Fasting-Mimicking Diet and the Thyroid: What's Known

The Fasting-Mimicking Diet (FMD) is a 5-day low-calorie protocol with metabolic and immune-modulatory signals. Trials show transient T3 drops; safety in Hashimoto on levothyroxine is unstudied. Consult your endocrinologist before trying it.

Why a "fast" affects the thyroid

The thyroid axis is deeply tied to energy availability. When the body senses scarcity, the deiodinase enzymes that convert inactive T4 into active T3 down-regulate, and reverse T3 — a metabolically inert form — rises. The classical demonstration came from Spaulding 1976, where switching healthy adults from a normal diet to caloric restriction produced a measurable T3 drop within days, with reverse T3 climbing in parallel [C5]. This is not a thyroid disease; it is the thyroid axis behaving as designed.

The Fasting-Mimicking Diet (FMD), developed by Valter Longo's group at USC, was designed to provoke the same biological response as multi-day water-only fasting while still providing a small amount of food. The protocol is 5 consecutive days of a plant-based, very-low-calorie diet — roughly 1,090 kcal on day 1 and 725 kcal on days 2 to 5 — followed by 25 days of normal eating, repeated monthly in trials [C1][C2]. The macronutrient mix is low in protein and low in sugar, which is critical: the IGF-1 / mTOR / PKA signalling that FMD aims to lower depends specifically on amino-acid and glucose deprivation, not just calorie reduction [C4].

In trial volunteers, three monthly FMD cycles produce a coordinated pattern — lower fasting glucose, lower IGF-1, lower CRP, modest weight loss, and a transient drop in T3 during the 5-day cycle that rebounds during the normal-eating window [C1][C2]. Animal data show that the same protocol triggers hematopoietic stem-cell turnover and may reset some autoreactive immune cells [C3][C4]. These are striking signals — but they were all generated in healthy or metabolically unhealthy adults, not in patients with autoimmune thyroid disease or on levothyroxine.

What the trials actually show

The two largest published human trials are Brandhorst 2015 (38 healthy adults, randomized crossover) and Wei 2017 (100 generally healthy adults, 3 monthly cycles, parallel group) [C1][C2]:

• Body composition. Modest fat-mass and waist-circumference reduction after 3 cycles; lean mass largely preserved [C1][C2].
• Cardiometabolic markers. Fasting glucose, triglycerides, total cholesterol, and CRP fell after 3 cycles in participants who started elevated; little change in those already healthy [C2].
• IGF-1 and IGFBP1. IGF-1 dropped during the 5-day cycle and rebounded; IGFBP1 (an IGF-1 inhibitor) rose [C1][C2][C4].
• Thyroid axis. T3 dropped transiently during the 5-day fast (an expected adaptive response), with TSH variability and recovery within the 25-day refeed window [C1][C5]. Antibody titers (TPO, Tg) were not measured in these trials.
• Immune signals. Animal models showed reduced autoimmune EAE (a multiple sclerosis model) severity and partial myelin regeneration after repeated FMD cycles [C3]. A small human MS cohort embedded in the same publication showed feasibility but not efficacy [C3].

What's missing is what Hashimoto patients most want to know: does FMD lower thyroid antibodies? Does it reduce flares? Does it change the long-term levothyroxine requirement? No published trial has answered any of these questions.

What might recover or shift on adequate levothyroxine during FMD

If a patient with treated hypothyroidism cycles FMD, two things can happen to thyroid labs in the short term [C5][C6]:

• TSH may rise modestly during and after the 5-day cycle because peripheral T4-to-T3 conversion drops and the pituitary senses lower T3.
• Free T4 and free T3 may dip transiently during the 5-day window — this is the same physiological pattern seen in non-thyroid illness and prolonged caloric restriction [C5].

In a healthy person, these shifts normalize during refeed. In a Hashimoto patient on levothyroxine, the lab pattern can be misread as under-replacement, prompting a dose increase that would then become over-replacement once the patient returns to normal eating. The ATA hypothyroidism guideline specifically cautions against making dose changes based on labs drawn during caloric stress or acute illness — these are not steady-state numbers [C6].

