except within a trial or with your oncology team.
• Kidney or liver disease, or other serious chronic illness, without medical advice.
• Children and teenagers.

• Fatigue, weakness and headache, the most common complaints in trials.
• Hunger and irritability during days 2 to 5.
• Dislike of the food, the main reason patients stopped in the cancer trial.
• Low blood sugar in people on glucose-lowering drugs.
• Dizziness from lower blood pressure.