When FMD signals do NOT translate to Hashimoto

Several gaps between FMD trial data and Hashimoto care matter [C3][C6][C8]:

1. No randomized trial in autoimmune thyroid disease. The autoimmune signal in FMD is strongest in EAE (a mouse MS model) and in cellular markers — none of this has been replicated for thyroid antibodies in humans [C3].
2. Levothyroxine absorption during FMD is unstudied. The 5-day low-calorie window changes gastric emptying, bile acid flow, and meal timing — variables that all influence T4 uptake. No trial has measured peak T4 levels during an FMD cycle [C6].
3. Underweight and frail patients were excluded from FMD trials. Hashimoto patients with low BMI, eating-disorder history, or poor caloric reserve are the exact population FMD was never tested in [C1][C2].
4. Cycling stress on the HPA axis may worsen fatigue and sleep in patients whose thyroid is already producing subclinical hypothyroid symptoms [C6][C8].

What does NOT help

• Extending the FMD past 5 days ("more fasting must be better"). Brandhorst and Wei specifically tested 5-day cycles. Longer fasts in untreated hypothyroid patients can produce sustained T3 suppression without rebound [C1][C2][C5].
• DIY FMD without medical supervision in Hashimoto. The proprietary ProLon kit was designed to hit very specific macronutrient ratios. A self-assembled "1,000-calorie plant diet" is not the same intervention and has not been studied.
• Using FMD instead of levothyroxine. No trial has tested FMD as a substitute for thyroid hormone replacement, and there is no biological rationale for one [C6][C8].
• Repeating cycles more often than monthly. Trials used 1 cycle/month for 3 months. More-frequent cycling has not been studied for safety [C1][C2].

Practical guidelines

1. Talk to your endocrinologist before starting any FMD cycle. Bring the published Brandhorst and Wei papers if needed [C1][C2]; this is not a routine discussion most clinicians have prepared answers for.
2. Do not stop or skip levothyroxine during FMD. Continue the usual fasted-morning dose with water; levothyroxine absorption requires an empty stomach, which the FMD morning routine accommodates [C6].
3. Avoid labs during the 5-day cycle or in the first week after. Wait at least 2 to 4 weeks after the cycle ends to draw TSH/FT4 — the transient drop in T3 will distort the picture and your endocrinologist will not be able to interpret it as steady-state [C5][C6].
4. Watch for hypoglycemia, dizziness, palpitations, and severe fatigue. Stop the cycle and reach your clinician if these occur. Hashimoto patients with adrenal-axis comorbidities are at higher risk [C6][C8].
5. Skip FMD if you are underweight, pregnant, breastfeeding, planning pregnancy, have an eating-disorder history, or are over 70. These were exclusion criteria in the trials [C1][C2].
6. Recheck thyroid labs 4 to 6 weeks after finishing the cycle series so any sustained shift in TSH or FT4 can be addressed with a dose review [C6][C7].

Frequently asked questions

Will FMD cure my Hashimoto? No. No FMD trial has tested patients with Hashimoto thyroiditis, and no intervention currently reverses the underlying autoimmunity that drives the disease [C6][C8]. The animal data on autoimmunity is interesting but does not transfer directly to humans.

Will FMD lower my TPO antibodies? This has never been measured in a published FMD trial. The autoimmune signal exists in mouse MS models and in cellular markers, not in human thyroid antibody titers [C3].

Can I stay on levothyroxine during an FMD cycle? Yes — and you should. Stopping levothyroxine during a 5-day low-calorie cycle can precipitate hypothyroid symptoms. Continue your usual fasted-morning dose with water [C6]. Your endocrinologist will decide whether monitoring is needed.

Should I expect my TSH to change after FMD? Trial data show transient T3 drops during the 5-day cycle with recovery during refeed [C1][C5]. In a treated hypothyroid patient, a TSH check during or immediately after the cycle is unreliable — wait several weeks for steady state before drawing labs [C6].

Is there a safer way to get the "fasting" signal without 5 days of restriction? Time-restricted eating (eating windows of 8 to 12 hours) has been studied separately and is generally well-tolerated in treated hypothyroidism — see our intermittent-fasting-thyroid article. It produces a much milder version of the metabolic signal without the multi-day caloric deficit.

Bottom line

FMD is a serious metabolic intervention with measured signals on glucose, IGF-1, inflammation, and immune cell turnover [C1][C2][C3][C4]. The thyroid response — a transient T3 drop during the 5-day cycle — is an expected adaptive shift seen in any sustained caloric restriction [C5]. What is unknown is the most important question for Hashimoto patients: whether FMD changes antibody titers, flare frequency, or the long-term levothyroxine requirement [C6][C8]. Until trials answer those questions, FMD in autoimmune thyroid disease is investigational. Your endocrinologist will be able to assess whether your individual situation makes the unknown risks acceptable, and will adjust monitoring around any cycles you do [C6][C7].